Classics in Asymmetric Synthesis

1

Chapter 13: Olefinations 13.1. Introduction

Ever since Georg Wittig back in 1953 pioneered the first olefination of a carbonyl compound,1 the synthesis of alkenes has become an indispensable strategic tool for the construction of complex molecules. While alkenes can be obtained by functional group transformations, most notably by selective hydrogen transfer to alkynes, the most powerful olefination reactions proceed by concurrent carbon-carbon formation.

There are three general strategies that allow the synthesis of alkenes with a broad structural variety and functional group tolerance, which each emerged approximately two decades apart from each other. The reaction of carbanions 3, being stabilized by an electron withdrawing group LG that acts at the same time as a good leaving group, with aldehydes or ketones 2 has been for more than 50 years a most reliable tool for alkene synthesis (Scheme 1, strategy A).2 Depending on the group LG, these transformations are known as Wittig reaction (LG = PR3),3 for which the Nobel Prize was awarded in 1960, Horner-Wittig reaction (LG = P(O)Ph2), Horner-Wadsworth-Emmons (HWE) reaction (LG = P(O)(OR)2),4 Julia Olefination (LG = SO2R),5 or Peterson Olefination (LG = SiR3).6 Related to this strategy are carbonyl olefination reactions with metal carbene, carbenoid or gem-dimetal complexes, making use e.g. of titanium (Tebbe reagent), zinc, chromium, or zirconium. Reductive coupling reactions of carbonyl compounds using low valent titanium, most notably the McMurry reaction, have also seen a spectacular development over the years, beginning with the synthesis of symmetrical, unfunctionalized alkenes up to complex applications in natural product synthesis with high functional group tolerance.7

R3

R2

R1

R4

O

R2

R1

+ LG

R4

R3

R3

R2

Y

R4

R1 X +

R2

R1

+R3

R4

1

2 3

4 5

6 7

LG = PR3, P(O)R2, SO2R, SiR3

X = Hal, OTf, N2 , OP(OR)3Y = H, BR2, SnR3, Hal

A

B

C

Scheme 1. Widely used strategies for olefinations

In the seventies, transition metal catalyzed cross coupling reactions with alkenes 5 entered the synthetic arena (Scheme 1, strategy B).8 Direct coupling of alkenes (Heck reaction: Y = H) or alkenyl metals (Stille reaction: Y = SnR3; Suzuki-Miyaura reaction: Y = BR2; Negishi reaction Y = ZnR) with appro-priately activated aryl, alkenyl or even alkyl substrates 4 has been developed into most economic processes, which have found not only manifold examples as

R2R1

R2R1

R1 R2

Li, NH3

Pd/C, BaSO4, H2 (Lindlar Catalyst)

Oliver Reiser

2

key steps in natural product synthesis but are used in industrial large scale applications for common intermediates such as styrenes as well.

With the discovery of robust and readily available catalysts, about a decade ago molybdenum (Schrock) and especially ruthenium (Grubbs), catalyzed metathesis reactions (Scheme 1, strategy C) initiated a change in paradigm not only for the synthesis of alkenes, but also for the assembly of complex structures such as carbo- and heterocycles. 13.2. Olefination of carbonyl compounds with phosphorus reagents 13.2.1. The Wittig reaction

The reaction of phosphonium ylides with carbonyl compounds – named after its inventor the Wittig reaction – is one of the most reliable methods for the synthesis of alkenes. It occurs with complete regioselectivity, i.e. the carbonlyl group of the substrate is replaced by the alkene group without that double bond isomerizations occur, and also the double bond geometry can be controlled to a large extend.

There has been a heated debate on the mechanism of the Wittig reaction, which is still ongoing even today. The controversy is centered around the intermediate that is formed after addition of the ylide to the carbonyl compound, i.e. if the betaine plays a role along the reaction pathway, or if the phospha-oxetan is directly formed in a formal [2+2]cycloaddition. Based on low temperature 31P NMR studies and calculations, and also supported through X-ray structures of phosphaoxetanes the scales have been tipped over in favor of the latter not only as being the decisive intermediate prior to alkene formation through elimination of phosphinoxide. Thus, the currently accepted mechanism for the Wittig reaction is the formation of the phosphaoxetanes 10 and 12, which stereospecifically collapse to the Z- or E-alkenes 11 and 13 (Scheme 2). The sterically more crowded cis-phosphaoxetane 10 is apparently kinetically favored over the trans-phopha-oxetane 12 for reasons not yet fully understood. Consequently, non-stabilized ylides 9 will follow a kinetically controlled pathway via 10 to give preferentially rise to (Z)-alkenes 11. With stabilzed ylides an equilibration of the diastereo-meric phosphaoxetanes 10 and 12 is reached via back reaction to 8 and 9, thus ultimately favoring the formation of (E)-alkenes 13 via the sterically less crowded 12 under thermodynamic control (Table 1). As a notable exception, α-alkoxyaldehydes give rise to Z-alkenes, especially when the reaction is carried out in protic solvents.

O

Rs

RL

+ PPh3

R

8 9

PPh3

O

RL

R

PPh3

O

RL

R

–

–

R RL

R

RL

10

12

11

13

PPh3

O

PPh3

O

Rs

RsRs

Rs

Scheme 2. The mechanism of the Wittig reaction

PPh3

RCH2 X

RCH2 PPh3

X

Base

RCH PPh3

R1CHO

PPh3

OR1

RPPh3

OR1

R

betaine phosphaoxetane

Chart 1. Definition of stabilized and non-stabilized ylides

RCH PPh3

non-stabilized ylides:R = H, Alkyl, Arylsalt free (M!Li)generated in situreact with aldehydes and ketones

stabilized ylides: R = electronwithdrawing group e.g. CO2R, COR, CNcan be isolated and subsequentlyreacted with aldehydes

M X

RCH2 PPh3

X

M Base

Classics in Asymmetric Synthesis

3

Table 1. Wittig reaction of aldehydes 8 with phosphonium ylides 9

8: RL/RS 9: R MX/solvent 11:13 (E/Z) Yield (%) Ref. –(CH2)8–OAc/H n-Pent NaHMDS/THF 2:98 79 9a

O

OHC

MeO O

O

CO2Et

DMF CHCl3 MeOH

86:14 60:40 8:92

9b

Nevertheless, the synthesis of (E)-alkenes from non-stabilized ylides can be achieved using the Schlosser modification,10 which calls for the presence of lithium salts to form the betaine, followed by a deprotonation/reprotonation sequence under kinetically controlled conditions in order to drive the reaction the trans-phosphaoxetane trans-15.

PPh3

R

14

PPh3

OR1

R

(cis)-15

Br

1) PhLi

2) R1CHO

LiBr H R1

OLi

H R

Ph3P

PhLi H R1

O

Ph3P R

Li

Br

LiBr

(syn)-16 17

H+

H R1

OLi

Ph3P R

H

(anti)-16

PPh3

OR1

R

(trans)-15

R1

R

18 Scheme 3. The Schlosser modification of the Wittig reaction

A very useful extension of this variant is to trap 17 with formaldehyde, resulting in the formation of Z-configurated allylic alcohols (Scheme 4).11 It is important to note that only the secondary alkoxide in 19 forms an oxaphos-phetane that subsequently eliminates triphenylphosphine oxide to yield the alkene 20.

H R1

O

Ph3P R

Li

Br

Li

17

O

HH H R1

O

Ph3P R

Li

19

LiO

– Ph3POR1

R

HO

20 Scheme 4. Synthesis of Z-allylic alkenes from aldehydes

The synthesis of alkenes with perfluoroalkyl groups is usually not possible by direct Wittig reaction, since such substituents in α-position are

Chart 2. Factors that control the alkene geometry in the Wittig reaction E-alkenes stabilized Ylides aprotic solvents catalytic amounts PhCO2H Schlosser modification Z-alkenes non-stabilized ylides salt free conditions (M≠Li) sterically bulky aldehydes sterically bulky ylides α-alkoxy aldehydes, methanol

Chart 3. Examples for the Schlosser modification of the Wittig reaction (cf. Scheme 3) R R1 Y

(%) E/Z

Me n-Pent 70 99:1 n-Pent Me 60 96:4 n-Pr n-Pr 72 98:2 Me Ph 69 99:1 Et Ph 72 97:3

Oliver Reiser ! 11/7/04 2:56 PMComment: Check authors of reference 10

Oliver Reiser

4

stabilizing the ylide too strongly to render it nucleophilic enough to react with carbonyl compounds. Perfluoralkylated β-ketophosphonium ylides, however, provide a solution for the synthesis of such alkenes, demonstrating in an interesting way alternative applications of commonly employed phosphor ylides (Scheme 4).12 Thus, ylides 21 can be acylated with perfluoroalkyl substituted acid anhydrides to yield β-ketophosphonium salts 22, which have a sufficiently high carbonyl reactivity to be attacked by nucleophiles to form alkenes 23 with generally high (E)-selectivity.13

PPh3

R

21 22 23

(RfCO)2O

RPh3P

Rf

O R1Rf

RR1

R1

Rf = CF3, C2F5, n-C3F7

R, R1 = alkyl, aryl, allyl, benzyl

R2 = Ph, n-Bu, alkynyl

R2

R2Li

Scheme 5. Synthesis of perfluoralkyl substituted alkenes

Despite the great synthetic versatility there is one significant disadvantage that has hampered the application of the Wittig reaction on industrial scale: As a byproduct triphenylphosphine oxide is formed, disfavoring this transformation under the aspect of atom economy. A convincing solution for large-scale applications of the Wittig reaction was developed by BASF AG in the synthesis of 26 as a precursor towards Vitamin A. Triphenylphosine oxide (27) is recycled by chlorination with phosgene, followed by reduction of the resulting dichloride (29) with aluminum metal to give back triphenylphospine (28) along with aluminumtrichloride that is further used as a catalyst for Friedel-Crafts acylations of aromatic compounds.

O PPh3

PPh3

COCl2

PPh3

Cl

ClAl

AlCl3

25 26

27

28 29

OHC OAcPPh3 OAc

1) RBr2) NaOH

Scheme 6. Synthesis of the Vitamine A precursor 26 13.2.2. The Horner-Wadsworth-Emmons reaction

An analogous olefination of carbonyl compounds can be achieved if phosphonates instead of phosphonium ylides are used (Horner-Wadsworth Emmons – HWE – reaction) are used, however, distinct differences to the Wittig

Chart 4. Atom economy in the Wittig reaction

Ph3P CH2 + RCHO

R

OPPh3

278g/mol

In a Wittig reaction for each mol product one mole tri-phenylphospine oxide (278g) waste is produced.

Classics in Asymmetric Synthesis

5

reaction previously described exist. Carbanions obtained upon deprotonation of a phosphonate are generally much more reactive than phosphonium ylides, so that they need to be stabilized by an electron withdrawing substituents in order to give useful yields in a subsequent reaction with either an aldehyde or a ketone. As byproduct water soluble phosphates are formed which greatly facilitates the workup of the products. The stereoselectivity of the HEW-reaction can be controlled by the steric and electronic properties of the alcohol groups R2 in the phosphonate: While alkyl groups, especially bulky ones like i-Pr, generally result in high E-alkenes, trifluoroethyl or aryl groups or the use of cyclic phosphonates will lead preferentially to Z-alkenes.

HRO

CH3

O a, b or c

HRO

CH3

CO2R1

HRO

CH3

R1O2C

+

30 (E)-32 (Z)-32

CO2R1P

OR2O

R2O

31

31a: R2 = –CH2CH3; 31b: R2 = –CH2CF3, R2 =H3C

31a: 92 831b: 9 9131c: 1 99

Scheme 7. Reagents and Conditions: (a) 31a, NaH, THF, 83%; (b) 31b, KH, THF, 84%; (c) 31c, NaH, NaI, 88%.

A remarkable control of stereoselectivity with α-fluorinated phosphonates 34 has been reported solely by a combination of the appropriate choice of the group R and the base used for its deprotonation.

O

Ha or b

CO2RP

OR2O

R2O

34

F

H

F CO2R

33

H

RO2C F

+

(Z)-35 (E)-35

R = H 99 1R = Et 6 94

Scheme 8. Reagents and Conditions: (a) 34 (R = H), PriMgBr, 84%; (b) 34 (R = Et), BunLi, 83%.

Masamune and Roush first developed a very useful variant that allows carrying out the HWE-reaction with phosphonates 31 under mild basic conditions. Chelation of the carbonyl and the phosphonate oxygen with LiCl of MgBr2 activates 31 sufficiently to allow its deprotonation with weak bases such as triethylamine, and subsequent reactions with aldehydes yields predominantly E-alkenes.14 The activation by Lewis acids is not even necessary if the reaction is carried out at high pressure.15

Chart 5. The HWE-reaction

R1 H

O+ XP

OR2O

R2O

base

R1 H

X

X = CO2R, C(O)R, CN

E-alkenes Bulky alkyl groups R2 Bulky groups X Z-alkenes R2 = –CH2CF3 (Still-Genari) R2 = –Ar; (Ando) Cyclic phosponates

Chart 6. Common methods for the synthesis of phosponates

(RO)3P + R1CH2X

(RO)2P(O)CH2R1

a) Michaelis-Arbuzov Reaction

(EtO)2P

O

CH3

1) BunLi

2) CuX

3) RCOCl

(EtO)2P

O O

R

b) Alkylation of phosphonates

c) Acylation of phosphonates

1) NaH

2) R1X(EtO)2P

O O

OR

(EtO)2P

O O

OR

R1

Oliver Reiser

6

13.2.3. The Horner-Wittig reaction

Phosphine oxides are yet another class of phosphorus reagents that can be used for carbonyl olefinations, a process that is known as Horner-Wittig reaction. Mechanistically similar to the Wittig and the HWE reaction, an important difference is the possibility to isolate the intermediate β-hydroxy phosphine oxides, when lithium bases and low temperatures are employed. Subsequently, 37 and 39 can be individually transformed after separation by chromatography to the Z-alkene 38 or the E-alkene 40, respectively.

RP

OPh

Ph

36

1) BunLi

2) R1CHO

3) H+

Ph2(O)P

R R1

OH

Ph2(O)P

R R1

OH

R

R1

R1R

38

40

37

39

base

base

Scheme 9. Lithium bases in the Wittig-Horner reaction.

Like the Schlosser variation in the Wittig reaction, there is also the possibility to obtain preferentially E-alkenes 40 from non-stabilized phosphine oxides. Oxidation of the mixture of 37/39 to β-ketophosphine oxides 41, being also accessible directly from 36 by a Claisen type ester condensation, followed by reduction with NaBH4 preferentially gives rise to 39, which can be explained by applying the Felkin-Anh-model.16 Complementary to this, if the reduction is carried out using CeCl3/NaBH4 (Luche reduction) 37 is preferentially obtained via chelate control, which collapses to the Z-alkene 38.

36

37/39

PDC

1) BunLi

2) R1CO2Me

R

O

R1

Ph2(O)P

NaBH4

NaBH4/CeCl3

39

37

40

38

P(O)Ph2

HR

OR1

BH4

40

R

PH

R1O

BH4

38

O

Ph

Ph

Ce3=

41

Scheme 10. Warren variation of the Horner-Wittig reaction.

There are numerous examples for the application of Wittig-type olefinations in natural product synthesis.17 A particular impressive example can

Chart 7. Factors that control the alkene geometry in the Wittig-Horner reaction E-alkenes R anion stabilizing, moderate temperatures, non-Li base, thermodynamic control → one-step protocol Z-alkenes R not anion stabilizing, low temperatures, Li base, kinetic control → two-step protocol

Classics in Asymmetric Synthesis

7

be found in the synthesis of the antibiotic aurodox (42) from the Nicolaou group, in which in a total of six olefination reactions all three Wittig variants were applied.

Me

MeOH

OH

OH

HN

O

OMe

Me

OH H

HO OH

MeN

OH

O

Me

OMe

Horner-Wittig Horner-Wadsworth-Emmons Wittig 42 13.2.4 Asymmetric Wittig-type reactions

On a first glance, one might think that asymmetric olefination reactions are a contradiction in terms since no chiral centers are created by alkenylating a carbonyl group. However, by symmetry-breaking strategies, either by differen-tiating meso-compounds or, in a kinetic resolution, racemic carbonyl compounds asymmetric processes can be developed.18 Most successfully, asymmetric Horner-Wadsworth-Emmons reactions have been developed, making use of three principle strategies to render a phosphonate chiral: (a) Introducing a C2-symmetrical, chiral auxiliary in place of the alkoxy groups in the phosphonate; (b) introducing a non C2-symmetrical, chiral auxiliary in place of the alkoxy groups in the phosphonate, thus making the phosphorus atom itself a chiral center; (c) introducing a chiral auxiliary in the part of the phosphonate that is subsequently transfered to the carbonyl compound.

Phosphonates, being rendered chiral by BINOL19 or the corresponding phosphonic acid bis(amides)20 bearing C2-chiral 1,2-diamines instead of diols are particularly effective for the desymmetrization of prochiral carbonyl compounds. Thus, upon deprotonation with KHMDS and subsequent reaction with 43, the alkene 45 is obtained. Zinc chloride was found to be an effective additive, ensuring in general high yields and selectivities in such transformations.

HH

O O

O

HH

O O

O

OP

O

CO2Me

CO2Me

KHMDS, ZnCl2

43

44

4582%

91% ee

Scheme 11. Asymmetric HWE-reaction with phosphonates bearing a C2-symmetrical chiral auxiliary

The camphorquinone modified phosporamidate 47 also promotes olefinations with high selectivity (Scheme 11).21 The better enantioselectiviy obtained in the transformation of cyclohexanone 46 to 48 in comparison to the use of 44 might very well reflect the greater proximity of the chiral information to the reaction centers, being manifested in a stereogenic phosphorus atom in 47.

Chart 8. Strategies for asymme-tric olefination reactions

OR R

R1

*

R R *

O

O O

R

OR

H

1. Prochiral carbonyl group differentiation

2. Enantiotopic carbonyl group differentiation

3. Kinetic resolution

Oliver Reiser

8

O

But But

CO2Me

KHMDS

NO

P

O

CO2Me

46

47

48 (78%, 86%ee)

(ent)-48 (98%, 79% ee)44, KHMDS, ZnCl2

Scheme 12. Asymmetric HWE-reaction with phosphonates bearing a non C2-symmetrical chiral auxiliary and a stereogenic phosphorus center

Racemic α-chiral aldehydes have been successfully applied in asymmetric HWE-reactions using the strategy of kinetic resolutions.20,22 With configurationally labile substrates even dynamic kinetic resolutions can be carried out if the base used for the deprotonation of the phosphonate is sufficiently strong to racemize the aldehyde by protonation/deprotonation sequences via its enolate.23 Thus, alkenylation of α-aminoaldehyde 49 with the phosphonate 50 in the presence of KHMDS/18-crown-6 gave rise to 51 in good yield and selectivity.23b The stereochemical outcome of this reaction can be rationalized by a transition state 52, in which the chiral auxiliary dictates the face of the phosphonate Z(O)-enolate that is attacked, the R groups on the phosphonate dictate the relative configuration between C2 and C3, and finally selection of the aldehyde enantiomer occurs in accordance with the Felkin-Anh model. Support for this model comes from the fact that in this type of kinetic resolutions the other aldehyde enantiomer reacts preferentially under chelating conditions.24

O

H

NBnp-Ts

Ph + PO

OO

(CF3CH2O)2

Ph1.3 equiv 1.0 equiv

KHMDS (1.2 equiv)

18-crown-6N

Bnp-Ts

Ph

CO2R*

49 50 51(77%, 90%de)

O

P(O)(OR2)

NR2

OH

BnH

53

R2N

Ph

P(O)(OR2

O

CO2R*

52

Felkin-Anh

(Z)-selectivephosphonate

Scheme 13. Dynamic kinetic resolution of α-aminoaldehydes by an asymmetric HWE-reaction 13.3. The Peterson Olefination

Conceptually quite similar to the Wittig type reactions but quite different in scope and limitation are alkenylations that proceed through β-

Classics in Asymmetric Synthesis

9

hydroxysilanes, the so-called Peterson olefination. Most commonly, an α-silylated carbanion is added to a carbonyl compound to give rise to two diastereomeric β-hydroxysilanes, which can be isolated analogous to the previouls discussed Horner-Wittig reaction and separately transformed further to alkenes. In distinct contrast, however, by careful control of the reaction conditions either alkene geometry can be obtained from each individual β-hydroxysilane. Under basic conditions, a syn-elimation of the silyl and hydroxy group takes place, presumably through a pentacoordinated intermediate, which is formed by attack of the deprotoanted oxygen onto silicon. On the other hand, under acidic conditions protonation of the hydroxyl group followed by an E2-elimination occurs, reversing the stereochemical outcome with respect to the base variant.

Based on this mechanistic rational it is possible to synthesize selectively E-alkenes by utilizing the different reaction rates of the diastereomeric β-hydroxysilanes in the elimination step. Upon sulfur-lithium exchange of 54 and reaction with an aldehyde, only 57 undergoes is converted at low temperature to (E)-56, while syn-elimination of 55 is hampered due to steric hindrance between R and Ar. Subsequent addition of acid then converts 55 in an anti-elimination to (E)-56 as well.25

MeMe

Me

SPhMe3Si

R

1) LDMAN

2) ArCHO

O

Me3SiAr

RH

H

O

Me3SiAr

HR

H–78°C

fast

slow R Ar

R

Ar

syn

syn

3) AcOH

anti

54

55 (Z)-56

57 (E)-56 (68%)

x

OBn

OBnAr =

Scheme 14. Stereoconvergent Peterson reaction

Many methods have been utilized to generate α-silyl carbanions, which were subsequently reacted with carbonyl compounds to yield alkenes, but the most valuable strategies are those, which will form β-hydroxysilanes diastereo-selectively, thus making their separation or the combination of different work up protocols superfluous in order to arrive at geometrically pure products.

Peterson reagents being stabilized by electron withdrawing substituents such as aldehydes, ketones or esters, preferentially give rise to (E)-alkenes upon deprotonation with a lithium base without isolation of the intermediate β-hydroxysilanes. The origin of this stereoselectivity has not been extensively explored, but it appears plausible that similar arguments as previously discussed for stabilized Wittig ylides. For example, in the synthesis of maytansine, the trisubstituted alkene 60 was synthesized as a single diastereomer using 59, in which the aldehyde functionality is masked as an imine.

Chart 9. The Peterson olefination

R2R3CO

R1 SiMe3

R2

R2

Me3Si OH

R4R1 R3

R2

Me3Si OH

R3R1 R4

base–Me3SiOH (syn)

base–Me3SiOH (syn)

H=

–Me3SiOH (anti)

R3

R4R2

R1 R4

R3R2

R1

+

Chart 10. Common methods for the generation of α-silyl carbanions. 1. Deprotonation SiMe3

R

Lithiumbase SiMe3

RLi

R = Ar, CO2R, CN, OMe, N(Me)=CNR, Hal, SR, S(O)R, SO2R, P(O)(OR)2

2. Heteroatom Substitution SiMe3

X

Lithiumnaphthalid SiMe3

Li

X = SR, SeMe, SnBu3

or n-BuLi

3. Nucleophile Addition to Vinylsilanes

SiMe3

RLi

SiMe3

R

or RMgBr

R1 R1

M

Oliver Reiser

10

Cl

MeO N

CH3 OMe

OHC CH3

OMEM

COCF3

SS

NBut

HCH3

LiSiMe3

82%

OHC CH3

CH3

OMEM

59

58 60 Scheme 15. Stereoselective Peterson olefination in the synthesis of maytansine.

Alternatively, when non-enolizable alhdeydes are employed, alkenylation can be induced with fluoride, giving rise to (E)-alkenes with high stereoselectivity.26

In contrast, N,N-dibenzyl-(triphenylsilyl)acetamide has been reported to yield with high selectivity (Z)-alkenes upon metallation with LDA followed by reaction with aromatic aldehydes.27 SiPh3

O

NBn2

1) LDA

2) ArCHO

NBn2

OAr

60 61 Scheme 16. Stereoselective synthesis of (Z)-configurated α,β-unsaturated amides.

Allyltrimetylsilanes 62 can readily be deprotonated with s-BuLi, however, upon reaction with an aldehyde 4-hydroxy-1-alkenylsilanes by attack on the γ-position in 63 and not Peterson products, which would require attack on the α-position are often obtained. However, transmetallation of lithium by titanium, boron or magnesium changes the regioselectivity to the α-position, and moreover, anti β-hydroxysilanes are obtained with high diastereoselectivity, which is best explained by a Zimmerman-Traxler type transition state model. Acid or base induced elimination can then be carried out to yield either Z,E or E,E-dienes.

Classics in Asymmetric Synthesis

11

Me3Si R

62

s-BuLiMe3Si R

63

Li

R1CHO

Me3Si R

64

HO R1

1) MX

2) R1CHO

O

M

R1

Me3Si

R

R1 R

OH

SiMe3

KH

H2SO4

R

R1

R

R1

65 66

67

68M = MgBr2, B(OMe)3, Ti(OPri)4, R = SiMe3, SMe, R1 = alkyl, Ph

90-94%

91-95%

Scheme 17. Stereoselective synthesis of 1,3-dienes References 1 2 Modern Carbonyl Olefination; Takeda, T., Ed.; Wiley-VCH: Weinheim,

2004. 3 (a) Maryanoff, B. E.; Reitz, A. B. Chem. Rev. 1989, 89, 863; (b) Vedejs, E.;

Peterson, M. J. In Advances in Carbanion Chemistry; Snieckus, V., Ed.; Jai Press Inc: Greenwich, CT, 1996; Vol. 2.

4 5 6 7 8 9 (a) Bestmann, H. J.; Koschatzky, K. H.; Vostrowsky, O. Chem. Ber. 1979,

112, 1923. (b) Tronchet, Bentzle, Helv. Chim. Acta 1979, 62, 2091. 10 Schlosser, M. Angew. Chem. Int. Ed. Engl. 1966, 5, 126. 11 (a) Corey, E. J.; Katzenellenbogen, Posner, J. Am. Chem. Soc. 1967, 89,

4245. (b) Corey, E. J.; Yamamoto, J. Am. Chem. Soc. 1970, 92, 226, 6636 and 6637.

12 13 Shen, Y. C. Acc. Chem. Res. 1998, 31, 584. 14 (a) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.; Masamune,

S.; Roush, W. R.; Sasaki, T. Tetrahedron Lett. 1984, 25, 2183. (b) Rathke, M. W.; Nowak, M. J. Org. Chem. 1985, 50, 2625. (c) Moison, H.; Texier-Boullet, F.; Foucaud, A. Tetrahedron 1987, 43, 537. (d) Aar, M. P. M. v.; Thijs, L.; Zwanenburg, B. Tetrahedron 1995, 51, 9699.

15 Has-Becker, S.; Bodmann, K.; Kreuder, R.; Santoni, G.; Rein, T.; Reiser, O. Synlett 2001, 1395.

16 Mengel, A.; Reiser, O. Chem. Rev. 1999, 99, 1191. 17 Nicolaou, K. C.; Harter, M. W.; Gunzner, J. L.; Nadin, A. Liebigs Ann.

Chem. 1997, 1283.

Oliver Reiser

12

18 (a) Rein, T.; Reiser, O. Acta Chem. Scand. 1996, 50, 369; (b) Tanaka, K.;

Furuta, T.; Fuji, K. in Modern Carbonyl Olefination; Takeda, T., Ed.; Wiley-VCH: Weinheim, 2004, p 286.

19 Tanaka, K.; Ohta, Y.; Fuji, K.; Taga, T. Tetrahedron Lett. 1993, 34, 4071; 20 (a) Hanessian, S.; Bennani, Y. L.; Delorme, D. Tetrahedron Lett. 1990, 31,

6461; (b) Hanessian, S.; Bennani, Y. L. Tetrahedron Lett. 1990, 31, 6465; (c) Hanessian, S.; Beaudoin, S. Tetrahedron Lett. 1992, 33, 7655.

21 Denmark, S. E.; Rivera, I. J. Org. Chem. 1994, 59, 6887. 22 (a) Rein, T.; Kann, N.; Kreuder, R.; Gangloff, B.; Reiser, O. Angew. Chem.

Int. Ed. Engl. 1994, 33, 556; (b) Rein, T.; Anvelt, J.; Soone, A.; Kreuder, R.; Wulff, C.; Reiser, O. Tetrahedron Lett. 1995, 36, 2303; (c) Pedersen, T. M.; Jensen, J. F.; Humble, R. E.; Rein, T.; Tanner, D.; Bodmann, K.; Reiser, O. Org. Lett 2000, 2, 535-538.

23 (a) Narasaka, K.; Hidai, E.; Hayashi, Y.; Gras, J.-L. J. Chem. Soc. Chem. Commun. 1993, 102-103; (b) Rein, T.; Kreuder, R.; Vonzezschwitz, P.; Wulff, C.; Reiser, O. Angew. Chem. Int. Ed. Engl. 1995, 34, 1023-1025.

24 Kreuder, R.; Rein, T.; Reiser, O. Tetrahedron Lett. 1997, 38, 9035-9038. 25 Huang, W.; Tidwell, T. T. Synthesis 2000, 457-470. 26 27