A tandem enol silane formation-Mukaiyama aldol reaction mediated by TMSOTf

Article abstract of DOI:10.1016/j.tetlet.2007.03.088

A slight excess of silyl trifluoromethanesulfonate mediates a tandem enol silane formation-Mukaiyama aldol reaction in the presence of Hunig's base. Preformation of the enol silane is unnecessary for efficient reactions, which proceed in 75-97% yield for the addition of aryl methyl ketones and acetate esters to non-enolizable aldehydes. Mechanistic data suggests that free amine is crucial for full conversion.

Full text of DOI:10.1016/j.tetlet.2007.03.088

Tetrahedron Letters 48 (2007) 3559–3562
A tandem enol silane formation-Mukaiyama aldol
reaction mediated by TMSOTf
C. Wade Downey* and Miles W. Johnson
Department of Chemistry, University of Richmond, Richmond, VA 23173, USA
Received 22 January 2007; revised 6 March 2007; accepted 17 March 2007
Available online 21 March 2007
Abstract—A slight excess of silyl trifluoromethanesulfonate mediates a tandem enol silane formation-Mukaiyama aldol reaction in
the presence of Hunig’s base. Preformation of the enol silane is unnecessary for efficient reactions, which proceed in 75–97% yield for
the addition of aryl methyl ketones and acetate esters to non-enolizable aldehydes. Mechanistic data suggests that free amine is cru-
cial for full conversion.
Ó 2007 Elsevier Ltd. All rights reserved.
The Mukaiyama aldol reaction1 has become an impor-
tant tool in organic synthesis in recent decades,2 display-
ing utility in both fragment coupling3 and building block
construction.4 One drawback, however, is the necessity
of preforming the requisite enol silane nucleophile. Fur-
thermore, cationic silicon species generated under the
reaction conditions may interfere with stereoselective
Mukaiyama aldol processes mediated by exogenous
Lewis acids.5 In our efforts to address these problems,
we have discovered that under specific reaction condi-
tions TMSOTf acts as both a silylating agent and a
Lewis acid, thus mediating a tandem enol silane forma-
tion-Mukaiyama aldol reaction, which obviates the need
for preformation and purification of the enol silane.
OTMS
Ph
O
O
O
TMSOTf
2,6-lutidine or
Ph
Me
H
Ph
Ph
i
-Pr2NEt
90% conv
ð1Þ
Literature accounts of the ability of silyl trifluoro-
methanesulfonates to mediate Mukaiyama-type reac-
tions are somewhat conflicting. Pioneering work by
Noyori in 1980 described the use of TMSOTf to catalyze
the addition of enol silanes to dimethyl acetals, but no
reaction was observed with aldehyde electrophiles.8 A
decade later, Hanaoka reported that TMSOTf cleanly
catalyzes the addition of acetophenone enol silane to
benzaldehyde under conditions seemingly identical to
those employed by Noyori.9 Likewise, Bosnich observed
that TMSOTf was a very effective catalyst in the pres-
ence of tetrabutylammonium trifluoromethanesulfo-
nate.5 We set out to clarify this ambiguity via a
mechanistic investigation of our own system, with the
goal of providing synthetic chemists with a predictive
model for avoiding competitive catalysis by adventitious
cationic silicon during Mukaiyama aldol reactions.
Our efforts to develop a direct aldol reaction of ester and
ketone enolates led us to the investigation of a number
of metal catalysts in the presence of TMSOTf and an
amine base.6 After the identification of a promising reac-
tion, we were surprised when control experiments indi-
cated that the reaction occurred even in the absence of
a metal catalyst (Eq. 1).7 Given the reaction conditions,
we suspected that a Mukaiyama aldol pathway was a
likely possibility, wherein the requisite enol silane nucleo-
phile was formed in situ. Subsequent Mukaiyama aldol
addition, catalyzed by a small amount of unreacted
TMSOTf, might then explain the observed product.
To this end, the putative acetophenone enol silane inter-
mediate was independently synthesized and purified.10
Treatment of this enol silane with benzaldehyde and
0.1 equiv TMSOTf in methylene chloride provided no
aldol product, as suggested by Noyori’s experiments
(Eq. 2). Confronted by the apparent inadequacy of
TMSOTf to act as a Lewis acid, we turned to the
Keywords: Mukaiyama aldol; Silyl triflate; Silyl trifluoromethanesulf-
onate; Enol silane.
*
Corresponding author. Fax: +1 804 287 1897; e-mail: wdowney@
richmond.edu
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.03.088

3560
C. W. Downey, M. W. Johnson / Tetrahedron Letters 48 (2007) 3559–3562
Table 1. Nucleophilic additives in the Mukaiyama aldol reactiona
possibility that Hunig’s base might act as a catalytic
Lewis base under our reaction conditions, activating
the nucleophile by attack upon the silicon atom of the
enol silane.11 Treatment of the enol silane and benzalde-
hyde with Hunig’s base, however, also provided no
product (Eq. 3). Based on these results, we chose to
add catalytic amounts of both TMSOTf (0.1 equiv)
and Hunig’s base (0.2 equiv) to the same reaction mix-
ture, which afforded >90% conversion of the enol silane
to the silylated aldol product (Eq. 4). Interestingly, when
0.2 equiv TMSOTf and 0.1 equiv Hunig’s base were
employed, only low conversion was observed (14%).12
OTMS
O
0.2 equiv Nu
OTMS
Ph
O
Ph
H
Ph
0.1 equiv TMSOTf
Ph
Entry
Nucleophile
Conversionb (%)
1
2
3
4
5
6
7
i-Pr2NEt
2,6-Lutidine
Et3N
Imidazole
Ph3P
90
71
6
<5
<5
52
6
TBAI
TBABr
SH
8
0
O
O
O
OTMS
OTMS
OTMS
0.1 equiv TMSOTf
ð2Þ
ð3Þ
No Reaction
No Reaction
Me
H
H
H
Ph
Ph
Ph
Ph
Ph
Ph
a Reaction conditions: acetophenone enol silane (1 equiv), aldehyde
(1.4 equiv), TMSOTf (0.1 equiv), nucleophile (0.2 equiv), CH2Cl2
(0.2 M), 0 °C to rt, 24 h.
1.0 equiv
0.2 equiv
i
i
-Pr2NEt
-Pr2NEt
b Conversion calculated by 1H NMR spectroscopic analysis. No fur-
ther reaction was observed after 24 h.
OTMS
Ph
O
0.1 equiv TMSOTf
Ph
the effects of other potential Lewis basic catalysts upon
the TMSOTf-catalyzed Mukaiyama aldol reaction
(Table 1).
90% conversion
ð4Þ
ð5Þ
O
OTMS
0.1 equiv
i-Pr2NEt
OTMS
Ph
O
As expected, the two amine bases that mediate the tan-
dem enol silane formation-Mukaiyama aldol reaction
also are active catalysts for the TMSOTf-catalyzed
Mukaiyama aldol (Table 1, entries 1 and 2). In contrast,
strongly nucleophilic amines such as triethylamine and
imidazole provided low conversion to the Mukaiyama
aldol adduct (entries 3 and 4). We speculate that these
amines react irreversibly with TMSOTf to form a nucleo-
phile-TMS salt, removing the silicon catalyst from the
reaction mixture. The same trend was observed when
non-nitrogen nucleophiles were tested (entries 5–8); only
the very weakly nucleophilic tetrabutylammonium
iodide (TBAI) provided a conversion comparable to
Hunig’s base and 2,6-lutidine.13,14 Upon examination
of these data, it appears that only very weak nucleo-
philes are appropriate co-catalysts with TMSOTf, which
may be attributable either to reversible complexation of
the nucleophile to the silicon center or to the formation
of a pentavalent silicon species.
H
Ph
Ph
0.2 equiv TMSOTf
Ph
14% conversion
It is clear from these experiments that catalytic quanti-
ties of both TMSOTf and an amine base are necessary
for high conversion to the Mukaiyama aldol adduct.
Furthermore, an excess of Hunig’s base in comparison
to TMSOTf greatly increases the efficacy of the reaction
conditions. Based on these findings, we suggest that
TMSOTf alone is not capable of interfering with the ste-
reoselectivity of chiral Lewis acids in Mukaiyama aldol
reactions, but in the presence of a weakly nucleophilic
Lewis base, cationic silicon is a highly active achiral
Mukaiyama aldol catalyst. This data is of tremendous
value to the organic chemist for whom high stereoselec-
tivity is a paramount goal in complex fragment coupling
reactions, the formation of chiral building blocks, or
other related processes.
A general mechanistic scheme consistent with these data
is presented in Scheme 1. Treatment of acetophenone
with TMSOTf and Hunig’s base produces the enol
silane nucleophile. Unreacted TMSOTf then activates
the aldehyde, and addition occurs in the presence of
Hunig’s base. Silyl transfer from the aldol adduct to
another molecule of aldehyde completes the catalytic
cycle.
Although it is apparent that both TMSOTf and an
amine must be present, the exact role of Hunig’s base
is not entirely clear. In the absence of aldehyde, aceto-
phenone is converted to the enol silane as expected,
regardless of whether TMSOTf or Hunig’s base is in
excess. Even when aldehyde is present in the reaction
mixture, no aldol products are observed unless an excess
of Hunig’s base in comparison to TMSOTf is present,
although that excess may be as little as 0.1 equiv. It
has not been determined whether this additional amine
acts as a silyl transfer catalyst, as a nucleophilic activa-
tor of the enol silane, or as a nucleophilic activator of
TMSOTf. In each of these scenarios, however, the amine
must act as a Lewis base, which prompted us to survey
As our mechanistic studies progressed, we began to
investigate the synthetic applications of the tandem enol
silane formation-Mukaiyama aldol reaction. A brief
optimization survey determined standard reaction con-
ditions to be ketone (1 equiv), aldehyde (1.4 equiv),
TMSOTf (1.2 equiv),15 i-Pr2NEt (1.5 equiv),16 and
CH2Cl2 (0.2 M) at room temperature. Under these con-
ditions, the addition of acetophenone to a range of non-

C. W. Downey, M. W. Johnson / Tetrahedron Letters 48 (2007) 3559–3562
3561
rich and electron-poor aromatic aldehydes are reactive
substrates under these conditions. Heterocyclic alde-
hydes were also effective reaction partners (entries 5–
6), as were sterically demanding naphthyl derivatives
(entries 7–8). A slight decrease in yield was observed
with a,b-unsaturated aldehydes (entries 9–10), but the
synthetic utility of these products makes them particu-
larly attractive, since ozonolysis provides access to an
a-alkoxy aldehyde. Attempted addition of acetophenone
to enolizable aldehydes and to acrolein provided no de-
sired products, presumably due to competing polymeri-
zation pathways.
O
O
Ph
Me
-PrNEt
H
Ph
i
TMSOTf
TMSOTf
Me3Si
H
OTMS
R3NH
O
TfO
Ph
Ph
PhCHO
i
-Pr2NEt
We were particularly interested in the scope of the eno-
late partner for this tandem enol silane formation-
Mukaiyama aldol reaction, because the reaction condi-
tions obviate the need for preformation and purification
of the enol silane (Table 3). Aryl methyl ketones, includ-
ing acetophenone and acetonaphthones, were superior
substrates, providing high yields of aldol adducts. Grat-
ifyingly, acetate esters also reacted well, providing
b-hydroxy ester products in >80% yield. This yield is
outstanding when compared to the standard Mukai-
yama aldol procedures, which require a separate synthe-
sis and purification of the ester-derived silyl ketene acetal
nucleophile in 40–70% yield17 before the addition step
can performed. Thus, our system provides a significant
increase in yield over the standard Mukaiyama aldol
procedure for ester-derived enol silanes.
OTMS
Ph
O
R3N SiMe3
Ph
Scheme 1.
Table 2. Aldehyde scopea
1. TMSOTf, i-PrNEt
OH
O
O
O
R
Ph
Me
H
R
Ph
2. 1.0 N HCl, THF
Entry
R
Product
Yieldb (%)
1
2
3
4
X = H
1
2
3
4
96
93
92
92
X = Me
X = OMe
X = NO2
At present, addition reactions of alkyl–alkyl ketones
proceed in somewhat diminished conversion (50–
X
O
Table 3. Enolate scopea
5
6
7
5
6
7
93
93
85
1. TMSOTf,
2. 1.0 N HCl, THF
Product
i-PrNEt
OH
Ph
O
O
O
S
R
Me
H
Ph
R
Entry
1
RCOMe
Yieldb (%)
O
1
96
Me
O
8
8
9
95
2
3
11
99
89
Me
9
75
88
O
12
Me
10
10
Me
a Reaction conditions: (1) acetophenone (1 equiv), aldehyde
(1.4 equiv), TMSOTf (1.2 equiv), i-Pr2NEt (1.5 equiv), CH2Cl2
(0.2 M), 0 °C to rt, 24 h; (2) 1.0 N HCl (3 equiv), THF (0.05 M), rt,
1 h.
O
4
5
13
14
82
83
EtO
Me
b Isolated yield after chromatography.
O
i
-PrO
Me
a Reaction conditions: (1) enolate precursor (1 equiv), benzaldehyde
(1.4 equiv), TMSOTf (1.2 equiv), i-Pr2NEt (1.5 equiv), CH2Cl2
(0.2 M), 0 °C to rt, 24 h; (2) 1.0 N HCl (3 equiv), THF (0.05 M).
b Isolated yield after chromatography.
enolizable aldehydes was observed (Table 2). Benzalde-
hyde and its derivatives were uniformly excellent sub-
strates, providing the b-hydroxy aldol adducts after
desilylation in >90% yield (entries 1–4). Both electron-

3562
C. W. Downey, M. W. Johnson / Tetrahedron Letters 48 (2007) 3559–3562
75%),11a but further optimization of these substrates is
underway. Propionate nucleophiles such as propiophe-
none react more sluggishly, and with poor diastereo-
selectivity (70% conversion, 1.5:1 syn:anti).11a
2. For a review of the Mukaiyama aldol reaction, see:
Carreira, E. M. Comp. Asym. Catal. 1999, 3, 997–1065.
3. For recent examples, see: (a) Rech, J. C.; Floreancig, P. E.
Org. Lett. 2005, 7, 5175–5178; (b) Terauchi, T.; Terauchi,
T.; Sato, I.; Shoji, W.; Tsukada, T.; Tsunoda, T.; Kanoh,
N.; Nakata, M. Tetrahedron Lett. 2003, 44, 7741–7745; (c)
Savall, B. M.; Blanchard, N.; Roush, W. R. Org. Lett.
2003, 5, 377–379.
4. For examples of the construction of chiral building blocks,
see: (a) Ishitani, H.; Yamashita, Y.; Shimizu, H.; Ko-
bayashi, S. J. Am. Chem. Soc. 2000, 122, 5403–5404; (b)
Evans, D. A.; Kozlowski, M. C.; Murry, J. A.; Burgey, C.
S.; Campos, K. R.; Connell, B. T.; Staples, R. J. J. Am.
Chem. Soc. 1999, 121, 669–685; (c) Carreira, E. M.; Singer,
R. A.; Lee, W. J. Am. Chem. Soc. 1994, 116, 8837–8838.
5. Hollis, T. K.; Bosnich, B. J. Am. Chem. Soc. 1995, 117,
4570–4581.
6. Evans and co-workers have documented the use of
silylating reagents in conjunction with Lewis acid-cata-
lyzed aldol reactions, but there is significant evidence that
these reactions do not occur via a Mukaiyama aldol
pathway. See: (a) Evans, D. A.; Tedrow, J. S.; Shaw, J. T.;
Downey, C. W. J. Am. Chem. Soc. 2002, 124, 392–393; (b)
Evans, D. A.; Downey, C. W.; Shaw, J. T.; Tedrow, J. S.
Org. Lett. 2002, 4, 1127–1130; (c) Evans, D. A.; Downey,
C. W.; Hubbs, J. L. J. Am. Chem. Soc. 2003, 125, 8706–
8707; For an Yb-catalyzed silylative aldol reaction, see: (d)
Kagawa, N.; Toyota, M.; Ihara, M. Aust. J. Chem. 2004,
57, 655–658.
Other silyl trifluoromethanesulfonates are also effective
mediators of this aldol addition. When TMSOTf was re-
placed with TESOTf under otherwise identical reaction
conditions, acetophenone and benzaldehyde reacted in
97% yield (Eq. 6). The reaction rate with TBSOTf suf-
fered considerably, however, providing the aldol adduct
in only 54% yield after 72 h.18
1. R3SiOTf,
i-PrNEt
OH
Ph
O
O
O
Ph
Me
H
Ph
Ph
2. acid workup
1a
R3Si = TMS, 96% yield
R3Si = TES, 97% yield
R3Si = TBS, 54% yield
ð6Þ
In conclusion, we have developed a new tandem enol
silane formation-Mukaiyama aldol process that clarifies
our understanding of the power of silyl trifluorome-
thanesulfonates to catalyze Mukaiyama aldol reactions.
This new method proceeds without preformation and
purification of the enol silane nucleophile. The presence
of an amine base is crucial for high conversion to the
aldol adducts. Mechanistic data suggests that both the
amine and the silyl trifluoromethanesulfonate play dual
roles in the reaction mechanism. Based on these results,
synthetic chemists can more accurately predict when cat-
ionic silicon species may interfere with stereoselective
Mukaiyama aldol systems. Expansion of the reaction
scope and the development of stereoselective variants
of this transformation are underway.
7. A similar, intramolecular reaction with an ester enolate
has recently been reported by Hoye: Hoye, T. R.;
Dvornikovs, V.; Sizova, E. Org. Lett. 2006, 8, 5191–
5194.
8. Murata, S.; Suzuki, M.; Noyori, R. J. Am. Chem. Soc.
1980, 102, 3248–3249.
9. Mukai, C.; Hashizume, S.; Nagami, K.; Hanaoka, M.
Chem. Pharm. Bull. 1990, 38, 1509–1512.
10. Anson, C. E.; Creaser, C. S.; Malkov, A. V.; Mojovic, L.;
Stephenson, G. J. Organomet. Chem. 2003, 668, 101–122.
11. Lewis base-catalyzed Mukaiyama aldol reactions are
known. For example, see: (a) Denmark, S. E.; Fan, Y. J.
Am. Chem. Soc. 2002, 124, 4233–4235; (b) Oi, T.; Doda,
K.; Keiji, M. Org. Lett. 2001, 3, 1273–1276.
Acknowledgement
12. (a) No further reaction was observed after 24 h; (b) NMR
experiments suggest that Hunig’s base and TMSOTf form
an adduct under the reaction conditions.
13. No product was observed when TBAI was mixed with
acetophenone enol silane and benzaldehyde for 24 h,
eliminating the possibility of a pure Lewis base-catalyzed
reaction. Likewise, no aldol product was observed when
Hunig’s base was replaced with TBAI in the tandem enol
silane formation-Mukaiyama aldol reaction.
We gratefully acknowledge the Thomas F. and Kate
Miller Jeffress Memorial Trust (J-208) and the Univer-
sity of Richmond for funding. M.W.J. thanks the How-
ard Hughes Medical Institute for a summer research
´
fellowship. We thank Dr. Rene Kanters, Dr. Raymond
Dominey and Dr. Timothy Smith for assistance with
mass spectral data. We thank Dr. John Gupton for help-
ful discussions.
14. These results with TBAI are very similar to Bosnich’s
results with TBAOTf. See Ref. 5.
15. No product was observed with TMSCl.
16. Use of 2,6-lutidine resulted in a slight decrease in yield.
Poor yield (<20%) was observed with Et3N.
Supplementary data
17. For recent examples of the synthesis of the silyl ketene
acetal derived from ethyl acetate by treatment with
LiHMDS and TMSCl, see: (a) Oisaki, K.; Suto, Y.;
Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2003, 125,
5644–5645; (b) Ding, F.; Kopach, M. E.; Sabat, M.;
Harman, W. D. J. Am. Chem. Soc. 2002, 124, 13080–
13087.
Experimental procedures and spectral data are avail-
able. Supplementary data associated with this article
can be found, in the online version, at doi:10.1016/
j.tetlet.2007.03.088.
18. Reaction proceeded to completion as determined by 1H
NMR. In order to remove the TBS group, the initial
adduct was treated with trifluoroacetic acid rather than
HCl.
References and notes
1. Mukaiyama, T.; Banno, K.; Narasaka, K. J. Am. Chem.
Soc. 1974, 96, 7503–7509.