Tandem silylformylation-allyl(crotyl)silylation: A new approach to polyketide synthesis

Article abstract of DOI:10.1016/j.tet.2003.04.004

Tandem intramolecular silylformylation-allyl(crotyl)silylation reactions have been developed that allow the highly efficient synthesis of polyketide fragments. The substrates are subjected to Rh(I)-catalyzed silylformylation to afford β-(diallyl)silyl ald

Full text of DOI:10.1016/j.tet.2003.04.004

Tetrahedron 59 (2003) 8889–8900
Tandem silylformylation–allyl(crotyl)silylation:
a new approach to polyketide synthesis
*
Michael J. Zacuto, Steven J. O’Malley and James L. Leighton
Department of Chemistry, Columbia University, New York, NY 10027, USA
Received 5 April 2003; accepted 14 April 2003
Abstract—Tandem intramolecular silylformylation–allyl(crotyl)silylation reactions have been developed that allow the highly efficient
synthesis of polyketide fragments. The substrates are subjected to Rh(I)-catalyzed silylformylation to afford b-(diallyl)silyl aldehydes which
undergo spontaneous uncatalyzed allylsilylation. This unusual spontaneous allylsilylation reaction is driven by strain release Lewis acidity,
which arises from the ,958 O–Si–C bond angle in the oxasilacyclopentane product of the silylformylation reaction. The methodology has
been developed both for alkene and alkyne substrates, may be used to establish as many as three stereocenters, and has been shown to be
amenable to use in an iterative fashion.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
Tamao oxidation of the product,5 deliver 1,3,5-triols in a
remarkably efficient manner.
Due to the prevalence and clinical relevance of polyketide-
derived natural products with the skipped polyol structural
motif,1 we have been engaged in the development of
efficient strategies for their synthesis.2,3 One approach has
focused on olefin carbonylation reactions and the general
approach is summarized in Scheme 1. Special emphasis has
been placed on the direct production of suitably protected
polyhydroxy aldehydes, since the next step in the sequence
is the addition of an allyl group. In this fashion a highly
efficient three step iterative sequence—with no protection
steps or oxidation state adjustments—might be realized.
Another important advantage to the direct production of
aldehydes is that, at least in principle, it is possible to
envision the aldehyde allylation step being rendered in
tandem with the carbonylation reaction. Given that the
previously reported intramolecular alkene silylformylation
reaction produces b-silylaldehydes directly (Eq. (1)),2b and
that allylsilanes are well-known aldehyde allylation agents,4
it seemed plausible that silylformylation of a diallylsilane
might lead to intramolecular aldehyde allylsilylation
(Eq. (2)). If successful, this tandem reaction would, upon
ð1Þ
ð2Þ
2. Alkene substrates
Our studies commenced with diallylsilyl ether 1, readily
prepared from the corresponding homoallylic alcohol in a
single step (Scheme 2).6 Subjection of this silane to the
action of 3 mol% Rh(acac)(CO)2 in benzene under 1000 psi
CO at 608C in a stainless steel pressure reactor, followed by
evaporation of solvent and subjection of the residue to the
conditions of the Tamao oxidation (H2O2, NaHCO3, THF/
Scheme 1.
Keywords: polyketides; silylformylation; carbonylation; allylation; strain release Lewis acidity.
*
Corresponding author. Tel.: þ1-212-854-4262; fax: þ1-212-932-1289; e-mail: leighton@chem.columbia.edu
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.04.004

8890
M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
Scheme 2.
MeOH, reflux) led to a mixture of triols in 77% yield. The
major product, easily separable from a mixture of minor
diastereomers, was isolated in 59% yield and was identified
as the syn,syn triol 2.7 1H NMR analysis of the unpurified
product mixture prior to the oxidation revealed a complex
spectrum with no peaks in the aldehyde region. Although
the spectrum has resonances consistent with the dioxasila-
bicyclo[3.3.0]octane depicted in Eq. (2), the peaks are quite
broad and featureless. Mass spectral analysis indicated high
molecular weight material with a repeating fragment pattern
suggestive of an oligomeric mixture. Based on these data it is
proposed that the initially formed dioxasilabicyclo-
[3.3.0]octane undergoes a ring-opening oligomerization
through O–Si bonds, a process driven by relief of ring
strain. Regardless, it is clear that following the carbonyl-
ation an allyl group is transferred, forming a second C–C
bond and a second new stereocenter in a single tandem
reaction.
and alkenyl groups, and reveal the trend of increasing
diastereoselectivity with increasing size of the homoallylic
substituent. The increased diastereoselectivity of entry 4
relative to entry 1 due to the presence of an anti allylic
methyl group mirrors a similar increase observed in the
diphenylsilane silylformylation reaction.2b This suggests
that the modest diastereoselectivity observed for entries 1–3
is due mainly to the modest diastereoselectivity of the
silylformylation reaction, and that the diastereoselectivity of
the allylation reaction is quite high. The use of an achiral
substrate corroborates this interpretation and provides a
useful benchmark diastereoselectivity (93:7) for the allyl-
ation. Finally, we note that in every case we have thus far
studied, the major syn,syn triol is easily separable from a
mixture of the other diastereomers.
In principle, this reaction is amenable to the creation of up to
as many as four new stereocenters by appropriate substi-
tution of both the alkene substrate and the allyl groups on
silicon. Specifically, methyl substitutions could lead to the
polypropionate subset of polyketide-derived natural pro-
ducts. While we have found that substitution on the terminal
carbon of the substrate alkene has a dramatically deleterious
effect on the silylformylation reaction, substitution of the
allyl groups on silicon was far more productive. A synthesis
of di-cis-crotylsilane was developed2i,8 and we devised a
method for its efficient attachment to the substrate alcohols
based on hydrosilane alcoholysis.9,10 Constrained to cata-
lysts that would not also catalyze hydrosilylation of the
products,11 we employed NaH (20 mol%) to catalyze the
alcoholysis of di-cis-crotylsilane, and the use of hexane as
solvent was found to be important to minimize further
reaction of the product, which was observed in more polar
solvents such as THF. Using this procedure, homoallylic
alcohols 3a, 3b, and 3c12 were transformed into di-cis-
crotylsilyl ethers 4a, 4b, and 4c in 90, 99 and 85% yields,
respectively (Scheme 3). Subjection of silanes 4a, 4b, and
4c to the standard reaction conditions led to the isolation of
triols 5a, 5b and 5c, in which three new stereocenters and an
anti terminal propionate unit have been established in a
single tandem reaction. In every case the yield shown is the
isolated yield of purified major diastereomer. The dia-
stereoselectivities (major diastereomer: all other diastereo-
mers) parallel closely those observed in the corresponding
With the feasibility of the tandem reaction established, a
brief survey of the substrate scope was undertaken (Table
1). As with our previous silylformylation study,2b best
results were obtained in benzene under 1000 psi CO at
608C; below ,800 psi CO or ,458C, the rate of the
silylformylation reaction becomes impracticably slow.
Entries 1, 2 and 3 establish the tolerance for both silyloxy
Table 1.
Entry
R1
R2
R3
dra
Yield (%)b
1
2
3
4
5
H
H
H
H
Me
i-Pr
H
H
H
Me
H
77:23
71:29
69:31
92:8
59
45
50
59
65
TBSOCH2CH2
CH2vCHCH2
i-Pr
Me
93:7
All reactions were conducted in a stainless steel pressure reactor equipped
with a pressure guage and a glass liner.
P
syn, syn triol: all other diastereomers.
Isolated yield of purified major product.
a
b
Scheme 3.

M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
8891
Scheme 4.
Table 2.
Entry
R1
R2
R3
dra
Yield (%)b
1
2
3c
4
5
6
7
8
n-Pr
HCuCCH2
TBSOCH2CH2
i-Pr
Ph
t-Bu
i-Pr
i-Pr
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
Me
4:1
4:1
5:1
8:1
7:1
10:1
23:1
7:1
68
63
63
83
83
66
70
70
a
1,5-anti:1,5-syn; measured by gas chromatography.
Isolated yield of the purified mixture of diastereomers.
The TBS group is cleaved during the protodesilylation step and the triacetate is isolated.
b
c
diallylsilane reactions (see Table 1) and we therefore,
conclude that the crotylation event is essentially
stereospecific.
CO in benzene at 608C as the standard reaction conditions.
With optimal reaction conditions identified, we set out to
investigate the scope of the reaction with regard to
homopropargylic and propargylic substituents (Table 2).
Entries 1–6 clearly establish a direct correlation between
the steric size of the homopropargylic substituent and the
diastereoselectivity of the reaction, as well as a tolerance
for alkyne and silyloxy functional groups. The superior
diastereoselectivity of entry 7 indicates the stereo-
chemically reinforcing nature of a syn propargylic methyl
group, whereas with an anti propargylic methyl group (entry
8) useful diastereoselectivity is still observed.
3. Alkyne substrates
Alkynes are well-known substrates for silylformylation,
and it seemed plausible that tandem allylsilylation of the
resultant unsaturated aldehydes might occur as well
(Scheme 4). Interestingly, in this system the carbon bearing
silicon is no longer stereogenic. Thus, any diastereoselec-
tivity in the allylsilylation would have to derive from the
original homopropargylic stereocenter, and would consti-
tute an example of remote 1,5 stereoinduction. To test the
feasibility of this proposal, diallylsilyl ether 6 was prepared
and treated to the silylformylation conditions used for
alkenes (1.0 mol% of Rh(acac)(CO)2, 1000 psi of CO,
benzene, 608C). The unpurified material was directly
subjected to protodesilylation (n-Bu4NF, THF, reflux) and
peracetylation (Ac2O, pyr) to provide an 8:1 mixture of
anti-diacetate 713 and the syn diastereomer in 83% overall
yield. Thus, the allylsilylation proceeded with good 1,5
diastereoselectivity, providing the 1,5 anti product as the
major diastereomer, a result which is opposite to that
observed with alkene substrates.
As an alternative to the protodesilylative workup procedure,
we have also investigated an oxidative workup of the
initially formed vinylsilane products to provide b,b0-di-
hydroxyketone products. Subjection of silane 8 to the
standard tandem silylformylation–allylsilylation conditions
followed by evaporation of the solvent and subjection of the
residue to the conditions of the Tamao oxidation (H2O2,
NaHCO3, THF/MeOH, reflux) led to ketodiol 9 in 71%
yield with identical diastereoselectivity to entry 4 in Table 1
(Scheme 5). In a similar fashion, diallylsilane 10 was
transformed into ketodiol 11 in 65% yield and 23:1
diastereoselectivity. This oxidative workup is noteworthy
in the context of polyketide synthesis in that diastereo-
selective ketone reduction would allow access to stereo-
chemically diverse triol arrays.
Optimization of the reaction conditions revealed that the
catalyst loading could be reduced to 0.1 mol% with no
decrease in efficiency. At CO pressures significantly lower
than 1000 psi, and/or at temperatures significantly below
608C there was a small improvement in the diastereo-
selectivity, accompanied, however, by a significant drop in
reaction rate. We therefore, quickly settled on 1000 psi of
In an effort to expand the scope of the alkyne reaction, the
crotylsilane chemistry discussed above (Scheme 3) was
examined. Thus, alcohol 12 was silylated with di-cis-
crotylsilane and the resulting silyl ether was subjected to the
tandem silylformylation–crotylsilylation conditions using

8892
M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
Scheme 5.
Scheme 6.
the oxidative workup procedure. Ketodiol 13 was isolated in
65% overall yield as an 8:1 mixture of diastereomers
(Scheme 6). Using the protodesilylative workup procedure,
alcohol 14 was transformed into 1513 in 52% overall yield
and 96:4 diastereoselectivity. As above, the diastereoselec-
tivities parallel closely those observed in the corresponding
diallylsilane reactions (see Table 2 and Scheme 5), leading
to the conclusion that the crotylation event is stereospecific.
4. Mechanism
Several intriguing mechanistic issues demanded explica-
tion. First, and most importantly, why do the allyl(crotyl)-
silylations proceed spontaneously in the absence of an
external Lewis acid? Second, what is the origin of the
diastereoselectivity, and why do alkene substrates give
predominantly 1,5-syn products while alkyne substrates
favor the 1,5-anti products? And third, why do cis-
crotylsilanes provide the anti products, and trans-crotyl-
silanes the syn products?
As mentioned above, our attempts to prepare di-trans-
crotylsilane have been fruitless. The most obvious proposal
was to adapt the procedure for the synthesis of trans-crotyl-
trichlorosilane.14 Even under harsh conditions with many
different catalysts however, we could not induce a
successful reaction between dichlorosilane and trans-
crotylchloride. Desiring nevertheless to establish that
trans-crotyl groups would result in syn propionate units in
the tandem intramolecular silylformylation–allylsilylation
chemistry, we synthesized trans-crotyl-phenylsilane as
shown in Scheme 7. In order to avoid the creation of a
mixture of diasteromers in the silane alcoholysis reaction,
we employed achiral alcohol 16. As shown, the silylation
chemistry and the tandem intramolecular silylformylation–
allylsilylation chemistry work equally well with a trans-
crotylsilane to give diol 17 in 64% overall yield.
As expected, a syn-propionate unit was obtained
stereospecifically.
In order to gain mechanistic insight into this unusual
allylsilylation reaction, we began by establishing the
intramolecular nature of the process with the crossover
experiment shown in Scheme 8. Thus, a 1:1 mixture of
diallylsilane 18 and deuterium-labeled diallylsilane 19
(each allyl group carried one deuterium) was subjected to
the standard conditions and produced the four indicated triol
products, each of which was cleanly isolated. No evidence
for the incorporation of any deuterium in the products
derived from 18 could be detected by 1H, 2H and 13C NMR
spectroscopy.
Examples of intramolecular silicon-tethered allylsilane
additions to aldehydes,15 acetoxyazetidinones,16 and
acetals17 have been reported, but require the use of external
Scheme 7.
Scheme 8.

M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
8893
Scheme 9.
Lewis acids. Utimoto has reported uncatalyzed additions—
both inter- and intramolecular—of tetracoordinate allyl-
silanes to aldehydes and ketones based on the use of strained
allylsilacyclobutanes.18 This report followed the disclosures
of Myers19 and Denmark20 that enoxysilacyclobutanes
engage in uncatalyzed aldol addition reactions. These
uncatalyzed reactions proceed through the agency of what
Denmark has termed ‘strain release Lewis acidity.’21 In this
model the ,808 C–Si–C bond angle imposed by the
silacyclobutane favors complexation of an aldehyde with
concomitant Si rehybridization to a trigonal-bipyramid with
the silacyclobutane spanning apical and equatorial positions
(908 ideally). Depsite the fact that the present work involves
5-membered silacycles, we propose that strain release Lewis
acidity is operative and responsible for the spontaneous
allylation chemistry observed (Scheme 9). Two key
observations support this claim. First, the O–Si–C bond
angle in oxasilacyclopentanes is typically ,958,22 which
would be expected to favor a trigonal-bipyramidal geometry
at Si in analogy to the silacyclobutane chemistry.23 And
second, pyrolysis of oxasilacyclopentane 20 with benz-
aldehyde (6.0 equiv.) in a sealed tube at 1308C (Utimoto’s
conditions) produced homoallylic alcohol 7 in 87% yield.24
stereoselectivities, an analysis was carried out based on
the mechanism outlined in Scheme 9 above. To avoid chiral
at silicon complexes we have consistently employed diallyl-
and dicrotylsilanes, and this mechanism clearly implies
that the stereoselectivity of the allylsilylation event is
determined by the relative rates at which the two
diastereotopic allyl groups react. It was important to secure
evidence in support of this hypothesis as it suggested an
intriguing possibility: that if chiral at silicon substrates
bearing only one allyl group could be diastereoselectively
prepared, either diastereomeric product of the allylsilylation
event could be accessed. This would greatly increase the
scope of this methodology.
Analysis of the trigonal-bipyramidal (tbp) conformations
for the Si–O(aldehyde) complex of the intermediate
aldehyde with the constraint that both the oxasilacyclo-
pentane ring and the 5-membered ring formed by Si–O-
(aldehyde) complexation must span an apical and an
equatorial position, reveals only two reasonable possibilities
(tbp-1 and tbp-2, Scheme 10). Transfer of allyl group a in
tbp-1 (path a) is expected to be favored, since it is closer to
the aldehyde than allyl group b, and leads to the observed
major syn product. In principle, transfer of allyl group b
(path b-1) could be the source of the minor anti product, but
In an effort to develop a model for the observed
Scheme 10.
Scheme 11.

8894
M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
Scheme 12.
this pathway should be highly disfavored since the initial
product would be a highly strained trans-fused dioxasila-
bicyclo[3.3.0]octane. Instead, it is proposed that the minor
diastereomer is formed by way of a pseudorotation to
tbp-2,25 followed by transfer of allyl group b (path b-2),
leading to production of the illustrated cis-fused dioxasila-
bicyclo[3.3.0]octane.
determined by the relative rates at which the diastereotopic
allyl groups transfer (path a vs. path b). A simple working
hypothesis for the observed preference for a path would
invoke a destabilizing steric interaction between the i-Pr
group and allyl group b. The correlation between steric size
of the homopropargylic substituent and the selectivity
observed in entries 1–6 of Table 2 is consistent with this
hypothesis as is the high selectivity observed in entry 7. The
nearly identical selectivity observed in entries 4 and 8 is less
well rationalized by this model and is suggestive of more
subtle stereoelectronic effects. Along these lines it is
interesting to note that conformational analysis (envelope
conformation with the i-Pr group pseudoequatorial)
suggests that there should be a nearly antiperiplanar
relationship between a lone pair on the oxygen and the
Si–C bond of the back allyl group, and a nearly orthogonal
relationship between the other oxygen lone pair and the
Si–C bond of the front allyl group.
To shed light on this model, allylphenylsilane 21 was
prepared and subjected to the tandem silylformylation–
allysilylation reaction (Scheme 11). Presumably proceeding
by way of a mixture of diastereomeric aldehydes 22 and 23,
upon oxidative workup the reaction produced syn-triol 24
stereospecifically (recall that the diallyl version gave a 93:7
dr, Table 1, entry 5). None of the anti diastereomer could
be detected. When the unpurified reaction mixture was
1
analyzed by H NMR spectroscopy prior to the oxidation
step, for the first time we observed an aldehyde inter-
mediate. Unable to achieve a stereochemical proof for this
aldehyde due to its instability, we may only tentatively
assign it structure 23. Nevertheless, the apparent reluctance
of this aldehyde to undergo allylation to give the anti-triol
diastereomer may be rationalized within the context of our
model by the assertion that the initially formed complex
does not undergo pseudorotation to the reactive complex.
This pseudorotation places the phenyl group in an apical
position, and in his classic studies of pseudorotation of
pentacoordinate siliconates, Martin has provided evidence
that this would be expected to be disfavored relative to more
electronegative substituents.26
Like the model outlined above for the alkene substrate
reactions, this model posits that both of the diastereotopic
allyl groups can transfer and each leads to a different
product diastereomer, but without the need for pseudo-
rotation. It could therefore be surmised that selective
replacement of either allyl group with a non-transferable
group would lead to a stereospecific reaction. To gain
support for this hypothesis we prepared silane 26 as a 1:1
mixture of diastereomers. Subjection of this mixture to the
standard reaction conditions led, presumably by way of the
illustrated 1:1 mixture of aldehydes, to a 1:1 mixture of diol
diastereomers 27 and 28 (Scheme 13). The clear implication
of this experiment is that either product diastereomer could
be selected for, based only on the availability of the starting
chiral silanes in diastereomerically pure form. In fact, we
have recently reported a new asymmetric silane synthesis
and have employed it to establish conclusively the
correctness of this model.27
In analogy to this model, a similar scenario may be
developed for the reactions of alkyne substrates. A key
difference, however, is that the presence of the unsaturation
in the aldehyde intermediate obviates consideration of
pseudorotation of the type discussed above as this would
lead to more highly strained complexes. Thus, binding of
the aldehyde by the Lewis acidic silicon is followed by
intramolecular allylsilylation as depicted for aldehyde 25
(Scheme 12). In this model, the diastereoselectivity is
Finally, we note that the unusual observation of anti-propio-
nate units from cis-crotyl groups and of syn-propionate units
Scheme 13.
Scheme 14.

M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
8895
from trans-crotyl groups was predicted by, and is fully
consistent with the proposed models. Thus, the intra-
molecular nature of the reaction necessitates that the alkyl
chain of the aldehyde occupy a pseudo-axial position on a
chair-like arrangement of the 6 reacting atoms (Scheme 14).
Thus, the reacting diasteroface of the aldehyde is the reverse
of what is normally postulated, leading to the observed
reversal from the usual sense of induction.28
90%, with the more sterically hindered alcohols giving
better results.
6.1.1. Preparation of di-cis-crotylsilane. In an inert
atmosphere glove box a 75 mL sealed tube was charged
with Pd(PPh3)4 (290 mg, 0.25 mmol, 0.5 mol%). The tube
was fitted with a septum, removed from the glove box,
placed under an Ar atmosphere and cooled (2788C).
1,3-Butadiene (13.0 mL, 150 mmol; condensed in a cooled
(2788C) graduated cylinder fitted with a septum) was added
by cannula to the sealed tube. Dichlorosilane (4.2 mL,
50 mmol; condensed in a cooled (2788C) graduated
cylinder fitted with a septum) was added by cannula to the
sealed tube. The tube was sealed with a teflon stopper and
the solution was stirred at 2788C for 3 h. The solution was
then allowed to warm to ambient temperature, and the
resulting heterogeneous solution was vigorously stirred for
48 h. The resultant orange–brown heterogeneous mixture
was transferred to a 25 mL round bottom flask and distilled
under reduced pressure (reduced slowly, to allow for the
removal of excess 1,3-butadiene) to afford dichlorodi-cis-
crotylsilane (bp <608C at 5 Torr) as a clear, colorless liquid.
5. Conclusion
The methodology reported herein allows the highly efficient
assemblage of fragments useful for polyketide-derived
natural products synthesis. We have reported one such
application,29 and several others are currently in progress.
Of further note, the methodology typically requires the
use of only inexpensive and readily available reagents
(e.g. HSiCl3, AllylMgBr, CO, NaHCO3, H2O2, n-Bu4NF).
Conceptually, the focus on methods that directly produce
aldehydes (i.e. carbonylation) has indeed led to the
development of tandem reactions for polyketide synthesis,
and the use of carbonylation in this respect is certainly
novel. Perhaps more importantly, this work has led to the
discovery that strain release Lewis acidity-induced allyl-
silylation chemistry may be accessed using 5-membered
silacycles and is not restricted to the use of silacyclobutanes.
This discovery has profound implications for the develop-
ment of new asymmetric reaction methodologies as has
recently been demonstrated.30
A solution of the dichlorodi-cis-crotylsilane in 10 mL THF
was added dropwise by addition funnel to a solution of
LiAlH4 (1.05 g, 27. 5 mmol, 0.56 equiv.) in 40 mL THF.
The solution was heated at reflux for 22 h. The resultant
heterogeneous mixture was cooled, and filtered through a
pad of Celite, rinsing with pentane. The filtrate was
transferred to a separatory funnel and washed with H2O
(2£10 mL) and brine (10 mL). The organic layer was dried
over Na2SO4, filtered, and concentrated. The residue was
distilled under reduced pressure to afford 5.5 g (75%) of di-
cis-crotylsilane (bp <558C at 15 Torr) as a clear, colorless
1
liquid. H NMR (400 MHz, CDCl3) d 5.38–5.49 (m, 4H,
6. Experimental
Si(CH2CHvCHCH3)2), 3.67 (quint, 2H, J¼3.6 Hz, 2H,
SiH2), 1.65–1.68 (m, 4H, Si(CH2CHvCHCH3)2), 1.59 (d,
6H, J¼5.8 Hz, Si(CH2CHvCHCH3)2); 13C NMR
(100 MHz, CDCl3) d 125.5, 122.8, 12.5, 10.2; IR (thin
6.1. General procedure for the preparation of diallylsilyl
ethers
film) 3019, 2930, 2136, 1652, 1396, 1362 cm21
.
A three neck flask equipped with an addition funnel is
charged with a solution of HSiCl3 (15 mmol) in Et2O
(20 mL) and the solution is cooled to 2788C. A solution of
the homoallylic alcohol (10 mmol) in Et2O (10 mL) is
transferred to the addition funnel and added dropwise over
30 min, with venting of HCl to a bubbler charged with 2 M
NaOH. Upon completion of the addition, the reaction
mixture is allowed to warm to room temperature, and the
volatiles (HSiCl3, HCl and Et2O) are removed by evapo-
ration with a positive flow of N2. The residue is dissolved
in 20 mL of Et2O and the resulting solution is cooled to
2788C. Allylmagnesium bromide (20 mL, 20 mmol, 1.0 M
in Et2O) is then transferred to the addition funnel and added
dropwise with vigorous stirring. Upon completion of the
addition, the solution is warmed to room temperature, and
then filtered through a pad of Celite, with pentane washes.
The filtrate is concentrated in vacuo, and the residue is
filtered again through a pad of celite with pentane washes.
After concentration, the resulting material generally is
.90% pure and can be used without further purification.
The major byproduct is triallylsilane, which can be removed
in vacuo. If desired, the products could be purified by
distillation under reduced pressure. For the substrates
reported here, the yields of purified material were 60–
6.1.2. Preparation of trans-crotyl-phenylsilane. To a
solution of trans-crotyl-trichlorosilane14 (10.0 g, 52.7 mmol)
in 100 mL Et2O was slowly added phenylmagnesium
chloride (29.0 mL, 57.9 mmol, 2.0 M in THF) by cannula
at room temperature with vigorous stirring. Upon comple-
tion of addition, the reaction mixture was heated at reflux for
3 h, then cooled. The reaction mixture was filtered through a
pad of celite and the filtrate was concentrated. The residue
was distilled under vacuum (,1.0 Torr; bp 78–808C) to
give a clear liquid (trans-crotyl-phenyldichlorosilane). To a
solution of LiAlH4 (0.46 g, 12.3 mmol) in 50 mL THF was
slowly added a solution of this liquid in 25 mL THF.
The mixture was heated at reflux for 48 h, then cooled.
The reaction mixture was slowly added to a large flask
containing ice water, with vigorous stirring, and the
resulting mixture was extracted with pentane (3£50 mL).
The combined organic extracts were dried (MgSO4), filtered
and concentrated. The residue was distilled under vacuum
(,1.0 Torr; bp 56–588C) to give 3.17 g (38%) of trans-
crotyl-phenylsilane as a clear colorless liquid. 1H NMR
(400 MHz, CDCl3) d 7.59–7.56 (m, 2H, two of C6H5),
7.43–7.34 (m, 3H, three of C6H5), 5.51–5.38 (m, 2H,

8896
M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
CHvCH), 4.31–4.29 (apparent dt, 2H, J¼1.7, 3.7 Hz,
SiH2), 1.85–1.82 (m, 2H, SiCH2), 1.61 (m, 3H,
CHvCHCH3); 13C NMR (75 MHz, CDCl3) d 135.2,
132.1, 130.0, 127.9, 125.9, 125.2, 18.0, 15.6; IR (thin
film) 3070, 3052, 3015, 2960, 2933, 2917, 2883, 2855,
2138, 1429, 1304, 1163, 1117, 1068, 1048 cm21; LRMS
(EI) calcd for C10H14Si 162.09, found 162.01.
with EtOAc (5£10 mL). The combined organic layers are
dried over MgSO4, filtered, and concentrated. The residue is
purified by flash chromatography on silica gel with
EtOAc/hexanes as eluant. In every case reported here, the
major syn,syn-triol eluted first, followed by a mixture of
other triol diastereomers.
6.5. General procedure for the silylformylation–
allylation reaction of alkyne substrates
6.2. General procedure for the preparation of
di-cis-crotylsilyl ethers
A glass liner for a stainless steel 45 mL Parr high pressure
reactor equipped with a stir bar and septum is charged with a
solution of the diallylsilyl ether substrate (2.0 mmol) in
6.0 mL benzene. The solution is cooled to 2788C and
Rh(acac)(CO)2 (0.5 mg, 0.002 mmol) is added. The liner is
inserted into the Parr reactor, and the pressure gauge and gas
inlet assembly is attached. The reactor is charged to 500 psi
with CO, and vented. The reactor is charged to 1000 psi
with CO and is then immersed (,2.5 cm) in an oil bath at
608C. After 2–3 h, the reactor is cooled to 08C and then
vented. The solution is concentrated and the residue is
immediately subjected to either workup procedure without
further purification.
A round bottom flask is charged with sodium hydride
(8.0 mg, 0.20 mmol, 60% wt. dispersion in mineral oil) and
5 mL hexanes is added. The mixture is stirred vigorously
under Ar for 3 min, and then the hexanes/mineral oil
solution is removed by cannula. This process is repeated,
and 2 mL of hexanes is then added. Di-cis-crotylsilane is
then added by syringe (154 mg, 0.197 mL, 1.1 mmol),
followed by a solution of the homoallylic alcohol
(1.0 mmol) in 2 mL hexanes by cannula. The resulting
solution is heated at reflux for the required time (typically
45–90 min). The solution is cooled, diluted with hexanes,
and filtered through a pad of Celite. The filtrate is
concentrated to afford cude material which is generally of
sufficient purity for use in the tandem silylformylation–
crotylsilylation reaction. Analytical samples of the di-cis-
crotylsilyl ethers may be obtained by distillation under
reduced pressure.
Protodesilylation workup. To a solution of the residue from
the tandem silylformylation–allylsilylation in 10.0 mL
THF is added tetra-n-butylammonium fluoride (6.0 mL,
6.0 mmol, 1.0 M in THF). The solution is heated at reflux
for 2 h, and then cooled. Saturated aqueous NH4Cl is added,
and the mixture is extracted with Et2O. The combined
organic layers are dried (MgSO4), filtered and concentrated.
The resulting diols may be purified by chromatography on
silica gel.
Note. The time of the reaction must be controlled as the
monoalkoxy silane will convert to the dialkoxysilane at
longer reaction times. The reaction time necessary for full
conversion of the alcohol to the monoalkoxysilane varies
with the steric hinderance of the alcohol, and should be
optimized for each alcohol.
Tamao oxidation workup. To a solution of the residue from
the tandem silylformylation–allylsilylation in 3.0 mL THF
and 3.0 mL MeOH is added NaHCO3 (0.15 g, 1.8 mmol)
and H2O2 (1.5 mL, 35% in H2O). The solution is heated
at reflux for 3 h, and then cooled. Saturated aqueous NaCl
is added, followed by saturated aqueous Na2S2O3 and the
mixture is extracted with Et2O. The combined organic
layers are dried (MgSO4), filtered and concentrated. The
resulting ketodiols may be purified by chromatography on
silica gel.
6.3. General procedure for the silylformylation–
allylation reaction of alkene substrates
To an oven-dried 45 mL stainless steel Parr bomb equipped
with a glass liner, stir bar and septum is added the substrate
(1.0 mmol), by cannula, as a solution in benzene (8 mL).
The solution is cooled to 2788C until frozen. Rh(acac)-
(CO)2 (7.7 mg, 0.03 mmol) is then added, and the bomb is
assembled and pressurized to 1000 psi of CO and vented.
This purge is repeated twice and then the bomb is
pressurized to 1000 psi with CO at 2788C. The apparatus
is then immersed in an oil bath and heated at 608C for 24 h.
After cooling to 08C, the bomb is vented. The solution is
concentrated and the residue is used without further
purification in the Tamao oxidation reaction.
1
6.5.1. Entry 1, Table 1. H NMR (500 MHz, CDCl3) d
5.77–5.86 (m, 1H, C(9)H), 5.12–5.15 (m, 2H, C(7)H2),
4.32 (s, 1H, OH), 4.08–4.13 (m, 1H, C(7)H), 3.93–3.94 (m,
1H, C(5)H), 3.65–3.68 94 (m, 1H, C(3)H), 3.11 (s, 1H,
OH), 2.26 (s, 1H, OH), 2.23–2.26 (m, 2H, C(8)H), 1.62–
1.69 (m, 1H, C(2)HCH3), 1.52–1.62 (m, 4H, C(4)H2 and
C(6)H2), 0.92 (d, J¼6.8 Hz, 3H, C(1)H3), 0.91 0.92 (d,
J¼6.8 Hz, 3H, C(2)HCH3); 13C NMR (75 MHz, CDCl3) d
134.3, 118.2, 77.8, 73.9, 71.7, 43.0, 42.5, 39.9, 34.2, 18.2,
17.5; IR (thin film) 3400, 3078, 2958, 1737, 1644, 1440,
1375, 1247 cm21; HRMS (FAB Mþ1) calcd for C11H22O3
202.1569, found 202.1648.
6.4. General procedure for the Tamao oxidation
The product from the silylformylation–allylation reaction is
dissolved in a 1:1 mixture of THF and MeOH (6 mL total).
To this solution is added NaHCO3 (120 mg, 1.5 mmol),
followed by H2O2 (1.6 mL, 15 mmol, 30% solution in H2O).
The flask is then fitted with a reflux condenser and the
solution is heated to reflux for 30–60 min. The solution is
cooled to room temperature and 2.0 mL of saturated
aqueous Na2S2O3 is added. The biphasic solution is filtered
through a cotton plug into a separatory funnel and extracted
1
6.5.2. Entry 2, Table 1. H NMR (400 MHz, CDCl3) d
5.83–5.87 (m, 1H, C(9)H), 5.07–5.13 (m, 2H, C(10)H2),
4.53 (s, 1H, OH), 4.2 (s, 1H, OH), 4.12–4.16 (m, 2H,
C(1)H2), 3.83–3.96 (m, 3H, C(3)H and C(5)H) and C(7)H),
3.75 (s, 1H, OH), 2.21–2.27 (m, 2H, C(8)H2), 1.47–1.73

M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
8897
(m, 6H, C(2)H2 and C(4)H2 and C(6)H2), 0.89 (s, 9H, Si–
C(CH3)3), 0.08 (s, 6H, Si–(CH3)2; 13C NMR (100 MHz,
CDCl3) d 134.9, 117.4, 73.4, 73.3, 71.5, 62.6, 43.6, 42.8,
38.5, 25.8, 18.1, 25.6, 25.6; IR (thin film) 3372, 2930,
1H, CHOH), 3.42–3.65 (br s, 3H, three OH), 3.41 (dd,
J¼2.6, 8.9 Hz, 1H, CHOH), 2.20–2.27 (m, 1H, CHCH3),
1.85–1.89 (m, 1H, CH(CH3)2), 1.76–1.80 (m, 1H, one of
CH2), 1.60–1.64 (m, 1H, CHCH3), 1.41–1.47 (m, 1H, one
of CH2), 1.05 (d, J¼6.9 Hz, 3H, CHCH3), 0.99 (d, J¼
6.9 Hz, 3H, CHCH3), 0.85 (d, J¼6.8 Hz, 3H, one of
CH(CH3)2), 0.77 (d, J¼6.9 Hz, 3H, one of CH(CH3)2);
13C NMR (100 MHz, CDCl3) d 140.0, 116.1, 80.5, 77.8,
76.0, 44.6, 41.8, 37.0, 29.9, 20.2, 15.7, 13.9, 13.1; IR
(CH2Cl2) 3400, 3082, 2967, 2932, 2875, 1638, 1462, 1436,
1387, 1320, 1266, 1143, 1095 cm21; HRMS (FAB, MþH)
calcd for C13H26O3 231.1882, found 231.1953.
2852, 2332, 1636, 1464, 1435, 1361, 1322, 1253 cm21
;
HRMS (FAB Mþ1) calcd for C16H34O4 318.2226, found
318.2308.
1
6.5.3. Entry 3, Table 1. H NMR (400 MHz, CDCl3) d
5.75–5.88 (m, 2H, C(2)H and C(10)H), 5.10–5.16 (m, 4H,
C(1)H2 and C(10)H2), 4.33 (s, 1H, OH), 4.13 (quint., J¼
6.2 Hz, 1H, C(6)H, 3.91–3.95 (m, 2H, C(4)H and C(8)H),
3.19 (s, 1H, OH), 2.22–2.27 (m, 4H, C(3)H2 and C(9)H2),
1.55–1.60 (m, 4H, C(5)H2 and C(7)H2); 13C NMR
(100 MHz, CDCl3) d 134.3, 118.2, 73.4, 71.6, 42.9, 42.4;
IR (thin film) 3356, 3069, 2927, 2355, 1838, 1641, 1440,
1324 cm21; HRMS (FAB Mþ1) calcd for C11H20O3
200.1412, found 200.1496.
1
6.5.8. Triol 5c, Scheme 3. H NMR (400 MHz, CDCl3) d
5.72–5.81 (m, 1H, CHvCH2), 5.10–5.16 (m, 2H,
CHvCH2), 4.25–4.32 (m, 1H, CHOH), 3.72–3.76 (m,
1H, CHOH), 3.20–3.51 (br s, 3H, three OH), 2.19–2.28 (m,
1H, CHCH3), 1.71–1.79 (m, 1H, one of CH2), 1.47–1.56
(m, 3H, one of CH2 and CH2), 1.35 (s, 3H, one of C(CH3)2),
1.27 (s, 3H, one of C(CH3)2), 1.05 (d, J¼6.9 Hz, 3H,
CHCH3); 13C NMR (100 MHz, CDCl3) d 140.0, 116.4,
75.8, 71.4, 70.7, 48.2, 44.8, 40.4, 31.9, 27.7, 15.6; IR (thin
film) 3369, 3079, 2972, 2940, 1640, 1436, 1380, 1367,
1330 cm21; HRMS (FAB, MþH) calcd for C12H24O3
203.1725, found 203.1645.
1
6.5.4. Entry 4, Table 1. H NMR (400 MHz, CDCl3) d
5.77–5.87 (m, 1H, C(9)H), 5.11–5.15 (m, 2H, C(10)H2),
4.77 (s, 1H, OH), 3.87–3.95 (m, 2H, C(5)H and C(7)H),
3.52 (d, J¼1.9 Hz, 1H, OH), 3.39–3.42 (m, 1H, C(3)H),
3.02 (d, J¼3.9 Hz, 1H, OH), 2.23–2.27 (m, 2H, C(8)H2),
1.87 (m, 1H, C(4)H), 1.74–1.78 (m, 1H, one of C(6)H),
1.60–1.66 (m, 1H, C(2)H), 1.41–1.50 (m, 1H, one of
C(6)H), 0.99 (d, J¼6.92 Hz, 3H, C(4)HCH3), 0.85 (d, J¼
6.79 Hz, 3H, C(1)H3), 0.78 (d, J¼6.89 Hz, 3H, C(2)HCH3);
13C NMR (100 MHz, CDCl3) d 134.5, 118.1, 80.8, 77.7,
72.0, 42.5, 41.8, 39.7, 29.9, 20.1, 13.9, 13.0; IR(CDCl3)
1
6.5.9. Entry 1, Table 2. H NMR (300 MHz, CDCl3) d
5.70–5.53 (m, 2H, CHvCH2 and CH2CHvCH), 5.40
(dd, 1H, J¼15.4, 6.9 Hz, CHvCHCHOAc), 5.18 (apparent
q, 1H, J¼6.6 Hz, CHvCHCHOAc), 5.00 (m, 2H,
CHvCH2), 4.81 (m, 1H, n-PrCHOAc), 2.30–2.11 (m,
4H, CH2CHvCH and CH2CHvCH2), 1.95 (s, 3H, one of
CH3CO2CH), 1.94 (s, 3H, one of CH3CO2CH), 1.47–1.31
(m, 2H, CH3CH2CH2), 1.29–1.18 (m, 2H, CH3CH2CH2),
0.83 (t, 3H, J¼7.2 Hz, CH3CH2CH2); 13C NMR (300 MHz,
CDCl3) d 170.4, 169.9, 133.0, 130.8, 128.9, 117.6, 73.2,
72.6, 38.8, 36.9, 35.5, 21.0, 20.9, 18.3, 13.7; IR (thin film)
3079, 2960, 2875, 1740, 1435, 1373 cm21; HRMS (FABþ,
NaI) calcd for C15H24O4Na 291.1572, found 291.1582.
3434, 3021, 2969, 2401, 1522, 1469, 1429, 1216 cm21
;
HRMS (FAB Mþ1) calcd for C12H24O3 216.1725, found
216.1800.
1
6.5.5. Entry 5, Table 1. H NMR (400 MHz, CDCl3) d
5.76–5.86 (m, 1H, C(9)H), 5.11–5.15 (m, 2H, C(9)H2),
4.36 (s, 1H, OH), 4.24–4.30 (m, 1H, C(6)H), 3.95 (m, 1H,
C(4)H), 3.08 (s, 1H, OH), 2.89 (s, 1H, OH), 2.23–2.26 (m,
2H, C(7)H2), 1.72 (dd, J¼10.9, 14.4 Hz, 1H, one of C(3)H),
1.50–1.61 (m, 2H, C(5)H2), 1.46 (dd, J¼2.1, 14.4 Hz, 1H,
one of C(3)H), 1.33 (s, 3H, C(1)CH3), 1.26 (s, 3H, C(2)CH3:
13C NMR (100 MHz, CDCl3)) d 134.4, 118.2, 71.7, 71.6,
70.5, 48.3, 43.1, 42.5, 32.0, 27.7; IR (thin film) 3359, 2963,
2928, 1718, 1640, 1430, 1369, 1326, 1145 cm21; HRMS
(FAB Mþ1) calcd for C10H20O3 188.1412, found 188.1488.
1
6.5.10. Entry 2, Table 2. H NMR (300 MHz, CDCl3) d
5.70–5.42 (m, 3H, CHvCH2 and CHvCH), 5.17 (apparent
q, 1H, J¼6.5 Hz, CHvCHCHOAc), 5.00 (m, 2H,
CHvCH2), 4.87 (m, 1H, CH2CHOAcCH2), 2.45–2.26
(m, 6H, CH2CHOAcCH2 and CH2CHvCH2), 1.97 (s, 3H,
one of CH3CO2CH), 1.96 (m, 1H, HCuCCH2), 1.95 (s, 3H,
one of CH3CO2CH); 13C NMR (300 MHz, CDCl3) d 170.0,
169.9, 132.9, 131.6, 127.8, 117.7, 79.2, 73.1, 70.6, 70.4,
38.7, 35.5, 23.1, 21.0, 20.8; IR (thin film) 3293, 3088, 2989,
1732, 1434, 1242 cm21; HRMS (FABþ, NaI) calcd for
C15H20O4Na 287.1259, found 287.1259.
1
6.5.6. Triol 5a, Scheme 3. H NMR (400 MHz, CDCl3) d
5.69–5.77 (m, 1H, CHvCH2), 5.07–5.12 (m, 2H,
CHvCH2), 4.52 (s, 1H, OH), 4.05–4.10 (m, 1H, CHOH),
3.72–3.67 (m, 1H, CHOH), 3.67–3.62 94 (m, 1H, CHOH),
3.31 (s, 1H, OH), 3.16 (s, 1H, OH), 2.18–2.23 (m, 1H,
CHCH3), 1.47–1.67 (m, 5H, (CH3)2CHCHCH2CHCH2),
1.02 (d, J¼6.9 Hz, 3H, CHCH3), 0.91 (d, J¼6.8 Hz, 3H, one
of CH(CH3)2), 0.89 (d, J¼6.8 Hz, 3H, one of CH(CH3)2);
13C NMR (100 MHz, CDCl3) d 140.0, 116.3, 77.6, 75.7,
73.9, 44.7, 40.3, 39.9, 34.0, 18.1, 17.5, 15.5; IR (thin film)
3368, 3079, 2961, 1737, 2875, 1640, 1442, 1395, 1370,
cm21; HRMS (FAB, MþH) calcd for C12H24O3 217.1725,
found 217.1809.
1
6.5.11. Entry 3, Table 2. H NMR (300 MHz, CDCl3) d
5.75–5.57 (m, 2H, CHvCH2 and CH2CHvCH), 5.47 (dd,
1H, J¼15.4, 6.8 Hz, CHvCHCHOAc), 5.23 (apparent
q, 1H, J¼6.5 Hz, CHvCHCHOAc), 5.08–4.94 (m, 3H,
CHvCH2 and CH2CHOAcCH2), 4.06 (t, 2H, J¼6.4 Hz,
AcOCH2CH2), 2.38–2.12 (m, 4H, CH2CHvCH and
CH2CHvCH2), 2.02 (s, 3H, one of CH3CO2CH), 2.01 (s,
3H, one of CH3CO2CH), 2.00 (s, 3H, one of CH3CO2CH),
1.87–1.74 (m, 2H, AcOCH2CH2); 13C NMR (300 MHz,
CDCl3) d 170.9, 170.4, 170.1, 133.0, 131.5, 128.2, 117.8,
73.2, 69.8, 60.6, 38.9, 37.0, 32.4, 21.2, 21.0, 20.8; IR (thin
1
6.5.7. Triol 5b, Scheme 3. H NMR (400 MHz, CDCl3) d
5.72–5.79 (m, 1H, CHvCH2), 5.08–5.13 (m, 2H,
CHvCH2), 3.85–3.89 (m, 1H, CHOH), 3.68–3.72 (m,

8898
M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
film) 3079, 2962, 2856, 1732, 1435, 1242 cm21; HRMS
(FABþ, NaI) calcd for C16H24O6Na 335.1471, found
335.1464.
(m, 2H, CHvCH2), 4.61 (apparent t, 1H, J¼6.2 Hz,
(CH3)2CHCHOAc), 2.42 (m, 1H, CHOAcCH(CH3)),
2.37–2.30 (m, 2H, CH2CHvCH2), 2.02 (s, 3H, one of
CH3CO2CH), 2.01 (s, 3H, one of CH3CO2CH), 1.79 (m, 1H,
(CH3)2CH), 0.94 (d, 3H, J¼6.9 Hz, CH(CH3)CHvCH),
0.84 (d, 3H, J¼6.7 Hz, one of (CH3)2CH), 0.82 (d, 3H,
J¼6.7 Hz, one of (CH3)2CH); 13C NMR (300 MHz, CDCl3)
d 170.8, 170.1, 134.9, 133.2, 128.7, 117.7, 80.6, 73.3, 39.1,
38.8, 29.4, 21.8, 20.8, 19.1, 17.5, 17.2; IR (thin film) 3080,
1
6.5.12. Entry 4, Table 2. H NMR (300 MHz, CDCl3) d
5.71–5.53 (m, 2H, CHvCH2 and CH2CHvCH), 5.41 (dd,
1H, J¼15.5, 6.9 Hz, CHvCHCHOAc), 5.17 (apparent q,
1H, J¼6.6 Hz, CHvCHCHOAc), 5.00 (m, 2H, CHvCH2),
4.67 (m, 1H, i-PrCHOAc), 2.30–2.13 (m, 4H, CH2CHvCH
and CH2CHvCH2), 1.95 (s, 3H, one of CH3CO2CH), 1.95
(s, 3H, one of CH3CO2CH), 1.74 (m, 1H, (CH3)2CH),
0.82 (d, 6H, J¼6.8 Hz, (CH3)2CH); 13C NMR (300 MHz,
CDCl3) d 170.4, 169.9, 133.0, 130.6, 129.3, 117.6, 77.0,
73.2, 38.8, 34.2, 30.8, 21.0, 20.8, 18.4, 17.4; IR (thin film)
3080, 2966, 1732, 1644, 1468, 1242 cm21; HRMS (FABþ,
NaI) calcd for C15H24O4Na 291.1572, found 291.1581.
2968, 2937, 2877, 1737, 1644, 1462, 1434, 1372 cm21
;
HRMS (FABþ, NaI) calcd for C16H26O4Na 305.1729,
found 305.1729.
6.5.17. Ketodiol 9, Scheme 5. 1H NMR (300 MHz, CDCl3)
d 5.77–5.63 (m, 1H, CHvCH2), 5.03–4.97 (m, 2H,
CHvCH2), 4.05 (m, 1H, CH2CH(OH)CH2), 3.74 (m, 1H,
iPrCHOH), 2.52–2.35 (m, 4H, CH2COCH2), 2.22–2.06
(m, 2H, CH2CHvCH2), 1.56 (m, 1H, (CH3)2CH), 0.80
(apparent t, 6H, J¼6.6 Hz, (CH3)2CH); 13C NMR
(300 MHz, CDCl3) d 212.4, 133.9, 117.6, 71.9, 66.7, 49.3,
47.0, 40.0, 33.0, 18.1, 17.5; IR (thin film) 3418, 3075, 2962,
2933, 1708, 1642, 1386 cm21; HRMS (FABþ, NaI) calcd
for C11H20O3Na 223.1310, found 223.1301.
1
6.5.13. Entry 5, Table 2. H NMR (300 MHz, CDCl3) d
7.35–7.23 (m, 5H, C6H5), 5.77 (dd, 1H, J¼7.0, 6.5 Hz,
C6H5CHOAc), 5.71–5.54 (m, 2H, CHvCH2 and
CH2CHvCH), 5.44 (m, 1H, CHvCHCHOAc), 5.21
(apparent q, 1H, J¼6.6 Hz, CHvCHCHOAc), 5.05
(m, 2H, CHvCH2), 2.65–2.48 (m, 2H, PhCHOAcCH2),
2.37–2.20 (m, 2H, CH2CHvCH2), 2.04 (s, 3H, one of
CH3CO2CH), 1.99 (s, 3H, one of CH3CO2CH); 13C NMR
(300 MHz, CDCl3) d 172.2, 139.7, 133.1, 131.4, 128.3,
127.9, 126.4, 117.7, 74.9, 73.3, 39.1, 38.7; IR (thin film)
3067, 2940, 1732, 1643, 1496, 1234 cm21; HRMS (FABþ,
NaI) calcd for C18H22O4Na 325.1416, found 325.1406.
6.5.18. Ketodiol 11, Scheme 5. 1H NMR (300 MHz,
CDCl3) d 5.85–5.72 (m, 1H, CHvCH2), 5.12–5.07
(m, 2H, CHvCH2), 4.12 (m, 1H, CH2CH(OH)CH2), 3.59
(dd, 1H, J¼8.5, 3.1 Hz, iPrCHOH), 2.75–2.53 (m, 3H,
CH(CH3)COCH2), 2.26–2.21 (m, 2H, CH2CHvCH2),
1.64 (m, 1H, (CH3)2CH), 1.07 (d, 3H, J¼7.1 Hz,
CH(OH)CH(CH3)), 0.99 (d, 3H, J¼6.6 Hz, (CH3)2CH),
0.84 (d, 3H, J¼6.8 Hz, (CH3)2CH); 13C NMR (300 MHz,
CDCl3) d 215.5, 134.0, 118.0, 75.9, 67.2, 49.1, 46.9, 41.0,
30.7, 19.0, 18.9, 8.6; IR (thin film) 3419, 3078, 2963, 2933,
2876, 1704, 1642, 1460, 1384; HRMS (FABþ, NaI) calcd
for C12H23O3 215.1647, found 215.1655.
1
6.5.14. Entry 6, Table 2. H NMR (300 MHz, CDCl3) d
5.71–5.54 (m, 2H, CHvCH2 and CH2CHvCH), 5.42 (dd,
1H, J¼15.6, 6.7 Hz, CHvCHCHOAc), 5.18 (apparent q,
1H, J¼6.6 Hz, CHvCHCHOAc), 5.03 (m, 2H, CHvCH2),
4.71 (dd, 1H, J¼7.7, 2.8 Hz, (CH3)3CCHOAc), 2.33–2.23
(m, 3H, three of CH2CHvCH and CH2CHvCH2),
2.14–2.03 (m, 1H one of CH2CHvCH and CH2CHvCH2),
1.97 (s, 3H, one of CH3CO2CH), 1.96 (s, 3H, one of
CH3CO2CH), 0.85 (s, 9H, (CH3)3C); 13C NMR (300 MHz,
CDCl3) d 170.5, 170.0, 133.2, 130.4, 130.3, 117.6, 79.1,
73.1, 38.8, 34.3, 32.9, 25.8, 21.1, 20.8; IR (thin film) 3079,
2965, 1740, 1433, 1372, 1242 cm21; HRMS (FABþ, NaI)
calcd for C16H26O4Na 305.1729, found 305.1745.
6.5.19. Ketodiol 13, Scheme 6. 1H NMR (300 MHz,
CDCl3) d 5.80–5.68 (m, 1H, CHvCH2), 5.07–5.01 (m,
2H, CHvCH2), 3.97–3.93 (m, 1H, CHOH), 3.84–3.78 (m,
1H, CHOH), 3.03 (br s, 2H, two OH), 2.62–2.47 (m, 4H,
CH2COCH2), 2.24–2.17 (m, 1H, CHCH3), 1.67–1.58 (m,
1H, (CH3)2CH), 1.02–1.00 (d, 3H, J¼6.8 Hz, CHCH3),
0.89 (d, 3H, J¼6.7 Hz, one of (CH3)2CH), 0.87 (d, 3H,
J¼6.7 Hz, one of (CH3)2CH); 13C NMR (75 MHz, CDCl3) d
212.5, 139.5, 115.9, 72.0, 70.5, 47.3, 47.2, 43.3, 33.1, 18.2,
17.6, 15.6; IR (thin film) 3428, 3077, 2964, 2932, 2876,
1706, 1640, 1465, 1385, 1334, 1170, 1170, 1127,
1002 cm21; HRMS (FAB, MþH) calcd for C12H23O3:
215.1647, found 215.1667.
1
6.5.15. Entry 7, Table 2. H NMR (300 MHz, CDCl3) d
5.75–5.56 (m, 2H, CHvCH2 and CH2CHvCH), 5.45 (m,
1H, CHvCHCHOAc), 5.25 (apparent q, 1H, J¼6.3 Hz,
CHvCHCHOAc), 5.10–5.04 (m, 2H, CHvCH2), 4.69 (dd,
1H, J¼7.1, 5.4 Hz, (CH3)2CHCHOAc), 2.46 (m, 1H,
CHOAcCH(CH3)), 2.40–2.33 (m, 2H, CH2CHvCH2),
2.05 (s, 3H, one of CH3CO2CH), 2.03 (s, 3H, one of
CH3CO2CH), 1.85 (m, 1H, (CH3)2CH), 0.95 (d, 3H, J¼
6.8 Hz, CH(CH3)CHvCH), 0.85 (apparent t, 6H, J¼6.6 Hz,
(CH3)2CH); 13C NMR (300 MHz, CDCl3) d 170.9, 170.2,
135.6, 133.2, 129.2, 117.8, 80.3, 73.4, 39.0, 38.2, 29.5, 21.2,
20.9, 19.6, 16.5, 15.2; IR (thin film) 2968, 2936, 2877, 1738,
1462, 1372, 1240 cm21; HRMS (FABþ, NaI) calcd for
C16H26O4Na 305.1729, found 305.1741.
6.5.20. Ketone 15, Scheme 6. 1H NMR (300 MHz, CDCl3)
d 5.67–5.42 (m, 2H, CHvCH2 and CHvCH), 5.40–5.35
(m, 1H, CHvCH), 5.10–4.98 (m, 3H, CHOAc and
CHvCH2), 4.69–4.65 (m, 1H, CHOAc), 2.44–2.35 (m,
2H, CH(CH3)CHvCH and CH(CH3)CHvCH2), 2.02 (s,
3H, one of CH3CO2CH), 2.00 (s, 3H, one of CH3CO2CH),
1.84–1.82 (m, 1H, (CH3)2CH), 0.95 (d, 3H, J¼6.3 Hz,
CHCH3), 0.93 (d, 3H, J¼6.3 Hz, CHCH3), 0.84 (d, 3H,
J¼6.8 Hz, one of (CH3)2CH), 0.81 (d, 3H, J¼6.8 Hz, one of
(CH3)2CH); 13C NMR (75 MHz, CDCl3) d 170.8, 170.1,
139.5, 136.4, 126.7, 115.4, 80.3, 77.2, 42.0, 38.3, 29.5, 21.1,
20.8, 19.6, 16.4, 15.7, 15.3; IR (thin film) 3080, 2969, 2936,
1
6.5.16. Entry 8, Table 2. H NMR (300 MHz, CDCl3) d
5.73–5.54 (m, 2H, CHvCH2 and CH2CHvCH), 5.41
(dd, 1H, J¼15.5, 6.8 Hz, CHvCHCHOAc), 5.27–5.20
(apparent q, 1H, J¼6.6 Hz, CHvCHCHOAc), 5.08–5.02

M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
8899
2877, 1736, 1643, 1460, 1372, 1240, 1184, 1104, 1020 cm21
HRMS (FAB, NaI) calcd for C17H28O4Na: 319.1885, found
319.1903.
;
Chem. 1992, 428, 315–333. (g) Barton, D. H. R.; Kelly, M. J.
Tetrahedron Lett. 1992, 33, 5041–5044. (h) Burn, M. J.;
Bergman, R. G. J. Organomet. Chem. 1994, 472, 43–54.
(i) Gregg, B. T.; Cutler, A. R. Organometallics 1994, 13,
1039–1043. (j) Lorenz, C.; Schubert, U. Chem. Ber. 1995,
128, 1267–1269. (k) Chang, S.; Scharrer, E.; Brookhart, M.
J. Mol. Catal. A 1998, 130, 107–119. (l) Ito, H.; Ishizuka, T.;
Okumura, T.; Yamanaka, H.; Tateiwa, J.-I.; Sonada, M.;
Hosomi, A. J. Organomet. Chem. 1999, 574, 102–106.
(m) Blackwell, J. M.; Foster, K. L.; Beck, V. H.; Piers, W. E.
J. Org. Chem. 1999, 64, 4887–4892. (n) Le Bideau, F.;
1
6.5.21. Diol 17, Scheme 7. H NMR (300 MHz, CDCl3) d
5.81–5.62 (m, 2H, CHvCH and CHvCH2), 5.53–5.46 (m,
1H, CHvCH), 5.11–5.00 (m, 2H, CHvCH2), 3.96–3.92
(apparent t, 1H, J¼6.2 Hz, CH(OH)), 2.51 (s, 1H, one OH),
2.38–2.22 (m, 1H, CHCH3), 2.17–2.00 (m, 3H, CH2 and
one OH), 1.17 (s, 6H, (CH3)2C), 1.01–0.99 (d, 3H, J¼
6.9 Hz, CHCH3); 13C NMR (75 MHz, CDCl3) d 140.2,
134.7, 127.8, 115.4, 75.8, 70.5, 46.4, 43.7, 29.2, 28.8, 15.1;
IR (thin film) 3369, 3080, 2972, 2930, 1666, 1640, 1460,
1415, 1379, 1214, 1153, 1007 cm21; LRMS (FABþ) calcd
for C11H21O2 185.29, found 185.96.
´
Coradin, T.; Henique, J.; Samuel, E. Chem. Commun. 2001,
1408–1409. (o) Chung, M. K.; Ferguson, G.; Robertson, V.
Can. J. Chem. 2001, 79, 949–957.
10. For a seminal report on the monoalcoholysis of a dihydro-
silane, see: Corriu, R. J. P.; Moreau, J. J. E. J. Organomet.
Chem. 1976, 120, 337–346.
11. Indeed, tandem alcoholysis–hydrosilylation using dihydro-
silanes and allylic and homoallylic alcohols has been
described. See: (a) Corriu, R. J. P.; Moreau, J. J. E.
J. Organomet. Chem. 1976, 114, 135–144. (b) Xin, S.;
Harrod, J. F. J. Organomet. Chem. 1995, 499, 181–191.
(c) Wang, X.; Ellis, W. W.; Bosnich, B. Chem. Commun. 1996,
2561–2562.
References
1. Rychnovsky, S. D. Chem. Rev. 1995, 95, 2012–2040.
2. (a) Leighton, J. L.; O’Neil, D. N. J. Am. Chem. Soc. 1997, 119,
11118–11119. (b) Leighton, J. L.; Chapman, E. J. Am. Chem.
Soc. 1997, 119, 12416–12417. (c) Sarraf, S. T.; Leighton, J. L.
Tetrahedron Lett. 1998, 39, 6423–6426. (d) Sarraf, S. T.;
Leighton, J. L. Org. Lett. 2000, 2, 403–405. (e) Zacuto, M. J.;
Leighton, J. L. J. Am. Chem. Soc. 2000, 122, 8587–8588.
(f) Dreher, S. D.; Hornberger, K. R.; Sarraf, S. T.; Leighton,
J. L. Org. Lett. 2000, 2, 3197–3199. (g) Sarraf, S. T.;
Leighton, J. L. Org. Lett. 2000, 2, 3205–3208. (h) O’Malley,
S. J.; Leighton, J. L. Angew. Chem. Int. Ed. 2001, 40,
2915–2917. (i) Zacuto, M. J.; O’Malley, S. J.; Leighton, J. L.
J. Am. Chem. Soc. 2002, 124, 7890–7891. (j) Wang, X.;
Meng, Q.; Nation, A. J.; Leighton, J. L. J. Am. Chem. Soc.
2002, 124, 10672–10673.
12. All chiral alcohol starting materials used in this study were
racemic.
13. The diacetate was prepared to facilitate analysis of the
diastereoselectivity by gas chromatography. The diol may be
isolated.
14. (a) Furuya, N.; Sukawa, T. J. Organomet. Chem. 1975, 96,
C1–C3. (b) Kira, M.; Hino, T.; Sakurai, H. Tetrahedron Lett.
1989, 30, 1099–1102. (c) Iseki, K.; Kuroki, Y.; Takahashi,
M.; Kishimoto, S.; Kobayashi, Y. Tetrahedron 1997, 53,
3513–3526.
15. Reetz, M. T.; Jung, A.; Bolm, C. Tetrahedron 1988, 44,
3889–3898.
3. For reviews of other approaches, see: Ref. 1, and (a) Oishi, T.;
Nakata, T. Synthesis 1990, 635–645. (b) Schneider, C. Angew.
Chem., Int. Ed. Engl. 1998, 37, 1375–1378.
16. Uyeo, S.; Itani, H. Tetrahedron Lett. 1991, 32, 2143–2146.
17. Linderman, R. J.; Chen, K. J. Org. Chem. 1996, 61,
2441–2453.
4. Yamamoto, Y.; Asao, N. Chem. Rev. 1993, 93, 2207–2293.
5. For a review on the Tamao and Fleming oxidations, see: Jones,
G. R.; Landais, Y. Tetrahedron 1996, 52, 7599–7662.
6. The alcohol is treated with HSiCl3, followed by 2 equiv. of
allylMgBr. This reaction provides the desired diallylsilyl
ethers in 50–90% yields, with more sterically hindered
alcohols giving better results.
18. Matsumoto, K.; Oshima, K.; Utimoto, K. J. Org. Chem. 1994,
59, 7152–7155.
19. (a) Myers, A. G.; Kephart, S. E.; Chen, H. J. Am. Chem. Soc.
1992, 114, 7922–7923. For another example of an uncata-
lyzed silicon-mediated aldol reaction, see: (b) Myers, A. G.;
Widdowson, K. L. J. Am. Chem. Soc. 1990, 112, 9672–9674.
(c) Myers, A. G.; Widdowson, K. L.; Kukkola, P. J. Am. Chem.
Soc. 1992, 114, 2765–2767.
7. Stereochemical determinations were made using the method
of Rychnovsky: Rychnovsky, S. D.; Rogers, B.; Yang, G.
J. Org. Chem. 1993, 58, 3511–3515.
20. (a) Denmark, S. E.; Griedel, B. D.; Coe, D. M. J. Org. Chem.
1993, 58, 988–990. (b) Denmark, S. E.; Griedel, B. D.; Coe,
D. M.; Schnute, M. E. J. Am. Chem. Soc. 1994, 116,
7026–7043.
8. This procedure was adapted from the reported method for the
synthesis of (Z)-crotyltrichlorosilane: (a) Tsuji, J.; Hara, M.;
Ohno, K. Tetrahedron 1974, 30, 2143–2146. (b) Kira, M.;
Hino, T.; Sakurai, H. Tetrahedron Lett. 1989, 30, 1099–1102.
(c) Iseki, K.; Kuroki, Y.; Takahashi, M.; Kishimoto, S.;
Kobayashi, Y. Tetrahedron 1997, 53, 3513–3526.
21. Denmark, S. E.; Jacobs, R. T.; Dai-Ho, G.; Wilson, S.
Organometallics 1990, 9, 3015–3019.
22. Several relevant X-ray structures have been reported. See, for
example: Shaw, J. T.; Woerpel, K. A. J. Org. Chem. 1997, 62,
6706–6707.
9. The silane alcoholysis literature prior to 1985 has been
reviewed. See: (a) Lukevics, E.; Dzintara, M. J. Organomet.
Chem. 1985, 295, 265–315. See also: (b) Caseri, W.; Pregosin,
P. S. Organometallics 1988, 7, 1373–1380. (c) Yamamoto,
K.; Takemae, M. Bull. Chem. Soc. Jpn 1989, 62, 2111–2113.
(d) Luo, X.-L.; Crabtree, R. H. J. Am. Chem. Soc. 1989, 111,
2527–2535. (e) Doyle, M. P.; High, K. G.; Bagheri, V.;
Pieters, R. J.; Lewis, P. J.; Pearson, M. M. J. Org. Chem. 1990,
55, 6082–6086. (f) Bedard, T. C.; Corey, J. Y. J. Organomet.
23. Hoveyda has reported relevant observations regarding the
Lewis acidity of oxasilacyclopentanes of this type. See:
Young, D. G. J.; Hale, M. R.; Hoveyda, A. H. Tetrahedron
Lett. 1996, 37, 827–830.
24. These are the conditions employed by Utimoto. See Ref. 18.
25. Myers has elegantly established a mechanism involving

8900
M. J. Zacuto et al. / Tetrahedron 59 (2003) 8889–8900
pseudorotation for the uncatalyzed aldol reactions of O-silyl
ketene N,O-acetals. See Ref. 19c.
Chem. Lett. 1995, 281–282. (c) Chemler, S. R.; Roush, W. R.
J. Org. Chem. 1998, 63, 3800–3801. (d) Chemler, S. R.;
Roush, W. R. Tetrahedron Lett. 1999, 40, 4643–4647.
(e) Chemler, S. R.; Roush, W. R. J. Org. Chem. 2003, 68,
1319–1333.
26. (a) Stevenson, W. H., III.; Wilson, S.; Martin, J. C.; Farnham,
W. B. J. Am. Chem. Soc. 1985, 107, 6340–6352.
(b) Stevenson, W. H., III; Martin, J. C. J. Am. Chem. Soc.
1985, 107, 6352–6358.
29. Dreher, S. D.; Leighton, J. L. J. Am. Chem. Soc. 2001, 123,
341–342.
27. Schmidt, D. R.; O’Malley, S. J.; Leighton, J. L. J. Am. Chem.
Soc. 2003, 125, 1190–1191.
30. (a) Kinnaird, J. W. A.; Ng, P. Y.; Kubota, K.; Wang, X.;
Leighton, J. L. J. Am. Chem. Soc. 2002, 124, 7920–7921.
(b) Wang, X.; Meng, Q.; Nation, A. J.; Leighton, J. L. J. Am.
Chem. Soc. 2002, 124, 10672–10673. (c) Kubota, K.;
Leighton, J. L. Angew. Chem. Int. Ed. 2003, 42, 946–948.
28. Kira and Sakurai, and Roush and Chemler have made similar
observations during studies of intramolecular crotylsilylations
of b-hydroxy ketones and aldehydes. See: (a) Sato, K.; Kira,
M.; Sakurai, H. J. Am. Chem. Soc. 1989, 111, 6429–6431.
(b) Kira, M.; Sato, K.; Kazushi, S.; Gewald, R.; Sakurai, H.