Stereoselective Transformations Starting with Chiral (Aikoxy)Methyl-Substituted Organosilicon Compounds

AWARDS 1996 (NSCS) . PREISTRAGER 1996 (NSCG) 133

CHIMIA 5/ (1997) Nr. 4lApriIJ A a 1996 ( rager••

Chimia 51 (/997) 133-139 © Neue Schweizerische Chemische Gesellschaft ISSN 0009-4293

Stereoselective Transformations Starting with Chiral (AIkoxy)methyl-Substituted Organosilicon Compounds

Stefan Bienz* Stefan Bienz was born on Januar 17, 1958, in (Werner Prize 1996 of the NSCS) Lucerne (CH). He is citizen ofWolhusen (LU) and Lucerne, Switzerland. Academic Background: 1977-1983: Studies Abstract. The following short review summarizes the results we achieved with the in Chemistry at the Phi losophical Faculty II of investigation of chiral silicon groups as auxiliaries for the enantioselective synthesis. the University of Zurich, Switzerland. Gradu- (Alkoxy)methyl-substituted silicon compounds with 'Si-centered chirality', which ate research under the supervision of Prof. Dr. were prepared in optically active form by application of a bioreduction, have been M. Hesse at the Institute of Organic Chemistry towards the diploma (May 5, 1983; thesis: efficiently used as starting materials for a number of stereoselective reactions. Acylsi- 'Contribution to the Transformation of Car- lanes of this type upon treatment with organometallic reagents gave rise to 1,2-addition bocycIes to Lactams by Ring Enlargement products with high degrees of stereoselectivities. The respective a-hydroxysilanes Reaction') and 1983-1987 towards the Ph.D. could be stereospecifically desilylated to chiral secondary alcohols, or, depending on (December 18, 1987; thesis: 'Synthesis of the substitution pattern, further used as starting compounds for stereocontrolled Macrocyc\ic Natural Products by Ring En- oxidation, Cope- or Claisen-type rearrangement reactions. Chiral a-metallated vinyl- largement Reactions', distinguished for its silanes were converted to a-silyl-substituted allylic alcohols and to a-silyl-substituted excellence). 1988-1989: Postdoctoral studies at the Colo- a,j3-unsaturated ketones. The prior led to chiral allenes in a Peterson-type reaction - rado State University, Fort Collins, Colorado however, without stereoselectivity - the latter delivered stereoselectively j3-chiral (USA) with Prof. Dr. A. I. Meyers in the field silicon-free ketones upon stereocontrolled conjugate cuprate addition followed by of enantioselective synthesis. removal of the silicon auxiliary. Since 1990 at the Institute of Organic Chem- istry, University of ZUrich, Switzerland: Head assistant and lecturer; Summer 1995: Venia legendi (habilitation treatise: 'Enantioselec- 1. Introduction silanes 1that gave in highly enantioselective Synthesis Using Amino Acid or Silicon tive allylation reactions rise to a variety of Based Chiral Auxiliaries and Synthesis of Organosilicon chemistry has undergone a compounds [5][6]. The chiral allylsilanes Polyamine Toxins from Spiders and Wasps'). tremendous development over the last two 1,however, do not possess a 'chiral silicon Research interests: stereoselective synthesis or three decades. It is fair to state that moiety' which delivers stereochemical in- with organosilicon compounds, synthetic rel- nowadays, the majority of multistep or- formation onto a carbon framework; the evant mass spectrometric methods and frag- ganic syntheses makes use of organosili- stereochemical information is already mentations, unusual substrates for bio- transformations, physiologically interesting con compounds in one way or another: as stored on the reactive group, which itself polyamine conjugates. reagents for C,C-bond formations, forfunc- is attached to an 'achiral silicon group' tional-group transformations, or for func- (silanes of the type A with 'proximate C- tional-group protections. With the advanc- centered chirality' [4]). Transfer of chiral- attained in several transformations with ing interest in the preparation of enantio- ity from a chiral silicon auxiliary to a proline-derivedchiral silanes 2 (that are of merically pure compounds, it is not sur- neighboring prostereogenic group, how- the type B) [2]. The scope of reactions prising that also attention has been in- ever, has been realized with silanes ofthe with this type of substrates, however, is creasingly directed towards stereoselec- type Band C. These compounds have limited, since no common chiral silicon tiveprocesses involvingchiral silicon com- located the chirality either on an inert side precursor (like, e.g., a chlorosilane) is pounds [1-6]. Some rather impressive re- chain (compounds of the type B with 're- employed. So far, only little use of chiral sults were obtained, e.g., with chiral allyl- mote C-centered chirality') or- in form of silicon compounds of the type C has been a stereogenic center - on silicon itself made for the enantioselective synthesis (compounds of the type C with 'Si-cen- [1-5]. This is probably due to the low *Correspondence: PO Dr. S. Bienz tered chirality' [4]). It is just astonishing, stereoselectivities that were obtained in Organisch-chemisches Institut Universitiit ZUrich though, how little work has been per- several reactions where almost exclusi ve- Winterthurerstrasse J 90 formed with such substrates. Some re- Iy derivatives of 3 have been used as the CH-8057 ZUrich markable stereoselectivities were lately chiral starting compounds. AWARDS 1996 (NSCS) . PREISTRAGER 1996 (NSCG) 134

CHIMIA 5/ (1997) Nr. 4 (April)

Enhanced stereoselectivities were ex- external prostereogenic electrophiles, re- alents, followed by the respective func- pected to be attained, if (alkoxy)methyl- spectively, leading to novel chiral com- tional-group interconversion. They were substituted chiral silanes 4 were used in- pounds amenable to further transforma- alternatively transferred into (a-iodovi- stead of derivatives of3. These compounds tions. nyl)silanes 10 (the ultimate precursors of should be enabled to form cyclic chelate a-metallated vinylsilanes 6) by treatment intermediates or transition states with cat- with a suitable lithium acetylide, reduc- ionic additives (metals, M). As a conse- 2. Preparation of Chiral tion of the obtained alkynylsilanes 9 with quence of the rigid intermediary structure, (Alkoxy)methyl-Substituted Silanes diisobutylaluminum hydride (OIBAR), they should effect an improved discrimi- and quenching of the resultant vinylalane nation of the different spatial approaches Racemic chiral (alkoxy)methyl-substi- with iodine [8]. The stereochemistry of the reacting species. This is in fact tuted silanes were prepared straightfor- around the C=C bond of 10 can be control- observed, as will be shown in this short wardly from commercial (chloromethyl)- led by the proper choice of the reduction review: below, we describe the prepara- dichloro(methyl)silane (7) [7]: Stepwise conditions; for (E)-configurated silanes tion of chiral (benzyloxy)methyl-substi- substitution of the Cl-atoms (via bis(ben- 10, the C=C bond geometry can also be tuted silanes and some subsequent stereo- zyloxy)silane intermediates) - first at the inverted in the course of the subsequent selective transformations. Focus will be chloromethyl group with the benzyloxy metallation step. directed on the primary and subsequent residue, then at the silicon center with the Optically active acylsilanes 5 are ac- reactions of two types of chiral organosil- tert-butyl group - provided the racemic cessible by using a biotransformation step: icon compounds, namely acylsilanes 5 chlorosilane 8 (Scheme 1), which could be e.g., (±)-5a is recognized by several mi- and a-metallated vinylsilanes 6. These used as the common precursor for the croorganisms as a substrate and is reduced substrates should stereoselectively react synthesis of a variety of silanes. Com- stereospecifically to the corresponding a- either as electrophiles at the prostereogen- pound 8 was thus converted into acylsi- hydroxysilanes (+)-(SiR,R)-lla and (+)- ic carbonyl group or as nucleophiles with lanes 5 by reaction with acyl-anion equiv- (SiS,R)-lla' (Scheme 2) [9][10]. These diastereomeric alcohols were separated by chromatography and reoxidized to (-)- R (R)-5a and (+)-(S)-5a, respectively. Alter- I natively, the two alcohols can be convert- R-Si~R" MeI - Type A I e.g. r=<,R' 1 ed to the hydrosilanes (+)-(S)-12a and R Me-~i"" Rz Chiral Reactive Group Me R (-)-(R)-12a by treatment with KH (Brook rearrangement [11][12]) and reduction of ('Proximate C-Centered Chirality') the thus obtained silyl ethers with LiAlH4 [13]. Best results in the biotransformation R* r step were obtained with immobilized rest- R-Si-Reeae ing cells of Trigonopsis variabilis (DSM Type B I e.g. ("'={OMe 2 R 70714), which provided the desired prod- Me-SI-Reeae Chirallnert Substituent I ucts not only in high chemical yields and Me with enantiomeric excesses > 96%, but ('Remote C-Centered Chirality') also allowed to develop a handy workup procedure. R1 I . Ph 2 R -ShReeae 1.* Type C e.g. a.-Npht- SI-Rreae 3 R3 I 3. Stereoselective Reactions Starting Chiral Silicon Center Me with Chiral Acylsilanes

('Si-Centered Chirality') 3.1. 1,2-Addition to the Carbonyl Group The 1,2-addition of organometallic re- Bn I agents to the carbonyl group of acylsilanes 0, 5 afforded the respective addition prod- ( '¥ .. ucts 13 with varying degrees of diastereo- Me""Si 6 selectivity [7][14] (Scheme 3, shown ex- Mej( r: emplary for the conversion of (-)-(R)-5a R Me R''---. I Me \ to (+)-(SiR,R)-13a). Particularly high se- 0, lectivities were attained, when the reac- ( 'M 5 prostereogenic tions were carried out in nonbasic sol- Me....:./Me""Si "R " vents, like, e.g., hexane (up to 99% de). ./\ eeae For instance, the treatment of (-)-(R)-5a Me Me ( with PhLi in hexane provided predomi- 4 nantly (SiR,R)-13a ((SiR,R)-13a/(SiR,S)- 13a = 10:1).The selectivity dropped dras- tically with increasing basicity of the solvent, and it was even reversed when the transformations were performed in TRF (ratio 1:3) or in Et20fDMPU (= 3,4,5,6- AWARDS 1996 (NSCS) . PREISTRAGER 1996 (NSCG) 135

CHIMIA 5/ (1997) Nr. 4 (April) tetrahydro-l ,3-dimethylpyri midin-2(] H)- Scheme 1 one) mixtures (ratio 1:4). No solvent de- pendence of the stereochemistry of 1,2- (CI oan additions to chiral acylsilanes derived from Me ( ° 3 were observed. These results - together CI-5i-CI Me+~i-{ I - with others - indicate that the 1,2-addi- Me - Me Me R tions proceed in nonpolar solvents, as ex- 7 8 5 pected, via intermediary chelate structures. '- - These structures are, however, not chair- ! like six-membered chelates, as initially oan Me proposed [14]: the stereochemical out- ( come of the transformations is just re- Me+~i--===--R - versed to that anticipated on the basis of Me Me such intermediates. As mentioned above, 9 (+)-(SiR,R)-13a was primarily formed from (-)-(R)-5a (92% ee, for absolute configuration see [15]) by reaction with Scheme 2 PhLi under 'chelate-controlled' conditions. This was proven by the conversion oan Trigonopsis variabi/is Me ( H of (SiR,R)-13a to (+)-(R)-I-phenyletha- (D5M 70714) nol «+)-(R)-15, 88% ee), applying the Me+~i~OH restingfreecellsor stereochemically well-established Brook restingimmobilizedcells Me Me Me rearrangement [16] (formation of (SiR,R)- (±)-5a (66--99%) (+)-(SiR,R)-11a >96% ee 14 by treatment of (SiR,R)-13a with KH), (+)-(SiS,R)-11a' followed by removal of the silicon group. 1) KH (ca!.) (90-92%) 2) LiA1H4 (80-84%) 3.2. Cope and Ireland Ester-Enolate ! Rearrangements Oan oan It cannot be the ultimate goal of the Me ( ° Me ( research with chiral silicon compounds to Me+~i-{ Me+~i-H prepare enantiomerically enriched second- Me Me Me Me Me ary alcohols like, e.g., (+)-(R)-15. In fact, (-)-(R)-5a [aJo=-10.7 {+)-{51-12a [aJO=+1.2 we were rather interested to use the above- (+)-(51-5a [alo= +10.9 {-)-{R)-12a [alo=-1.2 discussed 1,2-addition of organometallic reagents to chiral acylsilanes for the stereoselecti ve preparation of substrates ame- Scheme 3 nable to subsequent stereoselective transformations, making use of the newly in- OBn R OBn R troduced chirality of the molecules or of Me ( 0 Me7( ,PH \ the chirality still residing on silicon. Thus, Me-l"'~i-{ Me-l"1i~!Y silicon-substituted compounds that are MeMe Me (86%) MeMe Me Cope Ireland suitable for and ester-eno- (-)-(R)-5a (+)-(SiR,R)-13a late rearrangements were synthesized with (92% ee) this method. Reaction of either a,~unsaturated acylsilane 5b with allyl Grig- KH (kat.) nard reagents or of {3,y.unsaturated carb- (98%) onyl compounds 5c, d with vinyl-lithium ! R species afforded 1,5-dienes 16a-drequest- Me (OBn HiJR ed for the Cope rearrangement (Scheme 4) TBAF [19]. The first pathway to access these HO~tPh Me-+"~i-o-fPh compounds is unsatisfactory, though, be- Me (70%) Me Me Me cause it provides the desired products with (+)-(R)-15 (+)-(SiR,R)-14 low stereoselectivities only. Excellent se- (88% ee) lectivities were exhibited by the second procedure, but that entry to compounds of the type 16 is still problematic due to the difficult preparation of the {3,y.unsaturat- provided the respective acylsilanes 17a-d though not proven yet, are thought to pro- ed acylsilane precursors 5c, d: these sub- in high chemical and stereochemical yields. ceed via six-membered chair-like transi- strates isomerize readily - under mild ba- The rearrangement conditions are rather tion states locating the silyl group in equa- sic or acidic conditions, e.g., upon chro- drastic, but milder versions of the oxy- torial position. These transition states were matography - to the thermodynamically Cope rearrangement, i.e., base- or acid- calculated to be lowest in energy. more stable a,~unsaturated isomers. Nev- catalyzed variations, were not applicable Ireland ester-enol ate rearrangement ertheless, the compounds 16a-d were heat- due to undesired Brook rearrangements or precursors were obtained from an acylsi- 0 ed for a prolonged time to 200 , and they H20 eliminations. The Cope reactions, lane, too (Scheme 5) [20]. The major prod- AWARDS 1996 (NSCS) , PREIS TRAGER 1996 (NSCG) 136

CHIMIA 5/ (1997) Nr. 4 (April) ucts arising from the highly stereoselec- ment of these esters with UCA and tri- tion state with the ketene acetal (Z)-con- tive addition of vinyl organometallic spe- methylchlorosilane (TMSCI), the Ireland figurated. The creation of a-silylated car- cies to (±)-5a, a-silylated allylic alcohols ester-enolate rearrangement conditions boxylic acids in the Ireland ester-enolate 18a and 18b, and compound 18c that was [21][22], afforded the a-TMS-substituted rearrangement was rather astonishing, but formed stereospecifically from 18a by the y,O-unsaturated carboxylic acids of the is not without precedence: these types of action of lithium cyclohexyl(isopropyl)- types 21 and 22 (Scheme 6). The reaction products were often obtained as minor amide (UCA) were acylated via their mag- is stereospecific, proceeding most proba- side products, Their formation was sup- nesium alkoxide to 19a-c and20a-c. Treat- bly via a chair-like six-membered transi- pressed by the use of (tert-butyl)chloro- dimethylsilane (TBDMSCI) instead of Scheme 4 TMSCI [23][24]. In our case, however, attempts to perform the Claisen-type rear- R3SI* ,.-----"-----,0 rangement in presence of TBDMSCI led Bno, II # solely to compounds 23a-c/24a-c, the Me Si~Me starting esters silylated in a-position to the Me>( 'Me BrMg~R' Me 5b ~ - _ ester group. o Me 200· The removal of the TMS group of21a- c and 22a-c can be performed chemose- 4-7d R3Si*~ ca. Quant. R' lective1y by treatment of the compounds R' 17a-d with Et3N· 3 HF. Under these conditions, the chiral silicon moiety is not affected. R' = H (E) and (Z) > 95% de The reaction, however, is not stereoselec- oC R' = Me (E) and (Z) (FMe tive, giving rise to approximately equimo- BnO Me Me lar amounts of the epimeric products 26a, 5c,d band 26c, d starting from 22a or 22b, c, respectively. The configuration of the chiral center at e(2), however, could possibly Scheme 5 be adjusted as required in a later stage of a synthesis. At the moment, the a-desilylat- BnO\ 0 M"""Me BnO\ OH BnO\ OH ,Si-" • ,Si-f..Me ed vinylsilanes of the type 25 and 26 are Me...... ,.",·~\ Solvent Me...... ,.".' ,Si:{;~ ....Me + M M'''\ Me Me under investigation as starting materials e Me Me" \ Me :?'(~e\\ Me Me Me Me for substitution reactions with electro- (±)-5a 18a (SiR*,R*, Z) LICA 18d (SiR*,S', Z) philes. 18b (SiR',R', E) -.:::..:.... 18c (SiR',S', E)

1) MeMgBr 3.3, Reaction Cascade Leading to 2) RCH2COCI or A ldo 1-Type Products (RCH2C0l20 Allylic alcohols are prime candidates o for stereoselective Lewis-acid-catalyzed BnO\ 0-1° BnO\ 0-1- epoxidations with peroxides, and we an- Me...... ,. ••,·.si~''''MeLR~ Me...... ,.,••"~ (-Me LR Me" \ Me Me" \ Me \ ticipated a-silylated allylic alcohols of the Me Me Me Me type 18 as particularly attractive substrates 19a (SiR',R', Z), R = H (Scheme 7): they do not only possess the 19b (SiR',R', E), R = H 19c (SiR:S: E), R = H 20a (SiR*,R', Z), R = Me potential for an intramolecular chirality 20b (SiR',R*, E). R = Me 20c (SiR:S: E), R = Me transfer during the oxidation process, but also embody the possibility of a subsequent transformation leading to aldol-type Scheme 6 products. a-Hydroxy-j3,~epoxysilanes27, deriving from silanes 18, should be ame-

1) 4 eq. L1CA Me Me H MeHMe nable to Brook rearrangement, which 2) 3.5 eq. TMSCI ~C02H or ~C02H would lead, after eliminative opening of R3Si •••• R3Si " the oxirane, to silyl enol ethers 28 of 3) -aO· to -10· R'TMS R 'TMS aldols. Preparation of aldols on this path- I-Products u-Products (from (Z)) (from (E)) way would allow a stereoselective entry also to compounds that are not readily R=H 21b,c R = H 19a-c R=H 21a accessed by the aldol reaction itself; the 20a-c R=H 22a R=Me 22b,c R=Me acylsilane would have acted as a synthetic 11) 4eq.L1CA Et3N'3HF Et3N'3HF equivalent of an acyl cation, and a vinyl + 2) 3.5 eq. TBDMSCI ! ! o bromide as a P-hydroxy anion equivalent. The reaction of allylic alcohols of the O~TBDMS MeMeH MeHMe type 18 with tert-butyl hydroperoxide or ~C02H ~C02H (TBHP) in presence ofV(O)acac2 (acac R3Si~Me R3Si R3Si = Me R R acetylacetonate) gave in fact rise to aldol- type products (Scheme 8) [25]. However, 23a-c R= H 25a R=H 25b R=H not silyl enol ethers of the type 28 but a- 24a-c R= Me 26a,b R=Me 26c,d R=Me silylated p-hydroxyketones of the type 29 AWARDS 1996 (NSCS) . PREISTRAGER 1996 (NSCG) 137

CHIMIA 5/ (1997) Nr. 4 (April) were formed. Evidently, the intermediary Scheme 7 epoxides did not rearrange in a Brook- but rather in apinacol-type fashion (see Scheme H 9). The stereoselectivities ofthe reactions .....0.... _ involved in the cascade were virtually 1) Brook Rearr. R~~lJR'2) Elimination R~ quantitative. The ll"-faceselectivity of the 3) Prolonation epoxidation step, however, was not con- 18 27 sistent throughout the investigation of all compounds. It was strongly dependent on the substitution pattern of the starting materials. Whereas the attack of oxygen to 18a, c was stereochemically controlled by the chiral center at silicon, the carbinol chiral center was substantial for the ll"-face differentiation in compounds with spatial- ly increased groups atC(l) (e.g., 18e) orat Scheme 8 C(2) (e.g., 18t') of the allyl system. For HO V(O)aca~ 0 (I-PrO)4n o 9SI*R3 18a, c, the chelate transition structures TBHP cal. shown in Scheme 9 were proposed to ex- Me Me Me~Me plain the stereochemical outcome. It is H SSIR3 assumed that such chelate structures are 18a,c 29a 30a no more accessible for more encumbered starting materials, and consequently, a dif- V(Olaca~ 0 NaOEI ferent reaction course must be followed. TBHP cal. p~:3 The removal of the silicon auxiliary in compounds 29 presents at the moment a Me Ph Ph forward problem. Silicon shift from car- 18e 30e bon to the hydroxy O-atom by a 1,3-Brook rearrangement can be brought about by HO V(O)aca~ 0 (l-prOl.n o OSI*R3 TBHP cat. Lewis acid as well as by base catalysis, but Me MeYMe the yields of the respective products 30 are Me rather low. Elimination to a,,B-unsaturat- 18t 29t 30' ed ketones isthe major side reaction, which one Isomer only we have not been able to suppress so far.

Scheme 9 4. Stereoselective Reactions Starting with a-Metallated Vinylsilanes Me(5) Me \ ~L~~/(v) ." A,cld M si~o--V ('".OH I Me 4.1. Addition to Aldehydes and Forma- eJ:H" / ..•..•.~O Me / ~~----o --- R S·S ••.• tion of Chiral Allenes Me "" -OR 3 I Me (5) ··....Me ~ The addition of the chiral lithiated vi- H --.::.:::-. Re nylsilanes 31, prepared from the respective vinyl iodides 10 (preparation see H .---Re Scheme 1) either under preservation or I-! - -_ , _0, inversion of the C=C bond geometry, re- Me • -'Mel oc Bn ,'" acted with aldehydes to afford the corre- -- 0 ...V, ----+-- sponding ,B-silylated allylic alcohols 321 Me)-s(~ -\d~~v) 32' in good chemical yields (Scheme 10) Me /IMe ../ (5) Me (R) [8]. The stereoselectivities of the reactions, however, were fairly poor (67% diastereoselectivity at its best) and could not be improved with reasonable effort. reduction with LiAlH4 afforded the re- were gained. This is a remarkable im- This result was somewhat disappointing spective alcohols in ratios of ca. 10:1[26]. provement as compared to earlier attempts but was in fact not too surprising, since by With the allylic alcohols 32 at hand, of other groups that delivered allenes with considerations of structure models, no we tested their potential as starting mate- 60% yield at their best [27][28]. Though evidently favored approach of the chiral rials for the preparation of chiral allenes the reaction rates were higher with starti ng nucleophile to one of the ll"-facesof the by means of a Peterson-type olefination. compounds possessing better leaving prochiral electrophile could be recognized. The allene formation was optimized with groups (e.g., chioro- or dichloroacetates), Highly diastereomerically enriched sam- several O-acyl-substituted compounds of such substrates gave still lower yields of ples of alcohols 32 were finally still ob- the type 34. It was found that the treatment the desired allenes due to more side reac- tained in a roundabout way: oxidation of of the O-acetyl derivatives with CsF in tions. epimeric mixtures of 32/32' to their com- DMSO at 100-120° gives the best results: Much to our disappointment, thefluori- mon a,,B-unsaturated ketones 33 and re- yields of up to 95% of the parent allenes 3S de treatment of optically active silane (-)- AWARDS 1996 (NSCS) . PREISTRAGER 1996 (NSCG) 138

CHIMIA 5/ (1997) Nr. 4 (April)

(SiR,E)-34a - obtained from (-)-(R)-5a induction, was anticipated to be highly 5. The Fate ofthe Chiral Silicon Group by a-methylation, Shapiro reaction, reac- stereoselective, too. In fact, the reaction of tion with benzaldehyde, and acetylation several organocuprates with compounds Using a chiral auxiliary in stoichio- (Scheme 11) - gave the corresponding 33a-c delivered the corresponding addi- metric amounts, as we do it with our trans- allene 35a in low yields and as a racemate tion products 37, trapped as sily] enol formation, one must argue that it is neces- only. The earlier results from Torres et al. ethers 36, with high chemical and stereo- sary to recover this group in optically [29], who obtained the same product chemical yields (Scheme 12) [30]. Partic- active form for recycling. We were in fact through a similar pathway in 18% ee (as- ularly good results were obtained with the confident that this might be possible, since sumingly by an anti-elimination), could (Z)-configurated starting material 33a; but most transformations involving substitu- not be verified with our compound. Sup- the stereoselectivities were also satisfac- tion at silicon proceed with high degrees posedly, the formation of the allenes pro- tory with the (E)-configurated ketone 33b. of stereoselectivity [31]. Since we per- ceeds mainly by an EI rather than an E2 Only with 33c, where the group R 1 is formed our reactions only in a few cases mechanism. The EI mechanism is proba- rather bulky, the stereoselecti vities with optically active material, we simply bly particularly favored for the (E)-con- dropped to almost zero. This is probably did not have had the opportunity to test figurated starting materials, where the tran- due to unfavorable steric interactions be- this matter too closely yet. In the prepara- sition states for concerted E2 elimination tween the groups R I and R2, which prevent tion of (+)-(R)-l-phenylethanol (Scheme processes are strongly disfavored due to the formation ofchelate intermediates such 3), we were able to isolate the respective large A I.rstrain. The E2 mechanism could, as that shown in Scheme 13 with (+)-(R)- silicon component as the respective fluoro- however, still be preferred for the (Z)- 33a. Chelates of that type are thought to be and hydroxysilanes or, when the silicon configurated silylated elimination precur- responsible for the high n-face differenti- group was removed reductive]y by treat- sors. Since optically active (Z)-configu- ation in the addition reactions to the com- ment of (+)-(SiR,R)-14 with LiAIH4, as rated acetates of the type 34 are not acces- pounds 33. The stereochemical result of the hydrosilane. All three compounds sible to date, this cannot be tested momen- the reaction of optically active (+)-(R)- showed low optical rotations on]y. Similar tarily. 33a with Ph2CuLi is consistent with an results were obtained with the recovered attack of the cuprate at the least hindered silane components from the preparation 4.2. Conjugate Cup rate Additions face - opposite to the tert-butyl group - of of allene (±)-35a (Scheme 11) or ,B-chiral The a,,B-unsaturated a-silylated ke- the C=C bond of such a lithium chelate. ketone (-)-(R)-38 (Scheme 13). We have tones 33a-c, obtained by oxidation of the This attack of the reagent leads to (-)- not been able to detennine the enantio- alcohols 32a-c/32a-c', were assumed to (SiR,3S)-37a/37a' and finally to (-)-(R)- meric excesses of the respective silane be suitable substrates for the stereoselec- 38, which was compared with an authentic substrates yet. We expect, however, con- tive 1,4-addition of organocuprate rea- sample. The formation of an acyclic ,B- siderable loss of chiral information in the gents. Since LiAIH4 reductions already chiral ketone by means of an auxiliary- formation of the fluoro- and hydroxysi- led with high stereoselectivity to the re- assisted stereoselective cuprate addition lanes, since fluorosilanes racemize readi- spective allylic alcohols 32, the conjugate in an enantiomeric excess of 93% repre- ly under the influence of fluoride ions, and cuprate addition, which would also be sents to our knowledge the best example because the hydroxysilane was possibly stereochemically controlled by 1,3-chiral of such a reaction. formed by hydrolysis via the respective fluorosilane. To establish the chiral sili- Scheme 10 con auxiliary as a synthetically useful group, it remains still ample research to be done in respect to enable the recovery of the auxiliary in a chiral1y integer form. A~O.Py

(7D-80%) (86-95%) 6. Conclusion 31 32/32' (dr~ 3:1) 34 We have shown with a number of cr03!t LiAIH4 (dr~10:1) (50-95%) I g~so rather simple reactions that 'Si-centered' * 120° Bo chiral silicon auxiliaries can efficiently be o 0 Me ( )LR2 used as stereochemical directors in enantioselective syntheses of silicon-free or- Me+~i~l ganic molecules. The scope of transfor- Me Me R 35 mations that might be aided by chiral 2 33 R'=Me R =Ph RI •• Ph R2=Me, Ph, silicon is far from exhaustively explored Et,I·Pr yet, and we will further invest our full enthusiasm in the exploration of more applications of 'our' auxiliary. A draw- Scheme II back in the field of the chemistry with Bn Bn chiral (a]koxy)methyl-substituted silicon 1) LDA/Mel 0 (0 OAc 2) ArS02N2H3 CsFIDMSO H compounds, however, is still the rather Me ( 0 1200/48h 3) BuLi Me ~Ph H-c=c=d troublesome and laborious access to opti- Me-+"'1i~ Me-+"'1i ~ Ph"" \ 4) PhCHO (20%) Me cally active chiral silanes and, as dis- Me Me Me Me Me Me 5) MeMgBr / AC20 cussed above, the unsolved problem of (-)-(R)·5a (-)-(SiR,E)-34a (±)-35a recovery of the chiral auxiliary. Even though the biotransformation leading to AWARDS 1996 (NSCS) . PREISTRAGER 1996 (NSCG) 139

CHIMIA 5/ (1997) Nr. 4 (April) the 'enantiomerically pure' acetylsilanes Scheme 12 Sa can probably be scaled-up to our liking, it still requires the respective large-scale 3 instrumentation. A possible alternative 1) R 2CuLi I TMSCI might be to shift our attention more to Et20 chiral compounds of the type B that pos- 2) TMEDA sess chiral bidentate substituents. The ac- (60-90%) cess to a silicon compound that could be used as a common substrate for the prep- 33 R3:;; Me, Et, aration of a variety of silane derivatives BU,Ph would then be required. Selectivities ] would like to express my undivided grati- Bn Starting Material 33 Product 36 tude to my coworkers who all have contributed and some of them still contribute with their inde- R1 R2 Conf. Conf. dr M~l{O R2 fatigable and enthusiastic work to the results i a Me Ph (Z) (Z) 35:1 to >40:1 Me+S presented above (and to other results not shown I 3 b Me Ph (E) (Z) 9:1 to 40:1 Me Me R here). These are: S. Bratovanov, Dr. A. Chapeau- Rt H rouge, Dr. V. Enev, J. Fiissler, S. Gassmann, P. c Ph Ph (E) Mixtures 1:1 to 4:1 Koch-Huber, Z. Molnar, J. Schroder, R. Smith, 37 and Dr. D. Stojanova. I would like to thank also TMEDA = N,N,N',N'-tetramethylethylenediamine to the people in Stuttgart and Braunschweig, to Prof. C. Syldatk, Dr. M. Pietzsch, and Dr. L. Scheme 13 Fischer, for the fruitful collaboration in the bio- transformations, to Prof. M. Hesse and his group Bn for providing us with laboratory space and regu- O--Li+ laroccasions for discussions, and to the members ( 0 of the Institute of Organic Chemistry at the Uni- .. versity of ZUrich, particularly L. Bigler, N. Bild, Si Me- A. Guggisberg, Dr. R. Kunz, Dr. A. Linden, and Me 71"" '?I Ph Me ~ Dr. D. Nanz for supporting our activities with Me Me [CuPh2r their expertise. The Swiss National Science Foun- dation is gratefully acknowledged for the finan- (-l-(R)-5a (+)-(R)-33a cial support. 96% ee ~!

Recei ved: January 21, 1997 o Bno 0 TBAF/THF l [I] C.A. Maryanoff, B.E. Maryanoff, in 'Asym- ph .. Si metric Synthesis', Eds. J.D. Morrison and (87%) Me;{, 0Ph J.W. Scott, Acad. Press, New York, ]984, H Ph Me Me ~"Me Vol. 4. t H Ph [2] T.H. Chan, D. Wang, Chem. Rev. 1992,92, (-)-(R)-38 (-)-(SiR,3S)-37a/37a' 995. 93% ee [3] R.I.P. Corriu, C. Guerin, J.J.E. Moreau, in 'The Chemistry of Organic Silicon Com- TBAF = BU4NF pounds', Eds. S. Patai and Z. Rappoport, Wiley, Chichester, 1989, p. 305. [4] S.I. Hathaway, L.A. Paquette, 1 Org. Chem. [14] A. Chapeaurouge, S. Bienz, Helv. Chim. [22] P. Wipf, in 'Comprehensive Organic Syn- 1983, 48, 3351. Acta 1993, 76, 1876. thesis', Ed. B.M. Trost, Pergamon Press, [5] I. Fleming, J. Dunogues, R. Smithers, Org. [15] S. Bratovanov, A. Linden, S. Bienz, Acta Oxford, 1991, Vol. 5, p. 827. Reac.1989,37,57. Crystallogr .. Sect. C: Cryst. Struct. Com- [23] R.E. ]reland, R.H. Mueller, J. Am. Chem. [6] C.E. Masse, J.S. Panek, Chem. Rev. 1995, mun. 1996,52, 1045. Soc. 1972,94,5897. 95,1293. [16] The Brookrearrangement of a-aryl-substi- [24] R.E. Ireland, R.H. Mueller, J. Am. Chem. [7] S. Bienz, A. Chapeaurouge, He/v. Chim. tuted a-hydroxysilanes, i.e., the base-in- Soc. 1976,98,2868. Acta 1991, 74, 1477. duced simultaneous ] ,2-shifts of the sili- [25] J. Fassler, V. Enev, S. Bienz, in prepara- [8] S. Bratovanov, S. Bienz, Main Group Met. con group from carbon to oxygen and of the tion. Chem. 1996, 19,769. hydrogen from oxygen to carbon, was [26] V. Enev, unpublished results, University of [9] P. Huber, S. Bratovanov, S. Bienz, C. Syl- shown to proceed strictly under retention of Zurich, ]994. datk, M. Pietzsch, Tetrahedron: Asymm. configuration at silicon and inversion of [27] T.H. Chan, W. Mychajlowskij, Tetrahe- 1996,7,69. configuration at carbon (17]( 18]. dron Lett. 1974, 15, 171. [10] S. Bienz, S. Bratovanov, A. Chapeaurouge, [17] M.S. Birnbaum, H.S. Mosher, J. Am. Chem. [28] T.H. Chan, W. Mychajlowskij, B.S. Ong, P. Huber, L. Fischer, M. Pietzsch, C. Syl- Soc. 1971,93,6221. D.N. Harpp, J. Org. Chem.1978,43, 1526. datk, in 'Biochemical Engineering', Ed. [18] A.G. Brook, J.D. Pascoe, J. Am. Chem. [29] E. Torres, G.L. Larson, G.I. McGarvey, R.D. Schmid, Stuttgart, 1995, Vol. 3, p. 70. Soc. 1971,93,6224. Tetrahedron Lett. 1988,29, 1355. [II] A.G. Brook, C.M. Warner, Tetrahedron [19] P. Koch-Huber, Ph.D. thesis, University of [30] S. Bratovanov, S. Bienz, Tetrahedron: Lett. 1962, 18,815. Zurich, in preparation. Asymm. 1997,8, in press. [12] A.G. Brook, W.W. Limburg, 1 Am. Chem. [20] V. Enev, D. Stojanova, S. Bienz, Helv. [3]] L.H. Sommer, 'Stereochemistry, Mecha- Soc. 1963, 85. Chim. Acta 1996, 79, 391. nism, and Silicon', McGraw Hill, New [13] A. Chapeaurouge, Ph.D. thesis, University [21] R.E. Ireland, R.H. Mueller, 1. Am. Chem. York, 1965. of Zurich, 1996. Soc. 1972,94,5897.