five- and six-membered ring products, along with modest
quantities of simple 1,4-reduction products.
moving to more electron deficient phosphine ligands, 2b is
obtained in greater than 70% yield (Table 1, entries 5 and
6). To ensure the cycloreductions proceed in accordance with
the postulated mechanism, several control experiments were
performed. Exposure of the conjugate reduction product to
the reaction conditions does not produce 2b. Conversely, 2b
does not undergo retro-aldolization upon exposure to the
reaction conditions. Enal 2a is unreactive toward triarylphos-
phine addition, excluding tandem Morita-Baylis-Hillman
cyclization-conjugate reduction pathways. Finally, upon
omission of hydrogen, no reaction is observed. The structural
assignment of 2b was corroborated by single-crystal X-ray
diffraction analysis of the corresponding carboxylic acid
(Table 1).
Under these optimized conditions, the scope of the
catalytic aldol cycloreduction of keto-enals was explored.
As demonstrated by the catalytic cycloreduction of substrates
1a and 2a, five-membered-ring formation proceeds well to
provide the bicyclic aldol products 1b and 2b in 72% and
73% yield, respectively. As illustrated by substrates 3a and
4a, cyclization to form six-membered rings occurs in slightly
diminished yield due to increasing levels of conjugate
reduction. Here, aldol products 3b and 4b are produced in
63% and 59% yields, respectively. The structural assignment
of 4b was corroborated by single-crystal X-ray diffraction
analysis (Scheme 1).
The primary issues limiting the utility of aldehyde enolates
in cross-aldolizations with ketone partners involve polyal-
dolization along with a diminished thermodynamic driving
force.8 It was recognized that intramolecular aldolization
should attenuate polyaldolization. Additionally, as aldoliza-
tion is primarily driven by chelation,9 intramolecular al-
dolization should also favorably bias the enolate-aldolate
equilibria. Predicated on this analysis, catalytic hydrogena-
tion-aldolization of keto-enal 2a was attempted. Exposure
of a solution of keto-enal 2a in dichloroethane (DCE) at 40
°C to RhI(COD)2OTf under an atmosphere of hydrogen (1
atm) in the presence of triphenylphosphine and potassium
acetate gave the aldol product 2b in 23% yield, accompanied
by a 50% yield of the product of simple conjugate reduction
(Table 1, entry 1). The conjugate reduction manifold should
Table 1. Optimization of the Catalytic Aldol Cycloreduction of
Keto-Enal 1aa
solvent
(concn)
yieldb
Scheme 1. Catalytic Aldol Cycloreduction of Keto-Enals
1a-4a
entry
ligand
Ph3P
Ph3P
Ph3P
Ph3P
(p-CF3Ph)3P
(2-furyl)3P
additive
(1,4-reduction)
1
2
3
4
5
6
KOAc
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
DCE (0.1 M)
DCE (0.1 M)
DCE (0.05 M)
THF (0.05 M)
THF (0.05 M)
THF (0.05 M)
23% (50%)
40% (28%)
59% (29%)
65% (32%)
73% (22%)
73% (21%)
a Procedure: Toa25mLround-bottomedflaskchargedwithRh(COD)2OTf
(24 mg, 0.052 mmol, 10 mol %) and ligand (0.12 mmol, 24 mol %) was
added solvent. The mixture was stirred for 10 min under an argon
atmosphere, at which point 2a (100 mg, 0.52 mmol, 100 mol %) and base
(0.52 mmol, 100 mol %) were added. The system was purged with hydrogen
gas and the reaction was allowed to stir at 40 °C under 1 atm of hydrogen
until complete consumption of substrate. b Isolated yields after purification
by silica gel chromatography.
be attenuated by base-assisted entry into the monohydride
catalytic cycle (vide supra). Accordingly, substituting potas-
sium carbonate for potassium acetate, the yield of 2b is
increased to 40% (Table 1, entry 2). Reactions performed at
higher dilution provide 2b in 59% yield (Table 1, entry 3).
Under otherwise identical conditions, but in THF solvent,
the yield of 2b is increased to 65% (Table 1, entry 4). Finally,
To further explore the scope of this new catalytic variant
of the aldol reaction, a range of other substrates were
subjected to conditions for hydrogenation-aldolization. As
demonstrated by the cycloreduction of the indanedione
containing substrates 5a and 6a, aromatic ketones are viable
electrophilic partners. Keto-enals 7a and 8a highlight the
chemoselectivity of aldolization with regard to the use of
nonequivalent ketone acceptors. For 7a, steric bias in the
form of geminal dimethyl substitution induces addition to
the less encumbered ketone partner. In the case of 8a,
addition to the ketone occurs smoothly in the presence of
(8) As reported in ref 7, ab initio calculation (RHF/6-31G) revealed that
∆Hf for the â-hydroxyaldehyde (CH3)2C(OH)CH2CHO derived from
acetaldehyde and acetone is +21.155 kcal/mol, while ∆Hf for the isomeric
â-hydroxyketone CH3CH(OH)CH2COCH3 also derived from acetaldehyde
and acetone is -10.455 kcal/mol.
(9) The failure of tris(dialkylamino)sulfonium enolates to react with
aldehydes is attributed to unfavorable enolate-aldolate equilibria: Noyori,
R.; Nishida, I.; Sakata, J.; Nishizawa, M. J. Am. Chem. Soc. 1980, 102,
1223.
692
Org. Lett., Vol. 6, No. 5, 2004