Unexpected ambidoselectivity in crossed-aldol reactions of α-oxy aldehyde trichlorosilyl enolates

Article abstract of DOI:10.1016/j.tet.2007.04.009

The ambido-, stereo- and enantioselectivity of the phosphoramide-promoted aldol reactions of α-oxy aldehyde trichlorosilyl enolates with benzaldehyde has been investigated. Analysis of the products from α-tert-butyldimethylsilyloxy α-deuterioacetaldehyde trichlorosilyl enolate confirmed that this 1,2-bis-silyloxyethene derivative reacted as a tert-butyldimethylsilyl enolate rather than trichlorosilyl enolate in the aldol reaction with very high ambidoselectivity. The phosphoramide-coordinated trichlorosilyl group acted as an organizing center for the aldol reaction. From the aldol process, excellent anti-diastereoselectivity could be achieved. The enantioselectivity remained moderate to low for both anti- and syn-diastereomer with a wide range of phosphoramide catalysts. α-Triisopropylsilyloxy, phenoxy and benzyloxy acetaldehyde trichlorosilyl enolates also reacted in a similar fashion with benzaldehyde to give aldol products with varying degrees of selectivities.

Full text of DOI:10.1016/j.tet.2007.04.009

Tetrahedron 63 (2007) 8636–8644
Unexpected ambidoselectivity in crossed-aldol reactions
of a-oxy aldehyde trichlorosilyl enolates
*
Scott E. Denmark and Sunil K. Ghosh
Roger Adams Laboratory, Department of Chemistry, University of Illinois, 600 South Mathews Avenue, Urbana, IL 61801, USA
Received 31 January 2007; revised 3 April 2007; accepted 4 April 2007
Available online 12 April 2007
Abstract—The ambido-, stereo- and enantioselectivity of the phosphoramide-promoted aldol reactions of a-oxy aldehyde trichlorosilyl
enolates with benzaldehyde has been investigated. Analysis of the products from a-tert-butyldimethylsilyloxy a-deuterioacetaldehyde
trichlorosilyl enolate confirmed that this 1,2-bis-silyloxyethene derivative reacted as a tert-butyldimethylsilyl enolate rather than trichlorosilyl
enolate in the aldol reaction with very high ambidoselectivity. The phosphoramide-coordinated trichlorosilyl group acted as an organizing
center for the aldol reaction. From the aldol process, excellent anti-diastereoselectivity could be achieved. The enantioselectivity remained
moderate to low for both anti- and syn-diastereomer with a wide range of phosphoramide catalysts. a-Triisopropylsilyloxy, phenoxy and
benzyloxy acetaldehyde trichlorosilyl enolates also reacted in a similar fashion with benzaldehyde to give aldol products with varying degrees
of selectivities.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
crossed-aldol reaction of aldehyde TCS enolates is also de-
pendent on the enolate geometry. Thus the Z-enolate (Z)-1 af-
fords aldol product syn-2, whereas E-enolate (E)-1 affords
predominantly anti-2 (Scheme 1). Because of the instability
of the crossed aldehyde aldol products (3-hydroxy alde-
hydes) toward dehydration and oligomer formation, the
products are isolated as dimethyl acetals.
Lewis-base-catalyzed, enantioselective, aldol reactions are
now well-established as viable methods for the stereocon-
trolled construction of carbon–carbon bonds from a variety
of carbonyl compounds and enolate derivatives.1 Trichloro-
silyl (TCS) enolates derived from both ketones2 and esters3
react smoothly with aldehydes to provide aldol products in
high yield with good to excellent diastereo- and enantio-
selectivity. The diastereoselectivity of all these aldolization
processes is a near perfect reflection of the geometrical
composition of the TCS enolates.4 Thus both the syn- and
anti-aldol products can be accessed by the use of Z- and
E-TCS enolates, respectively.5
Me
Ph
N
O
N
P
N
Ph
Me
(0.05 equiv)
OH OMe
OMe
n-C5H11
syn-2
OSiCl3
3
Ph
Ph
n-C5H11
1. PHCHO / CH2Cl2 / –78 °C
2. MeOH
(Z)-1 (Z/E 99:1)
Although the crossed-aldol reactions between aldehydes are
known for their inherent problems of polycondensation and
other side reactions,6 we have recently reported the Lewis-
base-catalyzed aldol reaction of aldehyde TCS enolates
with aldehydes.7 Despite the relatively small size of the
aldehyde TCS enolate, this directed, crossed-aldol reaction
also affords products in excellent yields and diastereoselec-
tivity albeit with variable enantioselectivity. As is the case
with ketone TCS enolates, the diastereoselectivity in the
syn/anti 99:1
OH OMe
as above
OMe
n-C5H11
n-C5H11
(E)-1 (E/Z 30:1)
OSiCl3
anti-2
anti/syn 49:1
Scheme 1. Diastereoselective cross-aldol reaction of enoxytrichlorosilanes.
Application of the aldol reaction to the synthesis of carbo-
hydrates and polyhydroxylated carbon frameworks is well-
documented but requires the help of protecting group
manipulations and oxidation-state adjustments.8 A more
direct polyhydroxyl carbon skeleton synthesis can be envi-
sioned with the use of an iterative aldol sequence using
Keywords: Aldol; Enolate; Deuterium labeling; Enantioselective; Diastereo-
selective.
* Corresponding author at present address: 245 Roger Adams Laboratory,
Box 18-5, Department of Chemistry, University of Illinois, 600 South
Mathews Avenue, Urbana, IL 61801, USA. Tel.: +1 (217) 333 0066;
fax: +1 (217) 333 3984; e-mail: denmark@scs.uiuc.edu
0040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.04.009

S. E. Denmark, S. K. Ghosh / Tetrahedron 63 (2007) 8636–8644
8637
simple a-oxy aldehydes. Enolates of a-alkoxy ketones or
enol ethers/esters have been used in aldol reactions for the
synthesis of polyhydroxylated carbonyl compounds with
varying levels of stereocontrol.9 A few reports are also avail-
able for the crossed-aldol reaction of a-oxy aldehyde with
impressive success although the inherent problem of self-
aldolization remains and diastereoselectivity cannot be
controlled by substrate structure.10 The practical execution
of this strategy would require the implementation of a new
crossed-aldol technology: a selective aldol reaction of
a-oxy aldehyde substrate with another aldehyde with no
self-aldolization or polycondensation. On the basis of our
previous studies on crossed-aldol reaction of simple alde-
hyde trichlorosilyl enolates, we envisaged that a crossed
aldehyde aldol reaction of TCS enolates derived from a-
oxy aldehydes might offer a solution to this problem. Unlike
simple aldehydes, a-oxy aldehyde TCS enolate 4 has two
chemically reactive centers and can thus react via two dis-
tinct pathways as shown in Scheme 2. If the trichlorosilyloxy
group controls the aldol addition (normal mode), product 5 is
expected to form, whereas oxy-group-controlled aldol reac-
tion (reverse mode) would produce product 6. A previous
study from these laboratories showed that trialkylsilyl enol
ethers of acetaldehyde could engage in aldol additions to
aromatic aldehydes under the catalytic action of silicon
tetrachloride and a chiral bisphosphoramide.11 Thus, it was
uncertain if the activated trichlorosilyloxy group would
dominate over the (silyl)oxy group in controlling the course
of the aldol addition. We describe herein the synthesis and
the ambidoselective12 crossed-aldol reactions of a-oxy acet-
aldehyde TCS enolates.
ether with silicon tetrachloride has been described in
detail.2c,5,7c In these studies, the geometry of the resultant
TCS enolate could be controlled by appropriate selection
of the silylation method, but was largely dependent on the
structure of the starting ketone. Gratifyingly, the palladium
acetate catalyzed trans-silylation of either (E)- or (Z)-tri-
methylsilyl enol ethers of heptaldehyde with silicon tetra-
chloride afforded an 85/15 E/Z mixture of geometrical
isomers, the resulting TCS enolate. This outcome was
deemed sufficient to explore the aldol addition chemistry
of these species.
The tert-butyldimethylsilyl (TBS) group was initially se-
lected as the oxygen substituent because previous studies2
have shown the compatibility of the TBSO group with
TCS enolates and other trichlorosiliconium species.13 The
enol trimethylsilyl ether 7a of a-tert-butyldimethylsilyloxy-
acetaldehyde 8a was prepared using trimethylsilyl chloride
and triethylamine in hot acetonitrile (Scheme 3).14 The
1,2-disilyloxyethene 7a was isolated in 57% yield and
with a high geometrical purity (Z/E¼95/5). For the prepara-
tion of the TCS enolate 4a, the metathesis of enol trimethyl-
silyl ether 7a with silicon tetrachloride in the presence of
many metal salts under various reaction conditions was ex-
amined. A few representative examples are shown in Table 1.
The exchange reaction was extremely slow with Hg(OAc)2
and no reaction took place with Tl(OAc)3, but both
Pd(OAc)2 and Pd(TFA)2 were effective catalysts (entries
3–7). Interestingly, lowering the catalyst loading led to faster
reactions and also improved the yield of the distilled prod-
uct. It was found that if the palladium loading is too high,
the TCS enolate decomposes during distillation, producing
a significant amount of TBSCl as a by-product. Keeping
the reaction concentration low was also crucial to obtain
a good yield. The optimal conditions employed 2.0 equiv
of silicon tetrachloride and 0.034 equiv of Pd(OAc)2 at
0.5 M concentration in dichloromethane at room tempera-
ture (entry 6), which provided the TCS enolate 4a in 66%
yield. The geometrical purity of the enolate 4a was deter-
mined by the integration of the distinguishable peaks for the
olefinic protons and was found to be very high (Z/E¼95/5).
The diagnostic 13C NMR signals for the Z- and E-enolate
were also discernable. It is also worth mentioning that the
geometrical composition of the TCS enolate was indepen-
dent of the geometrical composition of the starting enol
trimethylsilyl ether used. This observation obviously re-
stricts this method to the production of E-TCS enolates.
OSiCl3
O
OH
O
"normal mode"
+
PO
controlled by OSiCl3 group
R
H
R
R
H
H
4
OP
5
O
OH
O
OP
"reverse mode"
controlled by PO group
+
Cl3SiO
R
H
4
OH
6
Scheme 2. Ambidoselectivity modes for aldolization of a-oxy trichlorosilyl
enolates.
2. Results
2.1. Synthesis of a-oxy acetaldehyde TCS enolates
2.2. Crossed-aldol reaction of a-tert-butyldimethyl-
silyloxyacetaldehyde TCS enolate 4a
To test the feasibility of this concept, the generation of stereo-
defined TCS enolates of a-oxy acetaldehyde was required.
Over the past decade, a number of synthetic methods have
been developed in these laboratories for the generation of
TCS enolates of ketones, esters and aldehydes. Amongst
them, the metal-catalyzed metathesis of a trialkylsilyl enol
2.2.1. Background reaction at L65 ꢀC. To assess the effi-
ciency of Lewis bases in promoting the aldol reactions, it
was necessary to perform a control reaction without pro-
moters. Thus, TCS enolate 4a and benzaldehyde were com-
bined in equimolar proportions at ꢁ65 ꢀC in chloroform.
OSiCl3
Me3SiO
TBSO
Me3SiCl, Et3N
O
SiCl4, Pd(OAc)2
TBSO
4a
TBSO
8a
MeCN, 80 °C
57%
CH2Cl2, rt
66%
H
7a
(Z/E = 95/5)
(Z/E = 95/5)
Scheme 3. Preparation of a-oxy trichlorosilyl enolates.

8638
S. E. Denmark, S. K. Ghosh / Tetrahedron 63 (2007) 8636–8644
Table 1. Survey of catalysts, loadings and reaction concentration for the
trans-silylation of 7a to 4a
of chiral phosphoramide 3 in dichloromethane at dry ice/
acetone temperature. The reaction was quenched with meth-
anol at low temperature, allowing the mixture to come to
room temperature and then pouring mixture into cold satu-
rated, aqueous sodium bicarbonate solution. The dihydroxy
dimethyl acetal products anti-9 and syn-9 (anti/syn¼3/1)
were isolated as an inseparable mixture in good yield
(Scheme 5). The relative configuration of the hydroxyl-
bearing centers as well as the absolute configuration of
anti-9 and syn-9 diastereomers was unambiguously estab-
lished by converting the acetals to known di-4-tolylthioace-
tals.15 To accomplish this, the mixture of dimethyl acetals 9
was treated with p-thiocresol in the presence of BF3$OEt2,
which provided an easily separable mixture of dithioacetals
anti-17 (w60/40 er) and syn-17 (w55/45 er). The absolute
configuration of the major enantiomer anti-17 was found
to be (2S,3S) while that for syn-17 was (2S,3R).15
OSiCl3
OSiMe3
SiCl4, MXn
CH2Cl2, rt
TBSO
4a
TBSO
7a (Z/E 95/5)
Entry MXn (equiv)
Time Concentrationa Conversionb Yieldc
(h)
(M)
(%)
(%)
1
2
3
4
5
6
7
Hg(OAc)2 (0.02) 72
Tl(OAc)3 (0.05) 48
1
1
1
1
0.3
0.5
1
5
NDd
No reaction ND
Pd(OAc)2 (0.2)
3
>95
>95
>95
100
16
24
19
66
38
Pd(OAc)2 (0.15) 16
Pd(OAc)2 (0.05) 12
Pd(OAc)2 (0.034) 12
Pd(OAc)2 (0.002) 38
>95
a
b
c
d
Refers to concentration with respect to 7a.
Monitored by 1H NMR.
Refers to isolated yield of distilled product.
ND¼not determined.
2.2.3. Ambidoselectivity of the aldol addition of 4a. The
formation of aldol products anti-9 and syn-9 (from quenching
with methanol) obfuscates the pathway by which 4a reacted,
i.e., as a TBS enol ether or as a trichlorosilyl enolate. To re-
solve this selectivity issue, the deuterium labeled enolate
4a-d1 was prepared. If 4a-d1 reacts as TBS enol ether in the
aldolization with benzaldehyde, then the deuterium will re-
side at the acetal carbon of the dihydroxy dimethyl acetal
18. On the other hand, if 4a-d1 reacts as the trichlorosilyl eno-
late, then the deuterium will be positioned at the hydroxyl
bearing carbon a-to dimethyl acetal group in 19 (Scheme 6).
Quenching the reaction with methanol after 4 h afforded the
aldol products in 12% yield along with 69% of benzaldehyde
dimethyl acetal 10 (Scheme 4). Thus, the uncatalyzed aldol
addition is a very slow process and it was appropriate to eval-
uate the ability of chiral Lewis bases to catalyze the addition.
OH OMe
Ph
+
OMe
O
OH
OSiCl3
1. CHCl3, -65 °C, 4 h
2. MeOH
+
9 (12%)
TBSO
Ph
H
Labeled TCS enolate 4a-d1 was prepared as shown in
Scheme 7. Dimethyl L-tartrate was converted to acetonide
2016 that was reduced with lithium aluminum tetradeuteride
(isotopic content >99%-d) to give the tetradeuterated diol
21. The acetonide group was removed and the primary
hydroxyl groups of the resulting tetrol 22 were selectively
protected as the TBS ethers to give 23. Sodium periodate
cleavage gave the a,a-dideuterated a-tert-butyldimethyl-
silyloxyacetaldehyde 8a-d2 in very good overall yield. The
enol trimethylsilyl ether 7a-d1 was prepared as before using
trimethylsilyl chloride and triethylamine in hot acetonitrile.
The 1,2-disilyloxyethane 7a-d1 was isolated in 64% yield
with a high geometrical purity (Z/E¼94/6). At this stage,
the isotopic composition of the enol TMS ether 7a-d1 was
found to be 99.2% enriched by mass spectrometric analysis.
Metathesis of enol TMS ether 7a-d1 with 2.0 equiv of sili-
con tetrachloride and 0.007 equiv of Pd(OAc)2 at 0.5 M in
dichloromethane at room temperature provided enolate
4a-d1 in 73% yield. As expected, enolate 4a-d1 was highly
4a (Z/E 95/5)
PhCH(OMe)2
10 (69%)
Scheme 4. Uncatalyzed, background aldol reaction of enolate 4a.
2.2.2. Crossed-aldol reaction of 4a in the presence of
phosphoramide promoters. As is the case with nearly all
other enoxytrichlorosilanes, the Lewis-base-promoted aldol
addition of enolate 4a was very fast. Using 20 mol % of
chiral phosphoramide 3 (Chart 1), the reaction between 4a
and benzaldehyde in CDCl3 was complete in less than an
hour at ꢁ70 ꢀC as judged by the disappearance of the enolate
by NMR analysis. For preparative purpose, the reaction be-
tween 4a and benzaldehyde was carried out using 20 mol %
Me
-Napth
Ph
N
N
N
O
N
O
N
O
1
P
P
P
N
enriched in the Z isomer (Z/E¼92/8) as judged H NMR
N
N
N
analysis.
Me
-Napth
Ph
Crossed-aldol addition of enolate 4a-d1 with benzaldehyde
in the presence of phosphoramide 3 took place smoothly
and a methanolic work-up provided a mixture of a pair of
anti- and a pair of syn-aldol products (anti/syn¼13/1,
Scheme 8). In both the anti- and the syn-manifold, the ratio
of 18/19 was found to be 94/6. Therefore, enolate 4a reacted
predominantly as a nucleophilic silyl enol ether rather than
as a trichlorosilyl enolate. This outcome was rather unex-
pected in light of the powerful catalytic effect of the phos-
phoramide. Clearly, the trichlorosilyl unit is playing an
important role to activate the addition, however, unlike the
3
11
12
Me
Me
N
Me
N
O
N
O
N
O
R
P
P
P
(CH2)5
N
N
N
Me
N
Me
Me
16
Me
2
13: R=Ph
14: R=1-piperidinyl
15
Chart 1. Chiral phosphoramide promoters used for asymmetric aldol
reactions.

S. E. Denmark, S. K. Ghosh / Tetrahedron 63 (2007) 8636–8644
8639
OH OMe
OMe
OH OMe
OMe
1. 3, (20 mol %)
OSiCl3
+
CH2Cl2
+
Ph
TBSO
PhCHO
Ph
–72 °C, 5.5 h
2. MeOH
74%
OH
OH
4a (Z/E = 95/5)
syn-9
anti-9
OH SC6H4(4-Me)
OH SC6H4(4-Me)
2
+
4-thiocresol, BF3.OEt2
77%
2
3
3
Ph
SC6H4(4-Me)
Ph
SC6H4(4-Me)
OH
OH
(2S,3S)-(–)-17
(2S,3R)-(–)-17
Scheme 5. Aldolization of TCS enolate 4a under catalysis with 3.
OH OMe
OMe
O
opportunities to probe the experimental and structural fac-
tors that control the diastereo- and enantioselectivity.
OH OMe
OMe
OSiCl3
cat.
+
or
TBSO
Ph
Ph
H
Ph
D
D
OH
OH
D
4a-d1
2.3. Optimization of the crossed-aldol addition of 4a
syn/anti-19
syn/anti-18
2.3.1. Influence of reaction conditions. The results of the
deuterium-labeling study clearly established that: (1) the
aldol addition of 4a is governed by the nucleophilicity of
the TBS enol ether and (2) the trichlorosilyloxy group plays
a major role in activating the aldehyde toward addition.
However, it was unclear if the TCS group was serving as
an organizational center as has been established in all of
the previous examples of these additions through the
Scheme 6. Aldolization of deuterium labeled TCS enolate 4a-d1.
reactions of simple aldehyde, ketone and ester TCS enolates,
in this case, it is the carbon bearing the trichlorosilyloxy
group that serves as the nucleophile. Thus, this transforma-
tion represents an unprecedented pathway for bond con-
struction with TCS enolates and therefore presents new
MeO2C
CO2Me
HOD2C
CD2OH
(MeO)2CMe2
LiAlD4 / Et2O
75%
dimethyl
(L)-tartarate
TsOH
O
O
O
O
84%
Me Me
Me Me
21
20
TBSCl
Amberlyst 15
MeOH / H2O
HOD2C
CD2OH
OH
CD2OTBS
OH
23
TBSOD2C
HO
imidazole / DMF
HO
55%
97%
22
NaIO4
TBSO
OSiMe3
TMSCl / Et3N
64%
D
D
THF / H2O / CH2Cl2
CHO
8a-d2
TBSO
D
77%
7a-d1 Z/E = 96:4
99.2% D
Pd(OAc)2
TBSO
D
OSiCl3
SiCl4 / CH2Cl2
73%
4a-d1 Z/E = 92:8
Scheme 7. Synthesis of D-labeled TCS enolate 4a-d1.
OH OMe
OMe
OH
anti-19
OH OMe
OMe
OH
anti-18
+
13
Ph
Ph
Ph
Ph
D
D
1. 3 (10 mol%)
O
OSiCl3
CHCl3/ –65 °C
+
96
:
4
TBSO
5.5 h
Ph
H
2. MeOH
D
OH OMe
OH OMe
OMe
4a-d1
93%
anti/syn = 13/1
+
OMe
Z/E = 92:8
99.2%-d1
D
D
OH
1
OH
syn-18
syn-19
96
:
4
Scheme 8. Crossed-aldol reaction of 4a-d1 with benzaldehyde.

8640
S. E. Denmark, S. K. Ghosh / Tetrahedron 63 (2007) 8636–8644
correlation of enolate geometry with product diastereoselec-
tivity. Obviously, the phosphoramide also plays a vital role
because the control reactions are extremely slow. Unfortu-
nately, the enantioselectivity in either the syn- or anti-
pathway is poor (Scheme 5). However, a large number of
different phosphoramides have been developed in these
laboratories for aldol reactions of TCS enolates of aldehyde,
ketone as well as esters with very high diastereo- and enan-
tioselectivity.1 It should also not be lost upon the reader, that
even if these are not TCS-organized transition structures, but
rather involve chiral siliconium ion activation of the alde-
hydes, excellent selectivities have been observed for addi-
tion of enoxysilane nucleophiles to aldehydes (with silicon
tetrachloride).13,17 Chart 1 contains various phosphoramides
employed in the present study for aldol reaction of enolate
4a and benzaldehyde. Optimization of the aldol addition
of 4a with benzaldehyde also involved a survey of the reac-
tion solvent and catalyst loading. For an appreciable reaction
rate, the reaction concentrations for all the studies were
maintained at 0.25 M.
bisphosphoramide 16 showed some bias for the syn-aldol
process although the major product was still anti-9. With
this phosphoramide, the enantioselectivity for the syn-aldol
product is also better than the monophosphoramides studied
so far.
2.4. Effect of substituent at the oxy group on the
selectivity of aldol reaction of a-oxy acetaldehyde
TCS enolates
Although the rate and diastereoselectivity of the aldol addi-
tion of 4a could be improved by modification of the catalyst
and reaction conditions, the enantioselectivity remained un-
acceptably low. Enantiomeric purity of the products could
not be improved and remained low to moderate with a large
number of catalysts surveyed. It would be interesting to
know the effect of the substituent at the oxy groups of a-
oxy aldehyde. Therefore, three substituents, triisopropylsilyl
(TIPS), phenyl (Ph) and benzyl (Bn), at the oxy group were
chosen. These groups have different steric or electronic
properties compared to the TBS group in 8a. The TCS eno-
lates 4b–4d of these a-oxy acetaldehydes 8b–8d were pre-
pared via their enol trimethylsilyl ethers 7b–7d as shown
in Scheme 9. The metathetical exchange of silicon group
of these enol ethers with silicon tetrachloride in the presence
of palladium catalyst provided the TCS enolates that were
not isolated but instead used in situ for aldol reaction. We
were gratified that all the TCS enolates reacted with benzal-
dehyde in the presence of phosphoramide catalysts to give
aldol products with good to excellent yields and varying de-
gree of diastereo- and enantioselectivity. The results of aldol
reaction of these enolates with benzaldehyde in the presence
of phosphoramide catalysts are compiled in Table 4.
To investigate solvent and catalyst loading, chiral phosphor-
amide 3 was chosen for study. It has already been shown that
3 was an effective catalyst for the aldol reaction and pro-
duced anti-9 as the major diastereomer. A number of solvents
were surveyed for the aldol reaction and chloroform was
found to be superior for this purpose in terms of rate and yield
(entries 1–3 and 7, Table 2). Interestingly, the diastereo-
selectivity of the aldol addition is highly dependent on the
catalyst loading in chloroform. The anti-selectivity increased
dramatically by reducing catalyst loading from 100 to
5 mol % without effecting the yield, whereas the enantio-
selectivity in either manifolds was unchanged. Accordingly,
the next objective was to survey phosphoramide structures.
The results in Table 4 show that TIPS protection behaved
more or less in the same way as TBS group. The phenoxy
group also behaved in the same way but the reactivity of
this enolate was much lower and the aldol reaction did not
proceed to completion under the standard conditions. The
benzyloxy group behaved in a different, but less satisfactory
way; the diastereo- and enantioselectivity is quite low for
this group.
2.3.2. Influence of catalyst structure. To optimize enantio-
selectivity, a series of monophosphoramides and one bis-
phosphoramide 16 (Chart 1) were examined at 10–20 mol %
loading and 0.25 M in chloroform at ꢁ65 ꢀC. Although
excellent yields and high anti-diastereoselectivities were ob-
served with many different catalyst structures, the enantio-
selectivities did not improve (Table 3). Interestingly, the
Table 2. Effect of solvent and catalyst loading on aldol reaction between 4a and benzaldehyde
OH OMe
OMe
O
OSiCl
OH OMe
3 1. 3 (mol %) / solvent,
+
–65 to –72 °C / 5.5 h
TBSO
+
Ph
Ph
H
Ph
OMe
2. MeOH
OH
4a (Z/E 95/5)
OH
syn-9
anti-9
Entry
Solvent
mol % of 3
Yielda (%)
anti/synb
anti erb (2S,3S)/(2R,3R)
syn erb (2S,3R)/(2R,3S)
1
2
3
4
5
6
7
8
9
10
CH2Cl2
Et2O
20
20
20
100
60
40
20
10
5
74
32c
43c
96
92
92
3/1
1.2/1
2.3/1
2.3/1
4/1
5/1
9/1
10/1
14/1
15/1
60/40
60/40
53/47
55/45
50/50
49/51
48/52
48/52
47/53
49/51
57/43
56/44
61/39
67/33
67/33
66/34
65/35
64/36
60/40
58/42
Toluene
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
93
94
82
2.5
64c
a
Yields of chromatographically homogeneous material.
Determined by CSP-SFC(AS; 150 bar, 3 mL/min, 2.5% MeOH).
Incomplete reaction.
b
c

S. E. Denmark, S. K. Ghosh / Tetrahedron 63 (2007) 8636–8644
Table 3. Effect of catalyst structure on aldol reaction between 4a and benzaldehyde
8641
Entry
Catalyst (mol %)
Yielda (%)
anti/synb
anti erb (2S,3S)/(2R,3R)
syn erb (2S,3R)/(2R,3S)
1
2
3
4
5
6
7
3 (20)
11 (10)
12 (20)
13 (10)
14 (20)
15 (20)
16 (10)
93
82c
74c
89
90
93
83
9/1
13/1
16/1
22/1
24/1
23/1
5/1
48/52
47/53
60/40
48/52
47/53
59/41
48/52
65/35
60/40
51/49
65/35
54/46
31/69
78/22
a
Yields of chromatographically homogeneous material.
Determined by CSP-SFC(AS; 150 bar, 3 mL/min, 2.5% MeOH).
Incomplete reaction.
b
c
OSiCl3
OSiMe3
O
Me3SiCl / Et3N
SiCl4, Pd(OAc)2
RO
RO
RO
MeCN, 80 °C
61-89%
H
CH2Cl2 / rt
"100%"
(Z/E >90/10)
(Z/E >90/10)
R = TIPS, 4b
R = Ph, 4c
R = Bn, 4d
R = TIPS, 7b
R = Ph, 7c
R = Bn, 7d
R = TIPS, 8b
R = Ph, 8c
R = Bn, 8d
Scheme 9. Preparation of a-oxy TCS enolates with various protecting groups.
Table 4. Crossed-aldol reaction of TCS enolates 4b–4d with benzaldehyde
OH OMe
catalyst (5 mol %)
OH OMe
OMe
OSiCl3
CHCl3, -65 °C
+
+
Ph
OMe
RO
PhCHO
Ph
5.5 h, then MeOH
OH
OH
4b-d (Z/E >90/10)
syn-9
anti-9
Yielda (%)
anti/synb
anti erb (2S,3S)/(2R,3R)
syn erb 10 (2S,3R)/(2R,3S)
,
Entry
TCS enolate
Catalyst
1
2
3
4
5
6
7
8
9
4b
4b
4c
4c
4d
4d
4d
4d
4d
3
14
3
14
3
12
13
14
15
66
62
30
38
83
83
83
83
74
6.1/1
15/1
21/1
20/1
1/1.4
2/1
3/1
1.9/1
1/1.1
47/53
43/57
48/52
46/54
57/43
61/39
30/70
31/69
46/54
45/55
38/62
42/58
54/46
35/65
60/40
51/49
50/50
52/48
a
Yields of chromatographically homogeneous material.
Determined by CSP-SFC (Chiralcel AS; 150 bar, 3 mL/min, 2.5% MeOH).
b
3. Discussion
bond forming process involves a five-membered, not the
usual six-membered chair-like transition structure.4,7 This
disparity is clearly illustrated in the divergent behavior be-
tween 4 and 1. In aldolizations of the latter TCS enolate,
the normal correlation of enolate geometry with product
configuration obtains ((Z)-1/syn-2 and (E)-1/anti-2).
However, in the present study, 4 of predominantly Z-geo-
metry gave the anti-diastereomer as the major product with
all catalysts. Moreover, the anti-selectivity increased with
decreasing catalyst loading. Unfortunately, the enantioselec-
tivities are low and variable for all cases. This outcome sug-
gests that there are competitive pathways operating to give
syn- and anti-aldol products.18
The aldol addition of TCS enolate 4a proceeded rather unex-
pectedly through a pathway in which the TBS enol ether
moiety served as the nucleophilic component. The trichloro-
silyl unit must still be playing an activating and perhaps or-
ganizational role, because the reaction rate is enhanced and
the diastereoselectivity is strongly influenced by phosphor-
amides. Without the benefit of detailed kinetic studies with
this nucleophile we cannot distinguish the two limiting
mechanistic scenarios: (1) the aldolization proceeds via
a closed transition structure consisting of one molecule
each of aldehyde and TCS enolate as was established for
all other TCS enolates,4,7 or (2) the aldolization proceeds
via an open transition structure involving two TCS enolates,
one to activate the aldehyde and a second to serve a nucleo-
phile as is seen in the SiCl4$phosphoramide catalyst sy-
tems.13 Even if we assume that the reaction proceeds
through a closed transition structure in which phosphor-
amides are present to activate the addition, it is still very
difficult to predict the stereochemical outcome because the
The increase in anti-diastereoselectivity with decreasing
catalyst loading suggests that the anti-product is generated
from a one-phosphoramide, cationic trigonal bipyramidal
transition structure (A, Fig. 1) and the syn-product is from
a two-phosphoramide octahedral, cationic transition struc-
ture (B, Fig. 1).1b,4 The benzyloxy-substituted TCS enolate
behaved differently and might be enabling another type of

8642
S. E. Denmark, S. K. Ghosh / Tetrahedron 63 (2007) 8636–8644
OP*
H
R
O
O
H
Ph
OP*
Si
OP*
O
Cl
Cl
OP*
H
H
Si
O
Cl
Cl
Cl
Ph
O
Cl
Cl Si
Cl
H
Cl
O
OR
O
H
H
OR
Ph
H
A
B
C
one-phosphoramide and
intramolecular oxy-coordinated
cationic hexacoordinate
transition structure
two-phosphoramide
cationic hexacoordinate
transition structure
one-phosphoramide
cationic pentacoordinate
transition structure
Figure 1. Hypothetically closed transition structures for the crossed aldolization of 4.
closed transition state structure involving coordination of the
alkoxy oxygen to the TCS group in a one-phosphoramide oc-
tahedral, transition state structure (C, Fig. 1). This transition
state can lead to either the anti- or the syn-product, thus mak-
ing the situation more complex. Open transition structures
are still more difficult to analyze because it is not known
where or how many catalyst molecules are bound. Unfortu-
nately, the low er for all catalysts does not permit additional
insights.
mass was poured into saturated NaHCO3 solution and ex-
tracted with pentane. The pentane extract was dried over
MgSO4, then filtered and evaporated. The residual oil was
purified by chromatography (SiO2, pentane/ether, 95/5) fol-
lowed by distillation gave 3.7 g (57%) of the enol ether (Z)-
7a as a clear colorless liquid. The Z/E ratio was found to be
95/5. Bp: 50 ꢀC (0.2 mmHg); 1H NMR (500 MHz, CDCl3):
6.41 (d, J¼10, 1H, (E)-7a), 6.35 (d, J¼10, 1H, (E)-7a), 5.55
(d, J¼3.2, 1H, (Z)-7a), 5.49 (d, J¼3.2, (Z)-7a), 0.93 (s, 9H),
0.18 (s, 9H), 0.13 (s, 6H); 13C NMR (101 MHz, CDCl3):
125.71, 124.45, 25.72, 18.49, ꢁ0.37, ꢁ5.21; IR (neat):
3042 (w), 2958 (s), 2931 (m), 2897 (w), 2888 (w), 2859
(m), 1663 (m), 1472 (w), 1464 (w), 1389 (m), 1292 (w),
1253 (s), 1147 (s), 1077 (s), 899 (s), 843 (s), 808 (m), 782
(m), 752 (w), 676 (w); MS (EI, 70 eV): 246 (M+, 3), 189
(Mꢁt-Bu+, 31), 147 (20), 73 (Me3Si, 100). Analysis,
C11H26O2Si2 (246.50) Calculated: C, 53.60; H, 10.63; Si,
22.79%. Found: C, 53.62; H, 10.54; Si, 22.41%.
4. Conclusion
The (Z)-TCS enolate of a-tert-butyldimethylsilyloxyacet-
aldehyde ((Z)-4a) can be prepared from the corresponding
trimethylsilyl enol ether by palladium-catalyzed metathesis
with silicon tetrachloride. The addition of (Z)-4a to benz-
aldehyde is susceptible to catalysis by phosphoramides
and the diastereoselectivity of the addition is dependent on
the catalyst loading. The ambidoselectivity of the aldol addi-
tion was elucidated by deuterium-labeling experiments that
confirmed that (Z)-4a reacted as a TBS enol ether in the aldol
process, and that the trichlorosilyl group acted as an organiz-
ing center for the reaction. From the aldol addition, excellent
anti-diastereoselectivity could be achieved. Unfortunately,
the enantioselectivity remained moderate to low for both
the diastereomers with a wide range of phosphoramide
catalysts.
5.2.2. General procedure II. Pd(OAc)2-mediated trans-
silylation of 7a: (1Z)-2-tert-butyldimethylsilyloxy-1-tri-
chlorosilyloxyethene (4a). A solution of the TMS-enolate
7a (Z/E¼95/5) (6.1 g, 24.8 mmol) in CH2Cl2 (25 mL) was
drop wise added to a solution of SiCl4 (5.6 mL, 48.3 mmol,
2.0 equiv) and Pd(OAc)2 (118 mg, 0.52 mmol, 0.034 equiv)
in CH2Cl2 (20 mL) at room temperature. After 12 h, solvent
and volatiles were removed under vacuum (up to 10 mmHg).
The residual oil was distilled to afford 5.0 g (66%) of the eno-
late 4a as a mixture of isomers (Z/E¼95/5). Bp: 50–55 ꢀC
1
(0.1 mmHg); H NMR (400 MHz, CDCl3): 6.67 (d, J¼10,
5. Experimental section
5.1. General experimental procedures
See Supplementary data.
1H, (E)-4a), 6.45 (d, J¼10, 1H, (E)-4a), 5.85 (d, J¼3.1, 1H,
(Z)-4a), 5.67 (d, J¼3.1, (Z)-4a), 0.94 (s, 9H), 0.18 (s, 6H);
13C NMR (101 MHz, CDCl3): 136.08 ((E)-4a), 129.30
((Z)-4a), 127.57 ((E)-4a), 121.20 ((Z)-4a), 25.70 ((E)-4a),
25.59 ((Z)-4a), 18.39, ꢁ2.95 ((E)-4a), ꢁ5.32 ((Z)-4a).
5.2. Experimental procedures
5.2.3. General procedure III. Aldol reaction of isolated
trichlorosilyl enolate: (2S,3S)-1,1-dimethoxy-3-phenyl-
propane-2,3-diol (anti-9) and (2S,3R)-1,1-dimethoxy-3-
phenylpropane-2,3-diol (syn-9). Trichlorosilyl enolate (Z)-
4a (615 mg, 2 mmol) was added to a cold (ꢁ67 ꢀC) solution
of the phosphoramide 3 (148 mg, 0.4 mmol, 0.2 equiv) in
CH2Cl2 (7.5 mL) and the mixture was allowed to stir for
10 min. Freshly distilled benzaldehyde (0.204 mL,
2.0 mmol, 1.0 equiv) was then added. After 5.5 h, MeOH
(7.5 mL) was added and the mixture was slowly allowed to
attain room temperature. The reaction mixture was poured
into cold NaHCO3 solution and was stirred for 0.5 h. The
reaction mixture was diluted with EtOAc, filtered through
5.2.1. General procedure I. Preparation of enol
trimethylsilyl ethers of a-tert-butyldimethylsilyloxy acet-
aldehyde: (1Z)-1-tert-butyldimethylsilyloxy-2-trimethyl-
silyloxyethene ((Z)-7a). A solution of the aldehyde 8a19
(4.6 g, 26.4 mmol) in acetonitrile (20 mL) was added to
a stirred solution of TMSCl (6.7 mL, 52.8 mmol, 2.0 equiv)
in MeCN (30 mL). Triethylamine (14.8 mL, 105.6 mmol,
4.0 equiv) was then added and the reaction mixture was
warmed to 80 ꢀC (bath). After 2 h, the reaction mixture
was brought to room temperature. Excess TMSCl and vola-
tile components were removed under vacuum. The residual

S. E. Denmark, S. K. Ghosh / Tetrahedron 63 (2007) 8636–8644
8643
Celite, and washed well with EtOAc. The organic layer was
separated and the aqueous layer was once extracted with
EtOAc. The combined extracts were dried over MgSO4
and concentrated. Column chromatography (SiO2, hexane/
EtOAc, 40/60) of the residue afforded 313 mg (74%) of
a mixture of aldol products anti-9 and syn-9 as a clear
products (19 pages) are provided. Supplementary data asso-
ciated with this article can be found in the online version,
at doi:10.1016/j.tet.2007.04.009.
References and notes
1
colorless thick liquid (anti/syn, 3/1). H NMR (400 MHz,
CDCl3): 7.43–7.27 (m, 5H, H-Aryl), 4.91 (dd, J¼2.9, 5.1,
1H, syn-9), 4.79 (dd, J¼3.7, 6.1, 1H, anti-9), 4.36 (d, J¼
4.9, 1H, syn-9), 4.27 (d, J¼5.4, 1H, anti-9), 3.85 (ddd, J¼
3.9, 5.6, 6.3, 1H, anti-9), 3.74 (dt, J¼2.9, 5.1, 1H, syn-9),
3.49 (s, 3H, syn-9), 3.48 (s, 3H, syn-9), 3.46 (s, 3H, anti-
9), 3.42 (s, 3H, anti-9), 3.24 (d, J¼3.9, 1H, OH-anti-9),
3.03 (d, J¼5.1, 1H, OH-syn-9), 2.51 (d, J¼5.4, 1H, OH-
syn-9), 2.14 (d, J¼3.9, 1H, OH-anti-9); SFC: tR (2R,3S)-
syn-9, 4.87 min (10.6%); tR (2S,3R)-syn-9, 5.29 min
(14.1%); tR (2R,3R)-anti-9, 5.81 min (29.9%); tR (2S,3S)-
anti-9, 6.65 min (44.53%) (Column: AS, MeOH 2.5%, pres-
sure 150 psi, flow 3 mL); MS (FI): 212 (M+, 2), 181
(MꢁOMe, 11), 180 (MꢁMeOH, 100), 106 (PhCHO, 16),
75 (CH(OMe)2), 11), 74 (11).
1. (a) Denmark, S. E.; Stavenger, R. A. Acc. Chem. Res. 2000, 33,
432–440; (b) Denmark, S. E.; Fujimori, S. Modern Aldol
Reactions; Mahrwald, R., Ed.; Wiley-VCH: Weinheim, 2004;
Vol. 2, Chapter 7.
2. (a) Denmark, S. E.; Stavenger, R. A.; Wong, K.-T.; Su, X.
J. Am. Chem. Soc. 1999, 121, 4982–4991; (b) Denmark,
S. E.; Stavenger, R. A. J. Am. Chem. Soc. 2000, 122, 8837–
8847; (c) Denmark, S. E.; Fujimori, S.; Pham, S. M. J. Org.
Chem. 2005, 70, 10823–10840.
3. Denmark, S. E.; Fan, Y.; Eastgate, M. D. J. Org. Chem. 2005,
70, 5235–5248.
4. Denmark, S. E.; Pham, S. M.; Stavenger, R. A.; Su, X.; Wong,
K.-T.; Nishigaichi, Y. J. Org. Chem. 2006, 71, 3904–3922.
5. Denmark, S. E.; Stavenger, R. A.; Winter, S. B. D.; Wong,
K.-T.; Barsanti, P. A. J. Org. Chem. 1998, 63, 9517–9523.
6. (a) Heathcock, C. H. Comprehensive Organic Synthesis; Trost,
B. M., Fleming, I., Heathcock, C. H., Eds.; Pergamon: New
York, NY, 1991; Vol. 2, p 133; (b) Barbas, C. F., III;
Cordova, A.; Notz, W. J. Org. Chem. 2002, 67, 301–303; (c)
Nielsen, A. J. Am. Chem. Soc. 1957, 79, 2518–2524; (d)
Merger, F.; Fouquet, G.; Platz, R. Liebigs Ann. Chem. 1979,
1591–1601; (e) Day, A. R.; Lin, I. J. Am. Chem. Soc. 1952,
74, 5133–5135; (f) Mahrwald, R. Modern Aldol Reactions;
Mahrwald, R., Ed.; Wiley-VCH: Weinheim, 2004; Vol. 2,
Chapter 8.8.
5.2.4. General procedure IV. Aldol addition of in situ
generated trichlorosilyl enolate using 8b: (1S,4R)-5-tert-
butyldimethylsilyloxy-4-methyl-1-hydroxy-1-phenyl-3-
pentanone (syn-28a). The TMS-enolate 7b (Z/E¼94/6)
(0.33 mL, 1 mmol) was added drop wise to a solution of
SiCl4 (0.24 mL, 2.0 mmol, 2.0 equiv) and Pd(OAc)2
(2.2 mg, 0.001 mmol, 0.01 equiv) in CH2Cl2 (1.2 mL). After
13 h at room temperature, solvent and volatile components
were removed under vacuum (up to 0.5 MmHg) to afford
350 mg (100%) of the TCS enolate 4b (Z/E¼90/10). The
residue was diluted with chloroform to a total volume of
2 mL. A portion of this chloroform solution (0.5 mL,
0.25 mmol) containing enolate 4b was added to a solution
of the chiral phosphoramide 3 (9.5 mg, 0.10 mmol,
0.10 equiv) in chloroform (0.5 mL) at ꢁ63 ꢀC. After
10 min, benzaldehyde (26 mL, 0.25 mmol, 1.0 equiv) was
added into the reaction mixture. After 4 h, MeOH
(7.5 mL) was added and the mixture was slowly allowed to
attain to room temperature. The reaction mixture was poured
into cold NaHCO3 solution and was stirred for 0.5 h. The
reaction mixture was diluted with EtOAc, filtered through
Celite, and washed well with EtOAc. The organic layer was
separated and the aqueous layer was once extracted with
EtOAc. The combined extracts were dried over MgSO4
and concentrated. Column chromatography (SiO2, EtOAc/
hexanes, 60/40) of the residue gave 35 mg (66%) of a mix-
ture of aldol products anti-9 and syn-9 as a clear colorless
thick liquid (anti/syn, 6/1).
7. (a) Denmark, S. E.; Ghosh, S. K. Angew. Chem., Int. Ed. 2001,
40, 4759–4762; (b) Denmark, S. E.; Bui, T. Proc. Natl. Acad.
Sci. U.S.A. 2004, 101, 5439–5444; (c) Denmark, S. E.; Bui,
T. J. Org. Chem. 2005, 70, 10393–10399.
8. (a) Evans, D. A.; Hu, J.; Tedrow, J. S. Org. Lett. 2001, 3, 3133–
3136; (b) Davies, S. G.; Nicholson, R. L.; Smith, A. D. Synlett
2002, 1637–1640; (c) Sibi, M. P.; Lu, J.; Edwards, J. J. Org.
Chem. 1997, 62, 5864–5872; (d) Takayama, S.; McGarvey,
G. J.; Wong, C.-H. Chem. Soc. Rev. 1997, 26, 407–415.
9. (a) Noltz, W.; List, B. J. Am. Chem. Soc. 2000, 122, 7386–7387;
(b) Sakthivel, K.; Noltz, W.; Bui, T.; Barbas, C. F. J. Am. Chem.
Soc. 2001, 123, 5260–5267; (c) Enders, D.; Jegelka, U. Synlett
1992, 999–1004; (d) Mukaiyama, T.; Iwasawa, N.; Stevens,
R. W.; Haga, T. Tetrahedron 1984, 40, 1381–1384; (e)
Marco, J. A.; Carda, M.; Falomir, E.; Palomo, C.; Oiarbide,
M.; Oritz, J. A.; Linden, A. Tetrahedron Lett. 1999, 40,
1065–1070; (f) Kim, K. S.; Hong, S. D. Tetrahedron Lett.
2000, 41, 5909–5913.
10. (a) Northrup, A. B.; Mangion, I. K.; Hettche, F.; MacMillan,
D. W. C. Angew. Chem., Int. Ed. 2004, 43, 2152–2154; (b)
Casas, J.; Engqvist, M.; Ibrahem, I.; Kaynak, B.; Cordova, A.
Angew. Chem., Int. Ed. 2005, 44, 1343–1345.
Acknowledgements
We are grateful to the National Science Foundation for
generous financial support (NSF CHE0105205 and
CHE0414440).
11. Denmark, S. E.; Bui, T. J. Org. Chem. 2005, 70, 10190–10193.
12. For a definition of ambidoselectivity, see: Seebach, D. Angew.
Chem., Int. Ed. Engl. 1979, 18, 239–258.
13. Denmark, S. E.; Beutner, G. L.; Wynn, T.; Eastgate, M. D.
J. Am. Chem. Soc. 2005, 127, 3774–3789.
Supplementary data
14. (a) Trost, B. M.; Urabe, H. J. Org. Chem. 1990, 55, 3982–3984;
(b) Vorbruggen, H.; Krolikiewicz, K.; Bennua, B. Chem. Ber.
1981, 114, 1234–1235; (c) Corey, E. J.; Cho, H.; Rucker, C.;
Hua, D. H. Tetrahedron Lett. 1981, 22, 3455–3458.
Complete experimental details for all preparative procedures
along with full characterization of all starting materials and

8644
S. E. Denmark, S. K. Ghosh / Tetrahedron 63 (2007) 8636–8644
15. Matthews, B. R.; Jackson, W. R.; Jacobs, H. A.; Watson, K. G.
Aust. J. Chem. 1990, 43, 1195–1214.
16. Jellen, W.; Mittelbach, M.; Junek, H. Monatsh. Chem. 1996,
127, 167–172.
17. (a) Denmark, S. E.; Wynn, T. J. Am. Chem. Soc. 2001, 123,
6199–6200; (b) Denmark, S. E.; Beutner, G. L. J. Am. Chem.
Soc. 2003, 125, 7800–7801; (c) Denmark, S. E.; Heemstra,
J. R., Jr. J. Am. Chem. Soc. 2006, 128, 1038–1039.
18. For anti-selective catalytic aldol additions: (a) Evans, D. A.;
MacMillan, D. W. C.; Campos, K. R. J. Am. Chem. Soc.
1997, 119, 10859–10860; (b) Kobayashi, S.; Horibe, M.;
Hachiya, I. Tetrahedron Lett. 1995, 36, 3173–3176; (c)
Yamashita, Y.; Ishitani, H.; Shimizu, H.; Kobayashi, S.
J. Am. Chem. Soc. 2002, 124, 3292–3302; (d) See Ref. 12.
19. Nicaise, O. J. C.; Denmark, S. E. Bull. Soc. Chim. Fr. 1997,
134, 395–398.