.
Angewandte
Communications
With catalysts SpinPO 1a–g in hand, we then investigated
their catalytic performance in the direct double-aldol reac-
tion, choosing acetophenone (4a) and benzaldehyde (5a) as
the standard substrates. After screening a variety of reaction
parameters, including base, solvent, temperature, catalyst
loading, and substrate concentration (Tables S1–S5), it was
found that the reaction proceeded smoothly in CH2Cl2 (0.1m)
at À308C in the presence of catalyst (R)-1a (10 mol%) and
dicyclohexylmethylamine (cHex2NMe; 5 equiv) as the base,
to give the corresponding 1,3-diol (À)-6aa in 84% yield with
a diastereomeric ratio (d.r.) of 91:9 (chiro/meso) and 94% ee
for chiro-6aa (Table 1, entry 1). Under the optimized reaction
(entries 2 and 3 vs. 1). The introduction of electron-donating
MeO groups on the aryl rings attached to the P atoms led to
a slight decrease in the enantioselectivity of the reaction
(88% ee ; entry 4). Catalyst (S)-1e with 4-tolyl groups on the
P atoms performed comparably to the prototypical catalyst
(R)-1a in terms of reactivity, diastereoselectivity, and enan-
tioselectivity, affording the corresponding 1,3-diol product
(+)-6aa in 84% yield with a diastereoselectivity of 91:9 and
94% ee (entry 5). The examination of diphosphine oxides 1 f
and 1g indicated that the presence of the methylene bridges
between the spiro[4,4]-1,6-nonadiene skeleton and the diphe-
nylphosphine oxide units of the catalyst is critically important
for the stereoselectivity of the reaction (entries 6 and 7 vs.
entry 1). The use of diphosphine oxides with privileged
skeletons, such as (R)-BINAPO[13] (entries 8 and 9) or (R)-
SDPO[14d] (entry 10), as well as our aromatic spiroketal-based
diphosphine oxide (S,S,S)-SKPO[16b] (entry 11), gave only
inferior results (< 78:22 d.r. and < 70% ee) to those obtained
with SpinPO (entry 1). These results emphasize the advanta-
geous features of the spiro[4,4]-1,6-nonadiene-1,6-bis-
(methyl) skeleton of the bisphosphine oxides for this catalytic
system. The reaction in the presence of a monoxide catalyst
(SKPOꢀ; entry 12), did give the expected product, but only in
low yield (12%), poor d.r. (38:62) and modest ee (46%), in
comparison with the results obtained with SKPO (entry 11).
Although we are unable to clarify the underlying reason for
the outstanding catalytic performance of (R)-1a, it is clear
that fine-tuning of the steric and electronic environment of
the phosphine oxide catalyst is critically important in achiev-
ing maximum asymmetric induction.
Having established (R)-1a as the most effective catalyst,
we subsequently investigated the double-aldol reaction of
various ketones with benzaldehyde (5a) under the optimized
conditions. As shown in Table 2, the reaction of 5a with
aromatic (4b–k), heteroaromatic (4l and 4m), olefinic (4n) or
aliphatic (4o and 4p) ketones proceeded smoothly, giving the
corresponding double-aldol adducts in high yields with good
to excellent diastero- and enantioselectivities. The reaction
tolerates both electron-withdrawing and electron-donating
substituents on the aromatic ring of the ketones; however,
a slightly negative effect was observed in the case of
a sterically demanding 2-methyl (entry 3) or a strongly
electron-withdrawing 4-NO2 group (entry 7). In the reaction
of 4-bromoacetophenone with benzaldehyde, the chiro prod-
uct (À)-6ga was obtained with an ee value of > 99%. With
ketones 4l and 4m, which contain heteroaromatic rings,
remarkable reactivities (87% and 91% yield, respectively)
and enantioselectivities (96% ee) were observed for the
reaction with 5a (entries 11 and 12). Similarly, the reactions of
olefinic and aliphatic ketones afforded the corresponding
adducts in high yields and stereoselectivities (entries 13 and
14). In the double-aldol reaction of butanone 4p, a sterically
less-hindered ketone with two enolizable a-carbon positions,
the two aldol reactions occurred at the two different
a-carbons of the carbonyl group. Thus adduct 6pa was
obtained in good yield with excellent diastereo- and enantio-
selectivity (Scheme 3a), results comparable to those achieved
by Nakajima and co-workers using BINAPO as the cata-
lyst.[13b]
Table 1: Catalyst optimization for the asymmetric double-aldol reaction
of acetophenone (4a) with benzaldehyde (5a).[a]
Entry
Catalyst
Yield[b] [%]
d.r.[c]
ee[d] [%]
1
2
3
4
5
6
7
8
(R)-1a
(R)-1b
(R)-1c
(S)-1d
84
76
60
82
84
45
83
62
86
50
38
12
91:9
94 (S,S)
84 (S,S)
74 (S,S)
88 (R,R)
94 (R,R)
66 (R,R)
57 (R,R)
54 (R,R)
70 (R,R)
34 (R,R)
58 (S,S)
46 (S,S)
86:14
72:28
92:8
91:9
50:50
92:8
75:25
78:22
58:42
52:48
38:62
(S)-1e
(À)-1 f
(S)-1g
(S)-BINAPO
(S)-BINAPO
(R)-SDPO
(S,S,S)-SKPO
(S,S,S)-SKPO’[f]
9[e]
10
11
12
[a] Unless otherwise noted, the reactions were carried out by addition of
silicon tetrachloride (0.8 mmol) to a solution of 4a (0.2 mmol), 5a
(0.44 mmol), cHex2NMe (1.0 mmol), and the catalyst (10 mmol%) in
CH2Cl2 (2 mL) at À308C. [b] Yields of the isolated chiro-diastereomers.
[c] Determined by 1H NMR analysis of the crude reaction mixtures (chiro/
meso). [d] Determined by HPLC analysis on a chiral stationary phase
(AD-H column). The absolute configurations were assigned by com-
paring the optical rotation obtained with values reported in the
literature.[13a] [e] The data is cited from Ref. [13a]. [f] (S,S,S)-SKPO’ is the
monooxide counterpart of (S,S,S)-SKPO.
conditions, a variety of spiro[4,4]-1,6-nonadiene-based chiral
phosphine oxides, including (R)- or (S)-1b–e with different
aryl substituents at the P atoms, were then tested as catalysts
for the double-aldol reaction between 4a and 5a. As shown in
Table 1, the steric hindrance of the aryl groups on the P atoms
has
a significant impact on the asymmetric induction
(entries 1 and 2). With catalysts (R)-1b and (R)-1c bearing
3,5-xylyl and 2-tolyl moieties, respectively, on the P atoms, the
diastereo- and enantioselectivities decreased in comparison
with the results obtained with their structural analogue 1a
2
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 6
These are not the final page numbers!