l ‘ -, ~ 21‘ «My A...- (Al- “\\\ s ' ‘ r“ -.D\“algllf_ 3:21 uvcnwc 1’ Hill): 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book return to remove charge from circulation records THE CHEMISTRY OF BIS-(TRIMETHYLSILYL)KETENE AND A RATE DECOMPOSITION STUDY OF BIS-SILYLATED ACETATES By Behrouz Cyrous 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 CHEMISTRY OF BIS-(TRIMETHYLSILYL)KETENE AND A RATE OF DECOMPOSITION STUDY OF BIS-SILYLATED ACETATES By Behrouz Cyrous Methyl, ethyl, and is0pr0pyl bis—(trimethylsilyl)- acetate were prepared in 80-88% isolated yield by treating bis-(trimethylsilyl)ketene with the apprOpriate alcohol in the presence of methanesulfonic acid at room temperature. The esters were converted quantitatively to the corresponding lithium enolates by reaction with one equivalent of lithium diis0pr0pylamide (LDA) in tetrahydro- furan (THF) at -78°C. The THF solutions are stable for at least one hour at -78°C, but decomposes to bis-(trimethyl- silyl)ketene at higher temperatures. The less substituted members of the series decompose faster than the more highly substituted enolates; R=Me> Et> 1-03H7> t-CLI'H9 Bis—(trimethylsilyl)ketene was reacted with methyl- lithium, n-butyllithium and tert-butyllithium at room temperature to give the correSponding¢x,(x , -bis-(trimethyl- silyl)ketone enolate. o! .0‘ -bis-(trimethylsilyl) ketones were obtained in 75-85% yield by quenching the reaction mixtures with dilute acetic acid or H20. The lithium enolate of(X,O(-bis-(trimethylsilyl)- acetone was reacted with trimethylchlorosilane to give Behrouz Cyrous the corresponding silyl enol ether in quantitative yield. This enolate was also reacted with benzaldehyde to give the ot-silylvinyl ketone . The lithium enolate oftx,a -bis-(trimethylsilyl)- acetone failed to react with aldehydes or ketones possessing one or moreM—C-C—C-C-OR (2) I I I 012' | _ I e o 0 H30 my I u I n HC-C-C-C-OR .d-keto ester (b) An inverse Claisen condensation which entails removing a proton from the solvent or from amine and rapid condensation of the ester thus formed with another molecule of enolate3 (eq. 3). I a-keto ester (c) Initial decomposition to a ketene which then con- denses with a second molecule of enolate to give the observed 3-keto ester product (eq. 4). o l n I 9 \ M-C-C-OR ?MI 8 M-(‘l-C-OR————>MOR +/c=c=o I 1 C—C-C-C-OR (4) , . I 3 I 9 H3d5 ’ Ho- c- -c- -c- OR ,3- -keto ester A study of the kinetic behavior of the-decomposition of ester enolates, which showed first order kinetics,” provided evidence against direct enolate coupling (pathway a). The ketene intermediate (pathway c) to A-keto ester formation was suggested by Vaughan5 for the self conden- sation of the reagent prepared from ethyltX-bnflmOFAN isobutyrate and zinc metal (eq. 5). CH3? C35 Ban-IC- C- OEt -—-+’BanOEt + (CH 3)ZC=C=O BanC-COZEt CH CH 3 EB OZnBr 3 CH 0 CH H30 CH3 3 3 CH3 - o - c - OEt (5) H;C-C-C-fi-0Et 4r—* ,.c _ 0 CH ——-C 3 Ketene intermediates have also been proposed for the Elcb mechanism in the hydrolysis of malonic and,3-keto esters.6-lO 2 Recently, bis-(trimethylsily1)ketene was isolated from the decomposition of lithium bis-(trimethylsilyl) acetate enolate.l9 Since the direct A-elimination of a metal alkoxide is a very rare occurrence, and the trapping of the ketene 11 was often not possible (for example, ketene intermediate trapping with dimethylamine or detection by an IR study of the solution was unsuccessful for lithio isobutyrate and lithio t-butyl acetate decomposition)?2 it was of interest to study this reaction in more detail. In par- ticular, it was of interest to know what effect the R group in the alcohol portion of ester enolate has on the rate of elimination of metal alkoxide, M-OR. Since the elimination of MOR to form ketene from most . ester enolates is followed by a further rapid condensation of ketene, it was desirable to choose an ester which would produce a stable ketene, thereby simplifying the rate determination of the initial elimination. Both di-tert- butylketene,” z, and bis-(trimethylsilyl)ketene, 1, are reported to be very stable and unreactive to nucleophilic attack. (CH ) C Me 8' 3 :C:C:O 3 L:C:C:O (CH3)BC Me3Si Z l However, the esters leading to di-tert-butyl ketene are reported to be inert to amide bases such as LDA, presumably for steric reasons.1u Consequently, we decided to synthesize for our study a series of esters derived from bis-(trimethylsilyl)acetate, g. MeBSu O The synthesis of tert-butyl bis-(trimethylsilyl)- acetate (g. R=t-buty1) by successive C-silylation (attach- ment of silyl group to the carbon atom of tert-butyl acetate) is reported in the literature15’16 (eq. 6, 7). 1) LDA, ~78°C CH3C020(CH3)3 2) me $101 11, MeBSiCHZCOZC(CH3)3 (6) 3 98% 1) LDA, ~78°C M93Si\ MeBSiCHZCOZC(CH3)3 t’ "CH-CO§+- + 2) MeBSiC1 MeBSi OSiMe 70% Me 8' - / 3 \ 30% D‘A— The preparation of methyl, ethyl and isopropyl bis-(tri- methylsilyl)acetates by this procedure would be a difficult task because methyl, ethyl, and iSOprOpyl acetates in the first step give mostly the 0-sily1ated (attachment of silyl group to oxygen atom) product, Q, rather than the C-silyl- ated product17, 2 (eq. 8). For example: methyl acetate 4 gives 65% O-silylated and 35% C-silylated product. 1) LDA, -78°c ,OSi(CH3)3 . CHBCOZR 2) (CH ) s101” CH2=Q\ + (CH3)351\/COZR 3 3 0R (8) Q 2 (major product) (minor product) R=Me, Et Also, in the second step of the silylation of 2, O-silyla— tion would be a problem sincecx—substituted acetate enolates (except for tert butyl esters) give almost exclusively the O-silylated compound IQ;8 (eq. 9).. . o 1) LDA. -78°C (CH ) 31 . 0 v 3 3 .\CH-C-OR + 2) (CH3)BS101 (CH3)BS1 C-silylated ester . ,,OSi(CH (CH3)BS1CH=C\‘ (CH3)BSi\V,COZR / (9) 3’3 OR 19 O-silylated ester The O-silylated products obtained in the first and second step cannot be used for our study. A reasonable preparation of methyl, ethyl and iSOprOpyl bis-(trimethylsilyl)acetates would be the reaction of bis- (trimethylsily1)ketene with the corresponding alcohol (eq. 10). HQ (MeBSi)2C=C=O + ROH ——> MeBSi\ ,CHcozR (10) R=CH3,Et, iso-propyl MeBSl This stable ketene is readily available by the decomposition of lithio tert-butyl bis-(trimethylsilyl)acetate19 (eq. 11). LDA,-78°C Pi 25°C (Me Si) once-i— fl(Me s1) coo ——> (Me s1) C=C=o 3 2 2 THF 3 2 2 30 min. 3 i ) ll +-Lio-}— RESULTS Preparation of Bis-(trimethylsilyl)ketene Tert-butyl acetate was reacted with an equivalent of lithium diisopropylamide (LDA) in tetrahydrofuran (THF) at —78°C. The reaction mixture was stirred for 10 minutes at -78°C to complete enolate formation. One equivalent of trimethylchlorosilane was added and stirred for an addi- tional 30 minutes to convert the enolate to tert-butyl (trimethylsilyl)acetate (eq. 12). _+_ LDA,-78°C MeBSiCl CH 00 >_ LiCH co—i— D Me SiCH co 3 2 THF 2 2 -78°C.THF 3 2 2 ll 98% (12) Quenching the solution with dilute hydrochloric acid is not desirable because, not only does remaining acid or water in the mono-silylated tert-butyl acetate, ll, destroy the newly formed enolate in the next step, but also the. separation and distillation of ll is time consuming. Therefore, the solution of ll_without quenching was trans- fered to an equivalent of LDA in THF at -78°C. The solution was stirred for 15 minutes to convert the monoesilylated (tert-butyl acetate to the corresponding lithium enolate, 1;, 7 (eq- 13)- -78°c Li Me SiCH CC + LDA —DMe Sincoii— + HN CH(CH ) (13) 3 2 2 THF 3 3 2 lg 2 An equivalent of trimethylchlorosilane was added to the solution and the solution was stirred for 30 minutes at -78°C and for another 15 minutes at room temperature to convert 1g to tert-butyl bis-(trimethylsilyl)acetate 12 (70%) and l& (30%) (GQ- 14)- OSiMe / \ 0 3 ICHC02+— + MeBSch : c 1) LDA,-78°C MeBSi 2 ‘ 2) MeBSiC'l ’ MeBSi \ +_ (14) 3) H3069 11 (70% 1‘1 (30%) After quenching with 3M‘HCl and distillation, compound 13 was isolated in 60% yield. Compound 13 was recovered and converted to the mono- silylated ester 1; (eq. 15) in the quenching step. ll_was then recycled for bis-silylation. / OSiMe3 H306 MeBSiCH=C -——————A> MeBSiCHZCOZ+—- (15) a \0‘1— Q Pure 11 was treated with one equivalent of LDA in THF at -78°C for 15 minutes, then the solution was warmed to room temperature and stirred for 30 minutes to convert the enolate 15 to bis-(trimethylsilyl)ketene, 1, (eq. 16, 17). O . (M93Si)ZCHCO§+—- LDA, -78 o.’ Me381\ 9 THF Ll;p-C-0 (16) MeBSi 8 l5 30 min. Me BSi\ 1; e—’ /C=C=o + 1110‘)— (17) 25 C MeBSi L Quenching with even dilute hydrochloric acid destroyed at least some of the ketene. Quenching with distilled water separated the ketene l from the aqueous layer without loss. Separation and careful distillation of the organic layer under reduced pressure gave a 60% yield of bis-(trimethyl- silyl)ketene. Some of the ketene was found in the evaporated solvent. From this, it was concluded that using a lower boiling solvent and dialkylamine in the reaction would simplify the separation. With this in mind, diis0propylamine (B.P. 85°C) was replaced with the relatively volatile diethylamine (B.P. 55°C); THF (B.P. 67°C) was replaced with diethylether (B.P. 34°C). Then diethylether and diethylamine were eva- porated at 0°C under reduced pressure. The ketene was dis- tilled under reduced pressure using an oil bath at approxi- mately 30°C. With this improved procedure the yield of isolated ketene was increased from 60% to 80% (eq. 18). 1) LiNEt2,-78°C (Me Si) CHCO .. (Me Si) C=C=0 (18) 3 2 2f" 2) R.T., 30 min. 3 2 80% isolated yield Preparation of Bis-(trimethylsilyl) acetates Bis-(trimethylsilyl)ketene was added to an equivalent amount of the corresponding alcohol at room temperature without any solvent. No addition product was found after 9 reacting overnight. Only starting materials were found in the solution mixture. The reaction was only slightly catalyzed by glacial acetic acid, but it was catalyzed by stronger acids such as concentrated sulfuric acid and methanesulfonic acid at room temperature. Bis-(trimethylsilyl)ketene was added dropwise to an equivalent amount of the corresponding alcohol containing 0.01% mole equivalent of methanesulfonic acid. The solution was stirred for 1-2 minutes. GLC analysis of the solution confirmed that the ketene was converted to the corresPonding bis-(trimethylsilyl)acetate (eq. 19). CHBSOBH, R.T. Me38i\ 9 (Me3Si)ZC=C=0 + ROH p _ /CH-C-0R (19) 1-2 mins. Me3Si R=Me, Et, iso-C3H7 g The colorless aux bis-(trimethylsilyl) acetates, g, are stable at room temperature indefinitely. The products g were distilled under reduced pressure, with isolated yields ranging from 80% to 88%. The results obtained with methanesulfonic acid are summarized in Table 1. 10 Table 1. Preparation of Bis-(trimethylsilyl) acetates Reaction Isolated ROH Condition Product % yield MeOH No solvent (Me3Si)20HCOZCH3 ’ 80 R.T. l m1n. EtOH No solvent (Me 3Si)2CHC02Et 81 R.T.:1_min. ISOprOpyl No solvent (Me 3Si)20HC0§CBH 7 88 "OH RCTC l min. CH3S03H(0.01%) (Me3Si)2C=C=0 + ROH . (MeBSi)2CI-IC02R ROT. l min. 1 eq. ' 1 eq. R = Me, Et, lSO-C3H7 Decomposition study of Lithio tert- -butyl Bis-(trimethylsilyl) acetate Tert—butyl bis-(trimethylsilyl)acetate was added to an equivalent amount of LDA in THF at -78°C. The solution was stirred for 10 minutes to form the corresponding lithium enolate 15 (eq. 20). Me35i\ .0 -78°C Me3sf\ .9 /CH- -C- -0+ + LDA ——9 L1;C-C-O+— (20) Me3Si THF M9381 15 15 is stable in solution at -78°C for at least 2 hours, but decomposes to bis-(trimethylsilyl)ketene and lithium tert-butoxide within 30 minutes at room temperature (eq. 21} l\ “ THF, R. T. 3Li- /C- -0 -fi> (Me Si) C: C: 0 + LiO (21) 3 2 30 min. 3812 ll Although the decomposition of 15 is slow enough at room temperature to follow the rate of decomposition, because of the fast decomposition of the methyl analogues, even at 0°C, a temperature of -23°C was Chosen for a com- parison of rates. Five,#l of a THF solution of the lithium enolate of tert-butyl bis-(trimethylsilyl)acetate at -78°C, was withdrawn and injected quickly onto the GLC. Only bis- (trimethylsilyl)ketene was observed, in quantitive yield by internal standard. Similar behavior was observed for the other lithium enolates of the bis-silylated acetates. From these observations, it was concluded that lithium enolates of bis-(trimethylsilyl)acetate decompose to bis- (trimethylsilyl)ketene on the GLC column (SE-30.5% on chromosorb w). Therefore, a THF solution of enolate 15 was warmed to -23°C in a CCln-Dry Ice bath. An internal standard was added and the rate of decomposition was followed by removing a small aliquot with a cooled syringe and adding the aliquot to sufficient water to quench the enolate to the ester. After addition of pentane and drying over anhydrous magnesium sulfate, the amount of ketene formed was measured by GLC analysis. Results are shown in Table 2. 12 Table 2. The Rate of Decomposition of Lithio- tert- -butyl bis (trimethylsilyl) acetate at -230 C. Time (min. ) en81ate stirred at -2 %1ketenef9;med 7-5 13 30 28 60 33 100 37 200 50 Li -23°C (Me 3Si)ZCC02‘-— -—-> (Me 3Si)ZC= --=0 0 + Li0+— 12 These data are presented graphically in Figure 1, by plotting the percent of produced ketene versus time. These data show that at -23°C,3.3 hours are needed to convert half of the lithium enolate 15 to ketene. Decomposition Study of Lithio—isopropyl Bis-(trimethy1silyl) acetate Is0pr0py1 bis-(trimethylsilyl)acetate was.converted quantitively to the correSponding lithium enolate 16, by treating with one equivalent of LDA in THF at -78°C for 10 minutes (eq. 22). 0 o . Me Si\ LDA,- 8 C Me 31 0 3 \CH'E'O / 7 iv 3 '\ u 0 x/ 2 MGBSi/C ‘\ THF Ll/C-C- ‘-\\ (2 ) Me 81 3 -.1§. 13 The enolate 16 is stable at -78°C for at least 1.5 hr. As the temperature is increased to 25°C, the enolate decomposes to bis-(trimethylsily1)ketene and lithium isopropoxide (eq. 23). Li -78°C to 25°C (MeBSi)2 C=C=0 + Li0-<: 2-< (MeBSi)2C-CO 4 (23) 1g The THF solution of this enolate16_was warmed to -23°C in a CClu-Dry Ice bath and an internal standard was added. At appropriate intervals, ketene formation was determined as before. The results are presented in Table 3. Table 3. The Rate of Decomposition of Lithioaisopropyl Bis (trimethylsilyl) acetate at -23 C. *Ifiime (min) eno ate stirred at -23 C % ketene formed 6 15 10 20 20 27 [+0 36 60 41 120 51 . Li _2300 . . (M6381)2 C-002-< —> [(CH3)331] 2 C=C=0 1 L10-< 16 1.“ .saev ease ea .oommn pm mpwpmomadmaflmachmafihpvImam szsmupnwe canpfla mo scavwmoQSOCCQ mo mpmm .H cm on cm on c: on ON mmsmfim OH 4 d W u - + osmpmm R 1 0.— cm oh The data show that, after approximately 2 hours at -23°C, half of the Lithio-is0pr0pyl bis-(trimethylsilyl) acetate has decomposed to ketene. The ' data are presented graphically in Figure 2 as before. Rate of decom osition ofLithio-Ethyl and Methyl-Bis (trimethylsilyl) acetates Ethyl and methyl bis-(trimethylsilyl)acetatesr were treated similarly. The results are presented in Tables 4 and 5. Table 4. The Rate of Decomposition of Lithio—Ethyl-Bis (trimethylsilyl) acetate at -23 C. Tiflirfigéngte?3§3§e % ketene formed 5 29 10 ' [+3 20 52 40 54 60 69 Li -23°C THF (MeBSi)2 C’COZ Et - ( Si)2 C=C=0 + Lio Et These data show that less than 20 min. is needed for the decomposition of half of the lithio ethyl bis-(trimethyl- silyl)acetate. and less than 5 minutes for the methyl ester at -23°C. These data are presented graphically in Figures 3 and 4. . l6 .odmmu Pm mpdpmomfiahaflmahsmeanvImam azmoumomH ownpwq mo Cofipfimomsoomm e0 opmm .m mpsmflm A.:Hev mafia om . on cm on on on om 0H u q _ a - - a d 1 3381 a .. 02 Table 5. The Rate of Decomposition of LithiooMethyl Bis-(trimethylsilyl)acetate' at -23 C. ngirggénéi Egglgte % ketene formed 2 25 4 41 5 5o 10 65 15 70 2° 75 40 A ~81 Li -23°C,THF (M93Si)2 C‘Coz CH3 ,, (Me3Si)2 C=C=0 + L10 CH3 18 .oommu Pm mpmpoomAdhaamamnpoafiupwumdm Hanna oanpwq Ho cowpwmoQSOCCQ mo ovum .m musmfim A.:wsv mafia on on on om 0H . a . _ . . . 8381 s 1 ooa .oommu pm mpmvmomaahaflmaznposwppvImam Hznpms oazpfiq Mo CowpflmoQSoomQ e0 mpmm .: mpzmfim A.sfisv mafia o: mm om mm om ma oa m J _ J A _ 1 u — _ assess s 1.00H DISCUSSION Preparation of 0(,0( -B1s-(trimethylsilyl)acetates Trialkylchlorosilanes react with ketone enolates to 20-24 produce 0-silylated products exclusively (eq. 24). Ketones substituted on thecXposition with trialkylsilyl groups are unstable relative to their .0-silylated isomers25 and rearrange to the 0-silylated isomer under the influence of heat or catalyst26 (eq. 25). 1) LDA 9SiMe3 2) MeBSiCl 0 or OSiR / 3 R381 \/'\ R A » CH2=C\ (25) Hglz'll R Silylation of ester enolates with trialkylhalosilanes can occur on either the carbon or oxygen atom (Fig. 5). o d3 e, u / CH C-OR ‘——-D CH = C 2 2 \ OR Figure 5. Ester enolates resonance structures. The ratio of C to 0§sily1ated product depends heavily on the structure of the ester enolate used (eq. 26). 21 OLi Me SiCl O OSiMe \ / 3 _ I n \ / 3 .C=C ~=D Me Sl-C-C-OR + C=C (26) / \ 3 | / \ OR OR It has been reported17 that the reaction path is deter- mined by steric factors: substitution on the alcohol portion of the ester favors C-silylation while substitution on the alpha carbon of the ester favors 0-silylation (eq. 27, 28). OSiMe 1) LDA,-78°C o / 3 CH300§*— A MeBSi\/“\ 04'— + CH2=C (27) 2) MeBSiCl \0 98% 2% 3 3 1) LDA’_780C CH30 H CH3\ /OSiMe3 :CH-C-OEt . 1 M6381: -C- C- OEt + /C=C\ .(28) CH3 - 2) MeBS1Cl CH CH3 OEt 3 1% 99% Tert-butyl acetate enolate gives almost exclusively C-silylated ester while ethyl isobutrate gives exclusively 0-silylated product. Methyl, ethyl and isopropyl acetates give mostly (more than 60%) 0-silylated products. We thought that the preparation of methyl, ethyl, and isopropyl bis- (trimethylsilyl) acetate for our decomposition rate studies of the corresponding lithio enolates would be difficult, because silylation of the lithium enolate of these acetates would give mostly 0-sily1ated products. Furthermore, silylation of the mono-silylated acetates to produce the desired bis-silylated acetates would be much 22 more difficult, because as mentioned above, (x-substituted acetate enolates usually give almost exclusively O-silylated product. Therefore, methyl, ethyl, and is0propyl bis-(tri- methylsilyl)acetates 2 are unlikely to be produced in a sufficient yield by successive silylation of correSponding acetates in the manner of the tert-butyl analogues (eq. 29). 1) LDA, 78° c . 1) LDA,-78°C CH 3005}— , Me381CH2C02-w— . w 2) Me 3SiCl 87 2) MGBSlCl 9 . . ,ossLMe3 . (29) Me381CH=C\O + (Me381)ZCHCO§+- 30% 70% Methyl, ethyl and is0pr0pyl bis-(trimethylsilyl)- acetates were prepared by addition of bis-(trimethylsilyl)- ketene to methanesulfonic acid in a solution of the appro- priate alcohol (eq. 30). CH3803H(cat.) Me38i\\ 3 (Me351)20=c=o + ROH . 1,, ‘/CH-C-OR (30) R.T. l min. MeBSi R=Me, Et, isopropyl 2 To observe the effect of the R group on the rate of decomposition of the enolates of these esters, 2 was con- verted to the corresponding lithium enolate by treatment with LDA in THF at -78°C. It has been reportedln that the carbon analogues l2 are inert toward LDA (eqs. 31, 32). M93Si“CH 3 OR LDA -7800 M638?\ 3 - - -+ -————-+>I - _ _ 2 -. Me351 23 Me30\ 8 -78°C CH-C-OR + LDA -—+ N.R. (32) Me3C.( THF 11 This is probably due to two factors: (1) The Si-C (1.872) bond is longer than C-c (1.543)27 bond which decreases steric hindrance: in the case of the silicon analogues. (2) The enhanced acidity of a Si—C-H group, compared with C-C-H, may also be an important factor in the greater reactivity of silicon analogues towards base. Such acidity is due to the vacant 3d-orbitals on the silicon atom which are of suitable energy for back bonding with a filled 2p- orbital on the adjacent carbon atom, thus stabilizing the adjacent carbanion.28‘ Decomposition Study of Lithio Bis-gtrimethylsilyl)acetates. All of the lithium enolates of bis (trimethylsilyl) acetate 3 which we studied were stable at -78°C in THF solution for at least 1 hour (eq. 33), but as the temperature . Pi -78°c , THF 1130‘B . (Me351)20002R 1 hr?—_> (Me381)2CI-ICOZR (33) 2 98% R = CH3, CZHS’ i-C3H7' t-CuH9 was increased, the enolates decomposedto the stable bis (trimethylsilyl)ketene- and lithium alkoxide (eq. 34). 24 R.T. 3 _. (1v1e35:1)‘2 c=c=o + LiOR (34) Three mechanisms have been suggested for converting simple ester enolates to.6-keto esters (Figure 6). CO R—*HCCO R It? * GC-CO-CCO R I \ ' -MOR _.__ CflCO R c M?COZR———> —ogo _,__2 Figure 6. Three Proposed Pathways of Ester Enolates Decomposition. A study of the kinetic behavior in the decomposition of ester enolates“ provided evidence against direct enolate coupling (path a). Path a requires that the decomposition be second order in enolate concentration, while the kinetic study showed the decomposition to be first order. A ketene intermediate has been proposed for the decomposition of zinc enolates. The reaction of this intermediate with the ester enolate is believed to produce a dimer, 1g, which upon hydrolysis, yields‘S-keto esters5 (eq. 35). \ /OZnX \ XZn-CCOZR )c=c\ ——-—vXZnOR + c=c=o ., OR / I (.1an ' “ ‘ e c': - 5 (35) HC‘I-C-C‘J-COZR H39 \c\:// 8 Ketene is produced in the thermal decomposition of 25 silicon ester enolates29 (eq. 36). \ / 3 AA \ C=C '——————D' C=C=O + CH OSiR (36) / \OCH3 / 3 3 Bruice and Pratt8 have pr0posed an E1cb mechanism involving a ketene intermediate in the hydrolysis of ortho and para-nitrOphenyl acetate esters (eq. 37). 0 base 3N® o- c- CH-C- OEt : g€>mcio ‘Lc c- OEt . R ‘1 8 .9 8 (37) EtOC-C'IH-C-B 4_B__ so a + Etoc-o=c=o R N0 R 2 l2 They reported that this ketene intermediate, 12, was trapped using aniline buffers. 9 Rebek and co-workers have reported a technique for the detection of reaction intermediates from an insoluble poly- meric precursor. The intermediates are detected by trapping them on a second solid phase suspended in the same solvent. An acyl transfer to g; has been observed when the precursor 2Q was treated with triethylamine in dioxane (Figure 7). Unlike a classical trapping eXperiment, a direct reaction of precursor and trapping agent is physically prohibited in this three-phase system. This method is especially suited for the detection of nucleophilic catalysis. 26 e @— CH2 solid phase 9 .9 @— CHZ‘QOC-CHZ-C-OET H\C«—-C/0Et reagent .4 ‘§O - solution 04 n I? ' @— CHZ-NHZ @— CH§NHCCHZCOEt solid phase Figure 7. Three Phase System for the Trapping of Reaction Intermediates. Ketene 22 has been trapped by this procedure in the decomposition of 23 (eq. 38). H CH base NO2 O\75/CH é- N°2 Gamer" H3 ._ 15’" H3707 if (38) @° ° 21 o// . g_2_ In the reaction between isoalkyloxazolines and n-butyl- 10 lithium, Meyers and co-workers have suggested a ketene- imine intermediate (eq. 39) which can be trapped with trimethylchlorosilane. o n-BuLi,O°C o ' OLi MeBSiCl ’ "’ .__. ,_4, N/’ < H THF N)76 N=C=C ‘ OSiMe (39) C C N: : \ 27 Decomposition of thiol ester enolates to ketene or ketene polymers also has been reported.30 Preliminary evidence indicates that the sequence of events in the forma- tion of mercaptide ion involves initial removal of an (X-proton from thiol ester followed by the decomposition of .this intermediate into a mercaptide ion and a ketene (eq. 40). 0 u ' _ CH3 C-SR + LDA——>RS + ketene ——_. polymer l—x n'~ ' (40) R-s-Rl-fi-T' Under the basic conditions involved, ketene polymer- ization occurs readily.31’32 Thus, when a thiol ester of diphenylaeetic acid was used, a gum having characteristics of a diphenylacetic acid was obtained.30 However, there is no report in the literature on what effect R groups have on the rate of decomposition of thiol ester enolates. Also, isolation of stable ketene from decomposition of hindered thiol ester enolates has not been reported. Decomposition studies of a wide variety of acetate enolates have also shown some evidence to support ketene 12 Moreover, bis-(trimethylsilyl)ketene has intermediates. been successfully isolated from the decomposition of lithio bis-(trimethylsilyl)acetate.. Since there was some ambiguity in the decomposition studies of ester enolates and because the directA-elimination of a metal alkoxide is a very rare occurrence, it was of interest to study this 28 reaction in more detail. The objective of this study was to determine the rate of decomposition of different bis-silylated acetate enolates and to observe what effect the R group on the alcohol portion of these enolates had on the rate of decomposition. Data from Tables 2, 3, 4 and 5 indicates that the rate of decomposition of the enolates at -23°depends on the size of the R group. Comparison of the relative rates of decomposition between methyl, ethyl, is0propyl and tert-butyl bis-(tri- methylsilyl)acetate enolates, clearly indicates that the methyl ester enolate decomposition is the fastest while tert-butyl ester enolate decomposition is the slowest of the series (Figure 8). Me Si\ Me Si 3 3 \ Li-C-C0- CH:3 Li—CCO Et Li—CCO Li—CCO 2 2 / 2 23 hr MeBSi Me 3Si MeBSi ilife time-—> 20 min. . 3.3 hr. at -23°C Figure 8. Com arison of Half Life Stability (T% of Bis-(trimethylsilyl) acetate Enolates. As Figure 8 shows, the rate of decomposition for bis- silylated acetate enolates decreases as the alkyl substi- tution of R increases. The relative rate of decomposition is presented graphically on the following page (Figure 9). 29 hepeaesm eaanpag eepaaaaam-hem Haeam-p use HameumehH .Haapm .Hagpez me ceapahemseoeo me apex .m ehswam . o on om 0H A case away om mm) mm om ww _ a .14 a .13 0 .4 \l mu! r. Hapam-p ..o: I * HhmoumomH a, 1.00 Haapm « l ow Assam: ecepem e..ooH If steric relief is assumed to be the driving force in the formation of ketene, then the tert-butyl bis-silyl- ated ester enolate should be the least stable of the series, because the tert-butyl bis-silylated enolate is more hindered than the others. Therefore, other factors must be involved in the decomposition reaction. In studying lithio-tert butyl acetate, Woodbury12 observed that the rate of decomposition decreased in the presence of lithium— t-butoxide (eqs. 41, 42). THF/25° c LiCHZCOd— W —L—> CH3C02—+— ((+1) THF/25° c H 23% LiCHzcoj— + Lio+———> 2 hrs. —L—> (3130021— (42) 86% He concluded that the ketene formation was reversible (eq- 43)- H LiCHZCOj— .: H>c:=c:=o + Lio+— (43) As expected, an increase in the lithium t-butoxide concentration would force the equilibrium towards the enolate and slow the decomposition rate. This does not occur with ethyl acetate enolate, probably because ethoxide anion is a weaker base than tert-butoxide anion. If this occurs in the decomposition of lithio bis-(trimethylsilyl) acetates (eq. 44), then one would eXpect that with increasing PKb of RO-, the equilibrium would be driven 31 to the left in eq. 44. Li 1 I .____. (MeBSi)2CC02R -— (Me351)2c=c=o + LiOR (44) 2 Since bis-(trimethylsilyl)ketene was isolated quan- titively from the decomposition of corresponding enolates, it can be assumed that the reaction proceeds to completion. This is probably due to the steric effect of bis-(trimethyl- silyl)ketene which prevents the nucleophilic attack of alkoxides on the ketene. As the basicity of RO- increases, the corresponding enolate should become more stable, the relative PKb's of the leaving groups are: ‘o+ > ‘o—< > 'OEt > 'OMe If enolate decomposition is related to the basicity of the leaving group, then enolate decomposition should be in accordance with Figure 8. An ambiguity arose when we plotted the logarithm of the decomposition rate of enolates versus time. Contrary to our expectations, data from Tables 2, 3, 4 and 5 did not fit a straight line which would confirm the first order decomposition of enolates studied. In conlcusion, it seems that the stability of bis- silylated acetate enolates increases with increasing basicity of the leaving group (RD-). 32 EXPERIMENTAL '1. Materials 3.913.622 Tart-butyl acetate was commercially available from Aldrich Chemical Company, and used without further puri- fication. Ethers Anhydrous diethylether, was commercially available and used without further purification. . Tetrahydrofuran (THF), was commercially available from Aldrich Chemical Company. THF was distilled from the sodium ketyl of benZOphenone and stored under argon. Amines DiiSOprOpylamine was obtained from Aldrich Chemical Company. It was used after distillation from calcium hydride (bp 84°C, 760 mmHg) and stored under argon. Diethylamine was obtained from Matheson Coleman & Bell Company. It was distilled from calcium hydride (bp 55, 760 mmHg), stored under argon and molecular sieve. n-butyllithium n-butyllithium was obtained as a 1.5M hexane solution 33 from Aldrich Chemical Company and titrated prior to use by the Watson-Eastham method.33 Trimethylsilylchorosilane Trimethylchlorosilane was obtained from Aldrich Chem- ical Company. It was distilled '?(bp 57°C, 760mmHg) and stored under argon. Alcohols Absolute methanol, ethanol and is0propanol were obtained from Aldrich Chemical Company and used without further purification. 2. Preparation of tert-buty;_§;s-(trimethylsiiyllaeegate= A dry 500 ml side arm round-bottomed flask was equipped with a magnetic stirring bar, septum inlet, gas inlet valve and mercury bubbler. The flask was flushed with argon and charged with 100 ml of distilled THF and 13.4 ml (100 mmole)‘ of distilled diisoprOpylamine was injected by syringe. The flask was cooled to 0°C in an ice bath and 66.7 ml of 1.5 M solution of n-butyllithium was added drOpwise over 10 minutes. The ice bath was removed and.f1ask was immersed in a.Dry Ice-acetone bath. 13.4 ml (100 mmole)Q of tert- butyl acetate was added dropwise over a period of 10 minutes. The solution was stirred an additional 10 minutes. Then 12.8 ml (100 mmoles) of distilled trimethylchlorosilane was injected drOpwise over 10 minutes. The solution was stirred for another 15 minutes. The Dry Ice-acetone bath was removed and the solution in the flask was transfered by 34 Teflon tubing to a 1000 ml round-bottomed containing 100 m moles lithium diiSOprOpylamide under argon in 100 ml THF at -78°C. The solution was stirred for 15 minutes and another 12.8 ml (100 m mole) of distilled trimethylchloro- silane was added dropwise over a 10 minute period. The mixture was stirred for 30 minutes at -78°C; Then the dry, ice—acetone bath was removed and the solution was stirred for 15 minutes at room temperature. 100 ml of 3 M aqueous HCl was injected.‘ The aqueous layer was separated in a 1000 ml separatory funnel and discarded. THF and diis0pr0py1amine were evaporated from the organic layer under reduced pressure. The mono silylated tert-butyl acetate was distilled under reduced pressure (bp 67°C, 13 mmHg) and was used for further silylation. The tert-butyl bis-(trimethylsilyl) acetate was then dis- tilled under reduced pressure (bp 61°C, 0.4 mmHg). The yield was 14.4 g (60%) of pure tert-butyl bis-(trimethyl- si1y1)acetate. 3. Preparation of bis-(trimethylsilyl)ketene A dry 500 m1 side-armed round-bottomed flask was equipped with a magnetic stirring bar, septum inlet, gas inlet valve and a mercurry bubbler. The flask was flushed with argon and 125 ml of anhydrous diethyl ether was in- jected into the flask. 15.6 ml (150 m mole) of diethyl- amine was added. The flask was immersed in an ice bath. Then 100 ml of n-butyllithium (1.5M, 150 m mole) was added 35 by syringe over 15 minutes to the stirring solution. After stirring for an additional 10 minutes, the flask was put in a dry ice-acetone bath and 39 g (150 m mole) of pure tert-butyl bis-(trimethylsilyl) acetate was added over a 10 minute period. The reaction mixture was stirred for an additional 20 minutes at -78°C, warmed to room temperature and stirred 30 minutes longer. 100 ml distilled water was injected into the flask. The two layers were separated in a separatory funnel. The organic layer was poured into a 500 ml side armed round-bottomed flask equipped with a magnetic stirring bar, septum inlet, a short vertical con- denser and a gas inlet valve on t0p of the condenser. The flask was immersed in an ice bath and stirred. The solvents (hexane, diethylamine and diethylether) were removed under reduced pressure. The ice bath was removed for 30 minutes and the solvent evaporation was completed. The remaining ketene residue was distilled under reduced pressure (bp . 20°C, 2 mm Hg) and afforded 20.2 g (80%) of pure bis (tri— methylsilyl) ketene. lH""NMR(CCI,+)S'0.25(S) IR (neat) 2085cm-l and 1295em'l (Lit 2085, 1295em'l). 4. Preparation of Methyl bis-(trimethylsilyl)acetate; A dry 25 ml side armed round-bottomed flask was equipped with a magnetic stirring bar, septum inlet, gas inlet valve and a mercury bubbler. The flask was flushed with argon and was charged with 0.8 ml (20 m mole) of methanol and 5 micro liters (0.01 mole) of methanesulfonic acid as a 36 catalyst. Then 4.0 ml (20 m mole) of pure bis-(trimethyl- silyl)ketene was added dr0pwise to the solution at room temperature. GLC analysis of the solution on 6 foot 5% SE-30 column confirmed that all of the ketene was converted to methyl bis-(trimethylsilyl)acetate within one minute. Bulb-to-bulb distillation of this solution under reduced pressure (.95 mm, 49-510C) gave 3.5 g ( 80%) of methyl bis-(trimethylsilyl)acetate. Product Analysis: bp 49-51°0/0.95 mm, Density 0.911 IR (neat), 1700 our1 1 strong C=0 absorption H NMR (001),): g 0.14 (s, 18H), g 1.53 (s. 1H),g3.46(s, 3H) 5. Preparation of Ethyl bis-(trimethylsilyllacetate A 25 ml side-armed round-bottomed flask was equipped with a magnetic stirring bar, septum inlet, gas inlet valve and a mercury bubbler. The flask was flushed with argon and charged by 1.1 M'ml (20 m mole) of absolute ethanol and 5 micro liters (0.01 m mole) of methanesulfonic acid as a catalyst. Then 4.00 ml (20 m mole) of pure bis-(tri- methylsilyl)ketene was added dr0pwise to the solution at room temperature. After stirring 1 minute, GLC analysis of the solution of 61, 5% SE-30 column confirmed that the ketene was converted to ethyl bis-(trimethylsilyl)acetate. Bulb-to-bulb distillation of this solution under reduced pressure gave 3.9 g ( 81%) of pure ethyl bis-(trimethyl- sily1)acetate. 37' Product Analysis: B.P. 57-59°0/ 1.1 mm, Density = 0.8519 IR(neat) 1720 om‘l strong C=0 absorption lH NMR(CC14):SP 0.23 (s, 18H),(2.3 (t, 3H),{1.6 (s, 1H) S 4.1 (q. 2H) 6. Preparation of Is0pr0pyl bis-(trimethylsily12aceta§81 The same procedure was applied as with the methyl and ethyl analogues. Bulb-to-bulb distillation of the solution in final step gave 4.32 g (87.8%) of isopropyl bis-(tri- methylsilyl)aoetate., Product Analysis: B.P. 63-65°c/ 1.1 mm, Density = 0.8621 1 strong absorption for Céo IR(neat) 1720 cm- 1H NMR (0014).g 4.85(S, 1H),g 1.5 (s,,1H),g 1.2(d, 6H) 8 0.15(s, 18H). 7. Rate of Decomposition of gis-(trimethylsilyl) acetate L1thium Enolate The following procedure is representative for all rate decomposition studies of the lithio bis-silylated acetates.. A dried 10 m1 side armed round-bottomed flask was equipped with a magnetic stirring bar, septum inlet, gas inlet valve and a mercury bubbler. The flask was flushed with argon and charged with 1 m1 g dry THF. 0.155 (1.0 m mole) freshly distilled diiSOpropylamine was added. The flask was immersed in an ice bath. Then 0.73 ml of a 1.5 M 38 solution (1.1 m mole) of n-butyl lithium was added dr0pwise. After a few minutes the flask was placed in a dry ice- acetone bath. Then 1 m mole of desired bis-silylated acetate was added dr0pwise and stirred an additional 10 minutes at -78°C. The flask was then placed in a carbon tetrachloride-dry ice bath (-23°C) and 0.21 ml (1.00 m mole) of undecane was injected into the solution as the internal standard. At appropriate intervals, approximately 0.1 m1 aliquots of the solution were removed using a cooled syringe and quenched by injecting it into approximately 50 m1 H20 in a small vial. Approximately 0.15 ml of pentane was added into the vial. The solution was dried over anhydrous magnesium sulfate and analyzed for ketene. The results are presented in Tables 2, 3, 4 and 5. 39 CHAPTER II REACTIONS OF BIS-(TRIMETHYLSILYL)KETENE INTRODUCTION The annelation of 2-a1kylcyclohexanones 23 with methyl. vinyl ketones 25 is an important route to the fused poly- cyclic system 2§,3u (eq. 45). While condensation at the less substitutedlm carbon of the 2-alkylcyclohexanone can be carried out efficiently using the corresponding enamines,35 the conditions required for reaction at the more substituted site result in low yields due to polymerization of the vinyl ketone36 (eq. 45). This difficulty is avoided by the use of alkyl halides instead of vinyl ketones,37’38 but a number of steps are then required to transform the added alkyl group into the 3-keto alkyl function. R 0 2.4. 25 26 The problem with simple vinyl ketones, such as methyl vinyl ketone, stems from the similar base strengths and reactivities of the enolate ions derived from the starting material and the Michael adduct (1, 4 adduct). 40 There is good evidence that silicon with its vacant 3d orbitals can stabilize an adjacent negative charge39’u'o'u'l and canf' be easily removed after cyclization. Taking advan- tage of this principle, Storkuz has reported the synthesis and use ole-silylated vinyl ketones 2] (eq. 46) for the annelation of ketones. 0Li 0 CH =C/MgBr 1) RCHO. R \/”\"/S:LE1:3 ‘01?3 2 . . . 21 , \‘SiEtB 2) 0x1dat10n .. 4‘ However; this preparation oftx-silyl vinyl ketones, requires (46) 11 several steps from commercially available vinyl silanes.‘3 The synthesis ofcx-silyl vinyl esters 28 A by reacting lithio-tert-butyl bis-(trimethylsilyl) acetate 25 (obtained by reaction of the correSponding ester, 23, with LDA (eq. 47))and aldehydes (eq. 48) was reportedla. -78°c ' Pi 0Li [SiMe SiMe THF 3 ./ 3 25 + RCHO ———> R- CH—C- 0-02 ———-> RCH=C -78°C I \CO (48) SiMe3 2+— + LiOSiM:3 The reagent, 25, failaito react with ketones. Thus, addition of cyclohexanone to a solution of 25 gave only recovered tert-butyl bis (trimethylsilyl) acetate and the lithium enolate of cyclohexanone (eq. 49) which was trapped 41 by trimethylchlorosilane. O MeBSiCl OSiMe ow _. » t We thought that the ketone analogue of 25 (i.e. 4) would be more efficient and probably would be able to convert both aldehydes and ketones to their corresponding¢0H-c-co-CH3_,¢CH=C (59) 5 mln' SiMe3 COCH3 .2 The same procedure was followed with acetaldehyde and propionaldehyde. The products were identical to the 46 quenching product,n2 both by GlC retention time and fiNMR (eq. 60). /OLi R , 9 3 2; + R ,OLi 1)c=c\ R H C. With Ketones. One equivalent of methyl lithium was added to an ethereal solution containing one equivalent of bis-(tri- methylsilyl)ketene at room temperature to complete the reaction. Then one equivalent of acetone was added and the reaction mixture stirred for one hour. Analysis of the reaction mixture by GlC gave the quenching product, a .0( - bis-(trimethylsilyl)acetone (eq. 61). 0Li . _ z' . (MeBSl)ZC-C\ + CHBCOCHB—p(Me381)ZCH co CH3 (61) 2; CH3 /OLi CH = c 2 \ CH3 The same procedure was followed with cyclohexanone. In this case, a product other than choc -bis-(trimethylsilyl).- l-HF NMR? of acetone was found in the reaction mixture. The this product was similar to the self condensation product of the cyclohexanone enolate (eq. 62). 47 0Li O O 0Li / _ n (MeBSi)ZC=C\CH +‘—-—p(MeBS1)ZCHCCH3 + ‘ (62) 22 3 condensation? D. With Algylhalides. An equivalent amount of methyl or ethyl iodide was added to the enolate 12 formed as before. GlC analysis of the solution after stirring at room temperature gave several products with higher boiling points than the starting materials, but all attempts to isolate and identify these products were unsuccessful (eq. 63). /OLi M638? 9 /OR (MeBSi)ZC=:C\ + R-X ——> R-.--/C--C--CH3 or (MeBSi)ZC=C\ .(63) 2g CH3 M8381 CH3 /OSiMe3 3. Reaction of (Me3Si)ZC=C\ , 39, with Ketones. CH3 An ethereal solution of 39 was prepared. One equiva- lent amount of cyclohexanone was added to the solution at -7800. Then 3 dr0ps of BF3 OEt2 was injected as a catalyst. The solution was stirred for 1.5 hows at -78°C and then 30 minutes at room temperature. Only trace amounts of an unidentified compound was formed. It definitely was not the expected product 23 (eq. 64). O SiMe3 33 + BF3 OEtz \Z, .=c-c:OCH3 (64) or 1 eq T1014A 31 48 The same reaction was done with an equivalent of TiClu with similar results. More than 80% of cyclohexanone was recovered from the solution after work up. 4. Reaction of the Bis (trimethylsilyl) Ketene wifih Silylenol ethers. An equivalent amount of bis-(trimethylsilyl)ketene containing 10% mole equivalent of BF3-0Et2 as a catalyst was injected drOpwise to a preformed solution of cyclo- hexanone silyl enol ether. None of the expected addition product was detected by GlC after 30 minutes stirring at 0°C. A trace amount of an unknown compound was detected on the G10 after solution was stirred overnight at room temperature (eq. 65). BFj'OEtZ l0% 0 o + (Me Si) c=c=o. , ms. (65) / . 3 2 . SiMe _ '“, 49 DISCUSSION Synthesis of u-Silylated Viny2 Ketones. ”2 and Boeckmanu'u' have shown that o<-si1y1 vinyl Stork ketones undergo Michael addition reactions with a number of electrOphiles. A valuable synthetic application of SiEt3 I 0 E R\/ SiEt (69) 5r 3 Several steps are required to prepare Grignard reagent 51 4g ‘ from commercially available vinyl silanes.43 It has been reported18 that lithio tert-butyl bis- (trimethylsilyl)acetate 25, obtained by treatment of the corresponding ester 23 with LDA (eq. 70), reacts with aldehydes to givecx-silyl vinyl esters (eq. 71). . -78°C . L1 (Me381)2CHCO§+- + LDA-—-—;E;—o-(Me381)ZCCOZ+- (70) £1 £5 0Li SiMe3 ‘ SiMe \ I / 25 + RCHO ——> R-CH-clz-coz-f— -———> RCH=C (71) SiMe 3 \C021L‘ Reagent 25 fails to react with ketones presumably due to steric effects.18 Bis silylation of ketones on carbon has not yet been reported because the silylation of ketone enolates with trimethylchlorosilane usually leads to silyl enol ethers (O-silylated product) rather than C-silylated ketonesn5-u7 (eq. 72). o 1) LDA, -78°C OSiMe R ”fl\\ i” , R et4§st 3 2) MeBSiCl (72) Mono C-silylated ketones can be synthesized by other 'r methods. Hauser53 prepared C-silylated ketones by the reaction of acid anhydrides with trialkyl- silylmethyl magnesium halides (eq. 73). l R SiCHZMgCl + (R1C0)20.___—_4pRBSiCH2COR (73) El 3 52 Brook54 reported that acylsilanes react with diazo- methane to form a mixture of 0- and C-silylated products (eq- 74). A 0 do I d9 R S. g R1 . I 1 . 1 1 CHZNq . CH2 (74) R SiO 69 ‘ o 3 \ ‘/l J . . . I 1 l/C=CH2 ‘_ R381 1:? - R J H ;- R381CHZC-R o-SllYlated / C-Silylated R1 = alkyl C-silylation of the49-keto silanes has not yet been reported. Although it has been reported that hindered ketenes such as di-t-butyl ketene13 and bis (trimethylsilyl) ketene are unreactive to nucleOphilic attack, the ketone analogues of 25 were prepared by the addition of alkyl lithium reagents to an ethereal solution of bis (trimethyl- silyl) ketene which is formed by decomposing 25 at room temperature (eq. 75). 0Li ( ) R'T' ( ) / ( ) Me Si C=C=0 + RLi -———-——> Me Si C=C . 75 3 2 ether 3 2 \‘R 32. Using methyl lithium, the enolate formed was reacted with aldehydes and ketones. Only benzaldehyde was - converted to the expectedmx-silyl vinyl ketone (eq. 76). 53' 0L1 SiMe -LiOSiMe SiMe 2 + flCHO——v MCH-o-COCH3 3' 20H=c\ o (76) c ‘4 SiMe3 \CHB All'other aldehydes and ketones having at least one (X-hydrogen gave the same product with enolate 32 as did H20 or dilute acetic acid (eq. 77-79). HéO or H363 8 2; __,. (MeBSi)ZCH-C—CH3. (77) 42 R 32 + 1:CH-CHO —__. 9.2 (78) R O 22 + R \\//fi-. R1 42 g; condensation (79) That means that the enolate 32 is protonated by aldehydes and ketones having d-hydrogen, again probably due to steric effects. The reaction of alkyllithhum».with bis-(trimethylsilyl)- ketene can be used as a general route in preparing ma -bis- (trimethylsilyl)ketones. Because the Grignard reagents could easily be prepared from a wide variety of alkyl halides, an attempt was made to use Grignard reagents instead of alkyl lithium. But bis-(trimethylsilyl)ketene was totally unreactive towards Grignard reagents. During the GlC analysis Of’obO& -bis-silylated ketones, it was observed that a,o< bis-silylated acetone 42 partially 54 decomposed on the GlC column (30.5%-SE on chromosorb W). The 1H NMR investigation of the decomposed mixture showed that the decomposition occurred by the migration of a silyl group (eq. 80). Me381\ 9 on Glc Me381\ /’081Me3 /CH-C-CH3 ’ C=C (80) Me Si 53'” H/ \ CH 3 3 42 c is and trans The migration was confirmed by the following observations: 1) Pure (x,o( bis-(trimethylsilyl)acetone, prepared by treating methyl lithium with bis-(trimethylsilyl)ketene and quenching with 3M HCl, was heated in a sealed tube in oil bath. No change was observed until 200°C. When the temperature rose above 200°C a yellow color formed. 1 2) The H NMR of his (trimethylsilyl)acetdnem42fi in 0811+ possesses .2 (CH3)3Si\ 2'9 2. .z’CH-C-CHB (CH3)381 42 a singlet at 8 0.27 ppm (18H) for the methyl groups on silicon (protons g). a singlet atS 2.01 ppm (3H) for the methyl ketone group (protons g), and a singlet at32.3l ppm 1 (1H) for thecx-proton (proton Q). But the H NMR of the heated yellow solution (decomposed solution) possesses five 55 singlets at 0.1-0.27 ppm for the trimethyl silyl groups; one for the remaining bis-silylated acetone (9) and four 1, 2, pl) for the cis and trans migrated products (a. 2 (eq. 81). (a) (2) (CH3)3si\(;) (2) over 20000 (CH3)BSi /OSi(CH3)3 /CHCO-CH 0.1 b t? H/C=C\CH . 1 a (CH3)381 3 (2) cis (81). (a?) (Q) (CH3)BSi\ /CH3 H ” \OSi(CH3) (5;) (pi) trans 3 Two singlets at 1.8-2.00 ppm, for the cis and trans methyl groups (d, d1 protons), one singiet at 2.01 ppm for the methyl of the remaining bis-silylated acetone (e proton), one singlet at 2.31 ppm for the o<-proton of the remaining silylated acetone (f) and one singlet at 4.4 ppm for cis (g) and trans (g1) vinyl protons were observed. 3) The IR spectrum of o(,o< -bis-silylated acetone 1 possesses a strong carbonyl absorption at 1680 cm' . This strong absorption decreased after heating to over 200°C 56 1 and instead, an absorption appeared at 1620 cm- for afl= carbon-carbon double bond. 4) The 1 H NMR and IR Spectra of O(,O( bis (trimethyl- silyl) acetone, collected from the 010 column outlet (30.5%?SE) were the same as that for the heated solution discussed previously. From these data it was concluded that thecx,«.bis (trimethylsilyl) acetone decomposes by migration of a trimethylsilyl group (eq. 81) if heated over 200°C. A similar type of decomposition has been reported by 25,26 Brook forA-keto silanes (cc -silylated ketones) on prolonged heating at temperatures from 80-175°C (eq. 82). o u o R3SiCHZC-Rl R3Sl°\ - (82) /C=CH2 1 R Lil . Related arrangements of trialkylsilyl acetones cata- lyzed by mercuric iodide or trialkyl silyl iodide have also been described.54 It has been suggested that this reaction proceeds via a four-centered activated complex (eq. 83). o .0. . K/\\ 1 ' l R381\\r_,>/C-R ——————-—»- RBSl“‘ u”'C-R -—4> 43 CH2 ‘CH2 (83) A crossed aldol reaction has been reported between silyl enol ethers and carbonyl compounds activated by titanium terachloride55 (eq. 84). 57 Rl\ /OSiMe3 RI: T1014 {R3\|_H/.o/SiMe3y\C2-T1C13 2/C=C\ + /C=O ————9 1 ’fi _3 R R3 R5 R2/\Rl /C-0 R“ R5 t 54 I TiCl ‘ Ox' ““0 '. -MeBSiCl RBJSXR R5 4—~ H20 R1 R2 R4 (84) R3 //0 inf / L Li: J \ C\c/ C\- R5 / \ R4 R1 R2 Therefore, the enolate formed by addition of methyl- lithium to bis-(trimethylsilyl)ketene was reacted with trimethylchlorosilane (eq. 85) to yield the correSponding silylenolether (3Q). Me331 0L1 Me Si 3 \ \ / C=C=0 + CH3L1-—-——D’ /C=C\ MeBSiCl MeBSi M83S1 CH3 Me381\C C/081Me3 8 M ./-\ (5) e351 CH3 32 ‘ Silylenolether 39 was treated with cyclohexanone in the presence of BF3-0Et2 with no detectable amounts of the correSpondingcx-si1ylvinyl ketone formed (eq. 86). 58 p BF3 OEt2 . 3g + 2_,, condensation ? (86) The same reaction was performed in the presence of TiClu; none of the eXpectedCX-silylvinyl ketone was observed by GlC. It appears that the reaction is prevented by the steric effect of bulky trimethylsilyl groups which disfavor the formation of transition states 44 or 45 (see eq. 84). 59 EXPERIMENTAL 1. Materialg £23222 Bis-(trimethylsilyl)ketene was prepared from t-butyl acetate by the procedure described in Chapter I and stored under argon at 0°C. Alkyllithium Reagents Methyllithium, n-butyllithium and tert-butyllithium were available from the Aldrich Chemical Company as standard 56 solutions which were titrated prior to use. Trimethylchlorosilane Trimethylchlorosilane was commercially available from the Aldrich Chemical Company and used after distillation (bp 57°C. 760 mm Hg). Aldehydes Acetaldehyde was available from Aldrich Chemical Com- pany. It was distilled (bp 21°C 760 mm) prior to use. Benzaldehyde was also available from Aldrich Chemical Company and distilled prior to use (bp 178°C). Ketones - Acetone was commercially available from Aldrich Chemical 60 Company and distilled prior to use (b.p~ 56°C). Cyclo- hexanone was also available from Aldrich Chemical Company distilled prior to use (b.p 155°C). 2. Preparation of(ng -Bi§:(tr2methylsily2) Ketones. The following procedure is typical for preparing bis-(trimethylsilyl) ketones. Argon was flushed through a 25 m side-armed round-bottomed flask, which was equipped with a magnetic stirring bar, septum inlet, gas inlet valve and mercury bubbler. The flask was charged with 5 ml anhydrous ether and 0.9 ml (5 mmoles) of methyl lithium in ether was added dr0pwise over 2 minutes. The solution was stirred for 1 hour at room temperature. Then 2.5 ml of 2 M solution of HCl was added to quench the solution. The aqueous layer was separated and organic solvent was removed under reduced pressure. The product was identified by its SINWRand IR spectrum. The yield was determined to be 75%. 1H NMR of o(,o< -bis-(trimethylsily1 acetone in CClu: 0.27 ppm (S, 18H), 2.01 (S, 3H, 2.31 (S, 1H), relative 1 to TMS. IR, strong absorption at 1680 cm' for the carbonyl group. 3. Reactions of (Xxx -Bis- trimethy2sily2acetone Enolate. A. High Trimethylch2orosi2ggg. Argon was flushed through a 25 ml side-armed round- bottomed flask, equipped with magnetic stirring bar, septum inlet, gas inlet valve and mercury bubbler. The flask was 61 charged with 5 ml anhydrous ether. Then 0.9 ml (5 m moles) of bis-(trimethylsilyl)ketene was injected and 3.4 ml of a 1.6 M (5.44 m moles) solution of methyllithium in ether was added dropwise over a 2 minute period. The solution was stirred for 1 hour at room temperature to convert the ketene to the corresponding enolate. Then 0.64 ml (5 m moles) of distilled trimethylchlorosilane was added dropwise. Stirring was continued for 1 hour. The product, 39, was identified by its 1H NMR Spectrum in 0C1“: Three singlets for three different kinds of tri- methylsilyl groups at 0.27-0.4 ppm and one singlet for the methyl group at 2.00 ppm, relative to the TMS. B. With Benzaldehyde. A 10 ml side-armed round-bottomed flask was equipped .with a magnetic stirring bar,,septum inlet, gas inlet valve and mercury bubbler. The flask was flushed with argon, 2 m1 anhydrous ether and 0.18 ml (1 m mole) of pure bis- (trimethylsilyl)ketene were injected into flask. Then 0.7 ml of a 1.6 M (1.1 m mole) solution of the methyl- lithium in ether was added drOpwise. The mixture was stirred for 1 hour, then 0.056 ml (1 m mole) of distilled benzalde— hyde was injected dropwise. The solution was stirred for 5 minutes. GLC analysis showed that all of the enolate was converted to the desired product in quantitive yield. The 1H NMR of product, 32, in 0014 was: 0.33 (S, 9H) for the trimethylsilyl group, 2.02 (S, 3H) for the methyl group, 62 8 6.76 (S, 1H) for the vinyl proton (due to the phenyl ring) andS 7.58 (S, 5H) for the phenyl. C. With other Aldehydes. The same procedure was followed with acetaladehyde and is a butyraldehyde. The results were the same as in quenching the enolate. NotX-silylated vinyl product was detectable; instead,o(,0( bis-(trimethylsilyl)acetone:was found in the solution. D.. flith Ketones. The same procedure as before was followed to make 1 mmole- of the ketone enolate. 0.104 ml (1 mmole)‘ dis- tilled cyclohexanone was added dropwise. The reaction was stirred for 0.5 to 1.5 hours. After quenching with suffi- cient water, none of the desireth-silyl vinyl ketone was found in the solution. The same result was observed for acetone. E. Reaction of the Enolate w2th A2kylha22des. l mmole. of ketene enolate was made by the above procedure. Then 62 ml (1 mmode) of distilled methyl iodide was injected dropwise at room temperature. New products were observed by Glc but all attempts to isolate and identify these products were unsuccessful. The same result was observed with ethyl iodide. 63 OSlMe3 / Si) C=C , 39, with Ketones. 2 \CH 3 A l mmole‘ ethereal solution of 39 was formed as 4. Attempt to React (Me3 described previously in this chapter. The flask was immersed in a dry ice-acetone bath and 0.104 ml (1 m mole) of cyclohexanone was added over 1 minute. 3 dr0ps of BF OEt2 was added as a catalyst to the flask. The solution 3 was stirred for 1.5 hours at -78°C. Then the solution was warmed to room temperature and stirred 30 minutes longer. None of the expected product was detected by Glc analysis. The same experiment was done using an equivalent of TiClu instead of BFjOEtZ. The solution was quenched with sufficient water at -78°C. No product was formed and more than 80% of the starting cyclohexanone was recovered, using Glc analysis. 64 BIBLIOGRAPHY l) 2) 3) 4) 5) 6) 7) 8) 9) 10) ll) 12) 13) BIBLIOGRAPHY D. F. Sullivan, R. P. Woodbury, M. W. Rathke, J. Org. Chem . 42 2039 (1977) a. M. W. Rathke, D. F. Sullivan, J. Amer. Chem. Soc. , 25. 3050 (1973L b. L. Lochmann and D. Lim, J. Organometal. Chem., 50_, (1973L Yu- -Negkuo, J. Yahner, C. Ainsworth, J. Amer. Chem. Soc., 23, 6321 (1971). -D. Sullivan, "The reactiOns of ester enolates," Ph.D. Thesis, Michigan State University (1974). W. R. Vaughan, S.C. Bernstein, and M.E. Lorber, J. Org. Chem., 2_. 1790 (1965). W. Remero, R. Roth, M. Weiss, J. Org. Chem., 3_, 2910 (1965) B. Holmquist, T. Bruice, J. Amer. Chem. Soc., 22, 2993 (1969L R. Prath, T. Bruice, Ibid, 22, 5956 (1970). J. Rebek, D. Brown, S. Zimmerman, Ibid, 21, 454 (1975). . Meyers, E. Smith, M. Ao, J. Org. Chem., 3_, 2129 (1973 H. R. Yaughan and H. P. Knoess, J. Org. Chem., 35, 2394 1970 R.P. Woodbury, "The chemistry of amid enolates and a study of ester enolate stability," Ph.D. Thesis, Michigan State University (1976). M. Newman, J. Amgg;g§oc., 32, 2498 (1960). 65 14) 15) 16) 17) 18) 19) 20) 21) 22) 23) 21.) 25) 26) ' 27) 28) 29) 30) BIBLIOGRAPHY, Continued G.D. CoPedege. "The synthesis of esters of Di tert butylacetic acid and their attempted reaction with stron base." Master Thesis, Michigan State University, 1979 - S.L. Hartzell, D.F. Sullivan and M.W. Rathke, Tetra- hedron Lett., 1403 (1974). K. Shimoji, H. Taguchi, K. Oshima, H. Yamamoto and H. Nozaki, J. Amer. Chem. Soc., 2_, 1620 (1974). M. W. Rathke and D. Sullivan, Synth. Commun., 2, 67 (1973)- S.L.éHartzell, M.W. Rathke. Tetrahedron Lett., 2737 197 - D.F. Sullivan, R.P. Woodbury, M.W. Rathke, J. Org. Chem., 32. 2039 (1977)- H. House, L. Czuba, M. Call, H. Olmstead,.L Org. Chem;, 2. 2324 (1969)- A. N. Nesmeyanov, I. F. Lutsenko, K. A. Bratzev, Dokl. Akad. Nouk. SSSR, 128, 551 (1959). Y. Boukov, G.W. Burleohenko, I.F. Lusenko, J. Organo- metal. Chem., 2, 478 (1965). g. C. Pommier, M. Pereyre, J. Valade, Compt. Rend., 260, 397- G. Stork, P. Hudrlik, J. Amer. Chem. Soc., 2_, 4464 (1968) A. Brook, D. Macrae, W. Limburg, J. Amer. Chem. Soc., .2. 5493 (1967) I. Lutsenko, Y. Baukov, O. Dudukina, E. Kramarova, J. Organomet. Chem., ii, 35 (1968). A. Gordon, R. Ford, "The chemist's companion: A Hand- book of Practical Data Techniques." E. Colvin, Chem. Review, 1, 14 (1979). . Kuo, F. Chen, C. Ainsworth, J. Amer. Chem. Soc., 21, 4604 (1971). G.E. Wilson, Jr., and J.G. Riley, Tetrahedron Lett., 379 (1972)- 66 31) 32) 33) 34) 35) 36) 37) 38) 39) 40) 41) 42) 43) 44). 45) L16) 47) #8) BIBLIOGRAPHY, Continued G. Natta, G. Mazzanti, G.F. Pregaglia, M. Bina i and M. Gransbini, Makrcmcleckular. Chem., 51, 1&8 %1961). G. F. Pregaglia, M. Binaghi and M. Cambini, Makromcle- ckular. Chem., 53, 10 (1963). S.C. Watson, J.F. Eastham, J. Organometal. Chem., 2, 165 (1967)- Fcr a good review see E.D. Bergmann, D. Ginsburg, and R. Pappc, Ogr. React., 19, 179 (1959). E. Stork, A. Brizzolara, H. Landesman, J. Szmuszkovicz, and R. Terrell, J. Amer. Chem. Soc., Q5, 207 (1963). J.A. Marshall and W.I. Fanta, J. Org. Chem., 22, 2501 (1969) and references cited therein. G. Stork, S. Danishfesky and M. Ohashi, J. Amer. Chem. Soc., g2, 5959 (1967)- G. Stork, S. Uyeo, T. Wakamatsu, P. Grieco and J. Labo- vitz, J. Amer. Chem. Soc., 21, 49#5 (1971). L.F. Cason and H.G. Brooks, J. Amer. Chem. Soc., 23, #582 (1952). L.F. Cason and H.G. Brooks, J. Org. Chem., 12, 1278 (1959). D.J. Peterson, J. Organometal. Chem., 9, 373 (1967). G. Stork, B. Ganem, J. Amer. Chem.Soc., 6, 6181 (1974). A. Ottolen hi, M. Fridin, A. Zilkha, gap. J. Chem., 3;, 2977 (1963 - R.K. Boeckman, Jr., J. Amer. Chem. Soc., 6, 6179 (1974). A.D. Petrcv, S.I. Sadykh-Zadeh, ch1, Akad, Nauk, SSSR, 121. 119 (1958). J.K. Rasmussen, Synthesis, 91 (1977). P.F. Hudrlik, "In new applications of organometallic reagents in organic synthesis." D. Seyferth, Ed. Elsevier, Amsterdam, 1976, pp. 127-159. G. Stork, B. Ganem, J. Amer. Chem. Soc., 25, 6152 (1973)- 67 BIBLIOGRAPHY, Continued 49) E. Brown, M. Ragault. ietgahgdron ngto. 393 (1978)- 50) L.N. Mashlyakovskii and L.F. Chelpanova, Zh. Obsch. Khim (Eng. transl.) 25, 2009 (1965). 51) N.V. Komarov, V.B. Puchnarevich, S.P. Suschinskaya, G.A. Kalabin, and V.G. Sakarovskii, Izvest. Akad. Nauk SSSR Soc. Khim. (Eng. transl) 803 (1968). 52) A.G. Brook and J.M. Duff, Can. J. Chem., 5;, 2924 (1973). 53) C.R. Hauser and C.R. Hance, J. Amer. Chem, Soc., 2;, 5091 (1952)- 54) O.V. Litvinova, Yu. I. Baukov, and I.F. Lutsenko, Dokl. Akad. Nauk. SSSR, 123, 578 (1967). 55) T. Mukaiyama, K. Banno and K. Narasaka, J. Amer. Chem. 800-. 2Q. 7503 (1974)- 56) M.E. Lipton, C.M. Sorensen, A.C. Sadler and R.H. Shapiro. J. Organometal. Chem., 186 (1980) 155. 68