Acetic acid aldol reactions in the presence of trimethylsilyl trifluoromethanesulfonate

Article abstract of DOI:10.1021/jo100828c

(Figure presented) In the presence of TMSOTf and a trialkylamine base, acetic acid undergoes aldol addition to non-enolizable aldehydes under exceptionally mild conditions. Acidic workup yields the β-hydroxy carboxylic acid. The reaction appears to procee

Full text of DOI:10.1021/jo100828c

pubs.acs.org/joc
(TMSOTf) and an amine base, yielding β-hydroxy carboxylic
Acetic Acid Aldol Reactions in the Presence of
Trimethylsilyl Trifluoromethanesulfonate
acids (eq 2). Syntheses of these products are exceedingly rare, as
illustrated by the fact that only two of the 13 products described
in this manuscript have ever been fully characterized in the
literature.3
C. Wade Downey,* Miles W. Johnson,
Daniel H. Lawrence, Alan S. Fleisher, and Kathryn J. Tracy
Gottwald Center for the Sciences, University of Richmond,
Richmond, Virginia 23173
wdowney@richmond.edu
Received April 27, 2010
Motivated by our interest in silylation-driven additions to
carbonyl compounds,4 we recently reported the one-pot enol
silane formation-Mukaiyama aldol addition of esters to
nonenolizable aldehydes, in which TMSOTf served as both
silylating agent and Lewis acid catalyst (eq 3).4a,5 We specu-
lated that silyl esters would also be effective enolate pre-
cursors under similar reaction conditions and would give rise
to carboxylic acid aldol products after desilylative workup.
Bellassoued has pioneered Lewis acid catalyzed Mukaiyama
aldol reactions of bis-silyl ketene acetals but describes only a
single example of the addition of an acetate nucleophile to an
aldehyde, and the reaction requires preformation and puri-
fication of the nucleophile (eq 4).6 Accordingly, we set out to
apply our one-pot enol silane formation-Mukaiyama aldol
strategy to this problem.
In the presence of TMSOTf and a trialkylamine base, acetic
acid undergoes aldol addition to non-enolizable aldehydes
under exceptionally mild conditions. Acidic workup yields
the β-hydroxy carboxylic acid. The reaction appears to
proceed via a three-step, one-pot process, including in situ
trimethylsilyl ester formation, bis-silyl ketene acetal forma-
tion, and TMSOTf-catalyzed Mukaiyama aldol addition.
Independently synthesized TMSOAc also undergoes aldol
additions under similar conditions.
The use of carboxylic acid derivatives in R-substitution reac-
tions, including aldol addition reactions, is well developed.1 The
parent carboxylic acids, however, are seldom employed in aldol
reactions because their inherent Brønsted acidity results in
deprotonation of the acid proton rather than the R-carbon. A
second deprotonation to yield the dianion is possible, but
harshly basic conditions are required and the highly reactive
dianion is difficult to control (eq 1).2 The development of a mild
and general aldol reaction of carboxylic acids would be a
desirable addition to the field of organic synthesis because of
the synthetic versatility of the carboxylic acid group, which can
be easily converted to the corresponding ester, anhydride, or
acid halide. We now report that acetic acid undergoes one-pot
bis-silyl ketene acetal formation-Mukaiyama aldol reactions
in the presence of trimethylsilyl trifluoromethanesulfonate
We began our investigation by treating commercially
available trimethylsilyl acetate (TMSOAc) with i-Pr2NEt,
benzaldehyde, and TMSOTf in CH2Cl2 for 2 h at room
temperature and were pleased to observe >95% conversion
to aldol addition adducts (eq 5). No R,β-unsaturated aldol
condensation products were observed. Although the unpuri-
fied reaction mixture included products silylated at one,
(3) (a) Saito, S.; Kobayashi, S. J. Am. Chem. Soc. 2006, 128, 8704–8705.
(b) Bietti, M.; Capone, A. J. Org. Chem. 2006, 71, 5260–5267.
(4) (a) Downey, C. W.; Johnson, M. W. Tetrahedron Lett. 2007, 48, 3559–
3562. (b) Downey, C. W.; Johnson, M. W.; Tracy, K. J. J. Org. Chem. 2008,
73, 3299–3302. (c) Downey, C. W.; Mahoney, B. D.; Lipari, V. R. J. Org.
Chem. 2009, 74, 2904–2906.
(1) For recent reviews, see: (a) Geary, L. M.; Hultin, P. G. Tetrahedron:
Asymmetry 2009, 20, 131–173. (b) Abiko, A. Org. Synth. 2002, 79, 116–124.
(c) Zappia, G.; Cancelliere, G.; Gacs-Baitz, E.; Delle Monache, G.; Misiti,
D.; Nevola, L.; Botta, B. Curr. Org. Synth. 2007, 4, 238–307. (d) Kimball,
D. B.; Silks, L. A., III Curr. Org. Chem. 2006, 10, 1975–1992.
(2) For examples of the successful execution of this strategy, see:
(a) Parra, M.; Sotoca, E.; Gil, S. Eur. J. Org. Chem. 2003, 8, 1386–1388.
(b) Galatsis, P.; Manwell, J. J.; Blackwell, J. M. Can. J. Chem. 1994, 72, 1656–
1659. (c) Fringuelli, F.; Martinetti, E.; Permatti, O.; Pizzo, F. Gazz. Chim.
Ital. 1993, 123, 637–640. For an example of a soft enolization strategy similar
to the one reported here, using Bu2BOTf as the Lewis acid, see: (d) Evans,
D. A.; Nelson, J. V.; Vogel, E.; Taber, T. R. J. Am. Chem. Soc. 1981, 103,
3099–3111.
(5) For a similar strategy applied to intramolecular cases, see: (a) Hoye,
T. R.; Dvornikovs, V.; Sizova, E. Org. Lett. 2006, 8, 5191–5194. (b) Rassu,
G.; Auzzas, L.; Pinna, L.; Zombrano, V.; Battistini, L.; Zanardi, F.;
Marzocchi, L.; Acquotti, D.; Casiraghi, G. J. Org. Chem. 2001, 66, 8070–
8075.
(6) (a) Bellassoued, M.; Gaudernar, M. J. Organomet. Chem. 1988, 338,
149–158. (b) Bellassoued, M.; Gaudernar, M. J. Organomet. Chem. 1990,
393, 19–25.
DOI: 10.1021/jo100828c
r
Published on Web 06/30/2010
J. Org. Chem. 2010, 75, 5351–5354 5351
2010 American Chemical Society

Downey et al.
JOCNote
SCHEME 1. Proposed Mechanistic Scheme
TABLE 1. Aldol Addition of Acetic Acid to Various Aldehydes
both, or neither hydroxyl group, acid-catalyzed desilylation
of the product mixture afforded a single β-hydroxy car-
boxylic acid.
Encouraged by this result, we next tested the reactivity of
acetic acid itself under similar conditions. By changing the
reaction conditions to include an extra 1 equiv of both
TMSOTf and i-Pr2NEt, one-pot conversion of acetic acid
to the Mukaiyama aldol adduct was achieved (eq 6).
aConditions A: 1.0 mmol acetic acid, 2.5 mmol i-Pr2NEt, 1.4 mmol
RCHO, 2.2 mmol TMSOTf, 5.0 mL CH2Cl2, rt. Conditions B: 1.0 mmol
acetic acid, 2.8 mmol i-Pr2NEt, 1.4 mmol RCHO, 2.5 mmol TMSOTf,
5.0 mL CH2Cl2, rt. bIsolated yield after chromatography.
Our interpretation of the reaction mechanism is outlined
in Scheme 1. First, i-Pr2NEt and TMSOTf react with acetic
acid to form TMSOAc in situ. A second equivalent of
TMSOTf activates the silyl ester toward deprotonation,
which is carried out by a second equivalent of i-Pr2NEt,
forming the bis-silyl ketene acetal. Finally, residual TMSOTf
catalyzes Mukaiyama aldol addition of the bis-silyl ketene
acetal to the aldehyde. In our hands, partial desilylation of
the initial aldol product occurs either in situ or during a brief
filtration through a pad of silica, resulting in a mixture of
monosilylated, bis-silylated, and fully desilylated species.
Subsequent treatment with trifluoroacetic acid (TFA) in
95% ethanol provides the final β-hydroxy carboxylic acid
product. Control experiments confirm that TMSOTf is
required for desired product formation; neither TMSOAc
nor acetic acid provides aldol adducts in the absence of
trimethylsilyl trifluoromethanesulfonate.
A brief survey of the reaction conditions verified that our
standard one-pot Mukaiyama aldol procedure4a was opti-
mal for this one-pot, three-step process. Of the amines tested,
i-Pr2NEt performed more consistently than Et3N and 2,6-
lutidine when acetic acid was used as the enolate precursor.
Toluene, THF, Et2O, and acetonitrile were all notably inferior
solvents compared to CH2Cl2. When TMSOTf was replaced
with TESOTf, reactivity slowed considerably, affording less
than 50% conversion after 24 h. Desilylation with 95%
EtOH7 and TFA proved optimal, rendering aqueous workup
unnecessary afterremoval ofthe solventinvacuo. Attemptsto
isolate the products via acid-base extraction generally pro-
vided poor or irreproducible yields, but the products were
easily purified by flash chromatography.
We tested the scope of the reaction by adding acetic acid to
a variety of nonenolizable aldehydes (Table 1). Benzaldehyde
derivatives were outstanding electrophiles, including both
electron-rich and electron-poor acceptors (entries 1-4). One
exception was p-nitrobenzaldehyde, which generally reacted
with <50% conversion.8 Heteroaromatic substrates (entries 5
and 6) and naphthyl substrates (entries 7 and 8) also provided
good yields of the carboxylic acid aldol products. Cinnamalde-
hyde was a surprisingly poor substrate under our one-pot
conditions (entry 9), but R-methylcinnamaldehyde reacted effi-
ciently to yield the synthetically useful allylic alcohol (entry 10).
Subjection of enolizable aldehydes to these conditions did not
provide the desired products, presumably because of competing
aldehyde-aldehyde condensation and polymerization reactions.
Nonetheless, the success of this reaction is remarkable, as
illustrated by the fact that only two of the adducts in Table 1
(entries 13a and 33b) have been characterized in the literature
prior to this report.
We also tested the efficacy of carboxylate salts in this one-
pot reaction. Sodium acetate was unreactive and appeared
to be insoluble in the solvents tested under our reaction
conditions.9 In contrast, tetrabutylammonium acetate dis-
solved rapidly and added to benzaldehyde to provide the
aldol adduct in 78% yield (eq 7). In addition to providing a
useful alternative to acetic acid, the results of the tetrabutyl-
ammonium acetate experiment are in accord with our pro-
posed mechanism. Because deprotonation of the acetic acid
(8) We speculate that the nitro group may interact unfavorably with the
TMSOTf catalyst, although addition of 1.4 equiv of nitrobenzene to the
standard reaction with benzaldehyde does not affect conversion for that
substrate. Note that the other electron-poor aromatic aldehyde tested,
4-fluorobenzaldehyde, reacts in high yield.
(7) Although methanol was also an effective solvent for the desilylation
reaction, competing formation of the methyl ester via Fischer esterification
was observed.
(9) Addition of tetrabutylammonium salts as phase transfer catalysts had
no effect upon reaction converstion.
5352 J. Org. Chem. Vol. 75, No. 15, 2010

Downey et al.
JOCNote
TABLE 2. Aldol Addition of TMSOAc to Various Aldehydes
propionic acid added to 1-naphthaldehyde in 72% yield after
24 h (eq 8). Butyric acid behaved similarly under the same
conditions (67% yield, 24 h, eq 9).10,11 Although it is clear that
steric effects slow the reaction rate of this one-pot silylation-
Mukaiyama aldol process, they may also provide a benefit by
preventing the desired products from undergoing a second aldol
addition (eq 10), which could be a competing pathway under
our conditions.
aConditions C: 1.0 mmol TMSOAc, 1.5 mmol i-Pr2NEt, 1.4 mmol
RCHO, 1.2 mmol TMSOTf, 5.0 mL CH2Cl2, rt. Conditions D: 1.0 mmol
TMSOAc, 1.8 mmol i-Pr2NEt, 1.4 mmol RCHO, 1.5 mmol TMSOTf,
5.0 mL CH2Cl2, rt. bIsolated yield after chromatography.
In summary, this one-pot, three-step reaction allows acetic
acid to be used directly as an enolate precursor in aldol
reactions, providing versatile carboxylic acid products.
Commercially available TMSOAc is a convenient and in-
expensive surrogate. We anticipate the application of this
strategy to other classical enolate reactions, including the
development of asymmetric variants.
is unnecessary for this variant, less than 2 equiv i-Pr2NEt was
necessary to achieve full conversion, in contrast with our
results with acetic acid itself.
Experimental Section
Typical Procedure for the Addition of Acetic Acid to Alde-
hydes. Synthesis of Product 1.3a. To an oven-dried 10-mL round-
bottomed flask under N2 were added CH2Cl2 (5.0 mL), acetic
acid (57 μL, 60 mg, 0.996 mmol), i-Pr2NEt (435 μL, 323 mg,
2.50 mmol), benzaldehyde (142 μL, 148 mg, 1.40 mmol), and
TMSOTf (400 μL, 491 mg, 2.21 mmol). The reaction was stirred
at room temperature, and the colorless solution became pale
yellow. After 2 h, the reaction mixture was added to a 25-mL
round-bottomed flask with 10 mL of EtOH (95%) and 5-10
drops trifluoroacetic acid. The reaction mixture was concen-
trated by rotary evaporation. Flash chromatography (10-50%
EtOAc/hex) afforded the pure product as a white solid (99%):
mp 89-90 °C; IR (film) 3283, 2914, 2648, 1687, 1448, 1394,
Readily available and inexpensive, acetic acid is an ideal
building block for organic synthesis. Its trimethylsilyl ester,
TMSOAc, is also readily available, inexpensive, and reactive
under our conditions. This silyl ester is particularly attractive as
a reaction partner because it requires fewer equivalents of
TMSOTf and i-Pr2NEt to be converted into the reactive bis-
silyl ketene acetal intermediate. Accordingly, we surveyed the
ability of TMSOAc to add to various aldehydes (Table 2). These
results closely parallel the results for acetic acid illustrated in
Table 1. Notably, however, the yield for cinnamaldehyde im-
proved (entry 10), and even p-nitrobenzaldehyde reacted in
moderate yield under these conditions (entry 5).
Finally, we turned to carboxylic acids other than acetic acid.
Phenylacetic acid did not provide aldol products under our
conditions, although bis-silyl ketene acetal formation was
detectable by 1H NMR spectroscopy. Presumably, the phenyl
group reduces the nucleophilicity of the π bond both sterically
and electronically. Similarly, the sterically hindered isobutyric
acid was also unreactive. Initial experiments with propionic acid
were promising but sluggish, providing about 50% conversion
overnight. We speculated that the size of the nascent silyl pro-
pionate made it kinetically difficult to activate with TMSOTf
and/or deprotonate with i-Pr2NEt. When i-Pr2NEt was re-
placed with the less sterically demanding 2,6-lutidine, however,
1269, 1052, 1014, 883, 764, 710 cm-1 1H NMR (500 MHz,
;
acetone) δ 7.47-7.43 (m, 2H), 7.38-7.31 (m, 2H), 7.29-7.24
(m, 1H), 5.17 (dd, J = 5.4, 8.1 Hz, 1H), 2.74 (dd, J = 8.0, 15.5
Hz, 1H), 2.70 (dd, J = 5.3, 15.5 Hz, 1H); 13C NMR (125 MHz,
acetone) δ 172.4, 144.4, 128.2, 127.2, 125.8, 70.1, 43.8; HRMS
(ESI) exact mass calcd for C9H10O3Na [M þ Na]þ 189.0522,
found 189.0517.
(10) The major diastereomer of product 12 is syn, as proven by conversion
to the methyl ester (MeOH, TFA) and comparison to the known 1H NMR
spectrum. Configuration of the major diastereomer of compound 13 is
assigned as anti, by analogy to the 1H NMR spectra for the stereoisomers
of compound 12.
(11) Because of the symmetry of the bis-silyl ketene acetal derived from a
carboxylic acid under these conditions, enolization can only produce a
compound of Z geometry.
J. Org. Chem. Vol. 75, No. 15, 2010 5353

Downey et al.
JOCNote
Acknowledgment. We gratefully acknowledge Research
Corporation and the Thomas F. Jeffress and Kate Miller
Jeffress Memorial Trust for funding. D.H.L. acknowledges
the University of Richmond Department of Chemistry
for a Puryear-Topham fellowship. K.J.T. acknowledges
the University of Richmond College of Arts and Sciences
for a summer research fellowship. We thank Dr. Timothy
Smith and Dr. Diane Kellogg for assistance with mass
spectral data.
Supporting Information Available: Experimental proce-
dures, compound characterization data, and spectra. This
material is available free of charge via the Internet at http://
pubs.acs.org.
5354 J. Org. Chem. Vol. 75, No. 15, 2010