2
C.W. Downey et al. / Tetrahedron Letters xxx (xxxx) xxx
Crossed aldehyde-aldehyde aldol coupling is an enduring prob-
lem in organic synthesis [9,10]. Challenges stem from the fact that
the two aldehyde reactants and the product, which is also an alde-
hyde, are all potent electrophiles, which can lead to various unde-
sired side reactions. In order to achieve a selective and high-
yielding crossed aldol reaction, each of the aldehydes present in
the reaction mixture must fall into a strictly defined role (Fig. 1).
That is, only one aldehyde can serve as the enolate precursor; only
one aldehyde can serve as the electrophile; and the product alde-
hyde must be unreactive under the reaction conditions. Moreover,
dehydration of the aldol products to yield the
bonyl compound must be avoided.
a, b-unsaturated car-
Fig. 2. Effective crossed aldol system.
In the course of our study of the TMSOTf/R3N system and its
ability to promote the addition of indoles to aldehydes [11], we
observed a side reaction wherein isobutyraldehyde was efficiently
converted to its enol silane but did not undergo self-aldol reaction
[12]. When a solution of this in situ-generated enol silane was trea-
ted with benzaldehyde, addition to yield the crossed aldol product
occurred. Further additions to the formyl group in the product
were not observed. Notably, the trial reaction appears to meet all
the criteria for the ideal crossed aldol reaction described in Fig. 1.
In the observed reaction, the non-enolizable aromatic aldehyde
can only play the role of the electrophile (Fig. 2). Only the isobu-
Table 1
Scope for crossed aldehyde-aldehyde aldol addition.
Entry
R1
R2
Product
Yield (%)b
tyraldehyde possesses the
lization, so it must act as the pronucleophile. Finally, the
quaternary center at the carbon in the product renders dehydra-
a hydrogen necessary to achieve eno-
1
2
3
4
5
i-Pr
i-Pr
i-Pr
i-Pr
Cy
Ph
1a
1b
1c
1d
1e
84
75
85
71c
82
4-(CF3)Ph
4-MePh
4-MeOPh
Ph
a
tion impossible, and effectively blocks additional enol silanes from
attacking the adjacent formyl group, preventing unwanted
oligomerization.
a
Conditions: 1. CH2Cl2 (10 mL),
A brief survey of the reaction scope for this crossed aldehyde-
aldehyde coupling reaction is presented in Table 1. Performance
of this reaction in CH2Cl2 at 0 °C was optimal for most of the exam-
ples studied, including when cyclohexanecarboxaldehyde was
employed as pronucleophile. For the electron-rich anisaldehyde,
however, room temperature was required for full conversion, and
the use of 2,6-lutidine instead of Hunig’s base was found to be
slightly superior. In general, the reaction worked well for the addi-
CHO (1.00 mmol), i-Pr2NEt (1.70 mmol), TMSOTf (1.60 mmol),
CHO (1.40 mmol), 0 °C, 2 h. 2. 1.0 N HCl (2 mL), THF (6 mL), rt, 10 min.
b
Isolated yield after chromatography.
Reaction performed at room temperature with 2,6-lutidine (1.70 mmol) instead
c
of i-Pr2NEt.
tion of
but attempted reactions involving additions to aliphatic aldehydes
resulted in no conversion to the aldol products. When the -mono-
substituted propionaldehyde was employed as the enolate precur-
sor, it was rapidly consumed but no desired products were
observed.
a-disubstituted aldehyde enolates to aromatic aldehydes,
with a survey of the reactivity of various thioesters with benzalde-
hyde (Table 2). First, several thioacetates were examined, including
various thiophenol derivatives (entries 1–3). A benzyl mercaptan-
derived thioester also reacted in superior yield (entry 4). A propi-
onate derivative was also studied, providing product 2e as a mix-
ture of diastereomers in high yield (entry 5). Enolization of this
substrate appeared to be significantly less efficient than was the
case with thioacetates, however, which may be attributable to
a
Evidence suggests that
reaction partners for this one-pot enol silane formation-
Mukaiyama aldol reaction, which makes -monosubstituted b-
a-monosubstituted aldehydes are poor
a
hydroxy aldehyde synthons inaccessible through this method.
Indirectly, however, these building blocks are accessible from car-
boxylic acid esters [13] or through the Fukuyama reduction of
thioesters [14]. Accordingly, we targeted thioesters as convenient
pronucleophiles based on their performance in related reactions
under similar conditions. [7] Our study of these substrates began
increased hindrance at the
a carbon. When Hunig’s base was com-
pletely replaced by the less hindered Et3N, conversion also suf-
fered, possibly because of competing attack by the Et3N upon
TMSOTf itself. In order to circumvent these difficulties, a catalytic
amount of Et3N was added to the Hunig’s base in the reaction mix-
ture, which smoothly afforded the desired product.
After the establishment of the thioester scope, various non-eno-
lizable aldehydes were treated with a standard thioester, S-phenyl
thioacetate, to good effect (Table 3). Both electron-rich and elec-
tron-poor benzaldehydes reacted in high yield, as did the sterically
encumbered 2-naphthaldehyde. The sensitive heterocycle 2-
furanaldehyde also proved amenable to the reaction conditions,
and the convenient alkenyl substrate cinnamaldehyde provided
an excellent yield of the desired product.
Earlier work established that carboxylic acid esters are good
substrates for one-pot silyl ketene acetal formation-Mukaiyama
aldol reactions in CH2Cl2. [6] More recently, we became intrigued
by the possibility that a ubiquitous ester like ethyl acetate could
act as both solvent and enolate precursor in these reactions [15].
Fig. 1. Design of a crossed aldehyde-aldehyde aldol addition reaction.
Please cite this article as: C. W. Downey, G. J. Dixon, J. A. Ingersoll et al., One-pot enol silane formation-Mukaiyama aldol reactions: Crossed aldehyde-alde-
Downey, C. Wade
