S. Liu et al. / Tetrahedron 72 (2016) 5115e5120
5117
without compromise of enantiopurity (entry 11 vs entry 1). In ad-
dition to the organic base, the inorganic base could also serve as
effective additive, albeit relatively less reactivity improvement was
detected (entries 12e15). Notably, the involvement of NaOH totally
suppressed this transformation (entry 16).
substituents on the aromatic rings worked properly and furnished
the desired adducts 3abeah in high yields with excellent enan-
tioselectivities (Table 3, entries 2e8). This asymmetric process was
independent of the electronic character of substituted group. Both
electron-deficient cinnamones 1bee and electron-rich analogues
1feh undertook complete transformations and displayed high
degree of enantioselectivities (entries 2e5 vs entries 6e8). At the
same time, the steric hindrance imposed by the substituents lo-
cated on the aromatic rings exerted limited influence on the re-
activity and stereocontrol. The ortho-substituted electrophiles 1e
and 1g were fully converted into target products within 48 h (en-
tries 5 and 7). Compounds 1i and 1j, both possessing a bulky
naphthyl group on the terminal site of double bond, were also
suitable acceptors and afforded the desired adducts with 93% and
99% ee, respectively (entries 9 and 10). The heteroaromatic accep-
tors 1k and 1l were all well tolerated, however, prolonged reaction
time was required probably owing to the electron-rich property of
furanyl and thiophenyl groups (entries 11 and 12). In addition, al-
iphatic enones 1m and 1n were competent acceptors and afforded
the corresponding adducts 3am and 3an with high optical purities
(entries 13 and 14). In line with previous research,13iek these sub-
strates exhibited relatively poorer reactivity and 144 h was neces-
sary for their complete conversions. After a careful investigation of
substituted groups at the end of double bond, we next evaluated
Once indentifying proton-sponge as the optimal additive, we
turned our attention to investigate the effect of acid cocatalyst. A
variety of substituted aromatic acids proved to be efficient co-
catalysts (Table 2, entries 2e5). The model reaction proceeded
smoothly in the presence of these substituted acids and afforded
the desired adduct in 82e93% yields, however, slightly poorer en-
antiomeric excesses were afforded in contrast with benzoic acid
(entries 2e5 vs entry 1). Further study revealed that aliphatic acid
displayed diminished catalytic activity in comparison with aro-
matic acids (entry 6 vs entries 1e5). Notably, the model reaction
almost didn’t occur in the presence of commonly-used strong acids,
including trifluoroacetic acid (TFA), p-toluenesulfonic acid (TsOH)
and trifluoromethanesulfonic acid (TfOH) (entries 7e9). The re-
activity and enantioselectivity slightly decreased when benzoic
acid and 4 was combined in a 1:1 ratio (entry 10). Moreover, the
reactivity was also relatively sensitive to the loading of additive.
Diminished catalytic efficiency was observed when 10 mol % of
proton-sponge was utilized, however, only negligible reactivity
improvement was achieved in the presence of 40 mol % of proton-
sponge (entries 11 and 12). Furthermore, the yield of 3aa decreased
to 80% when the amount of aminocatalyst 4 was reduced to 10 mol
% (entry 13). To our delight, elevating reaction temperature exerted
beneficial effect on catalytic activity. Nearly quantitative desired
adduct was formed almost without loss of enantioselectivity at
40 ꢀC after 48 h (entry 14). After careful investigation, the optimal
reaction conditions were therefore found to be a combination of
the effect of substituent at the a
0-site of enones (entries 15e16). In
general, relatively poorer reactivity was observed for these com-
pounds in comparison with cinnamones bearing a methyl group
adjacent to the carbonyl group (entries 15 and 16 vs entry 1).13j,k
Even after 168 h, less than 80% isolated yields were attained in
the case of ethyl-substituted 1o and isopropyl-substituted 1p (en-
tries 15 and 16). The bulky alkyl group beside the carbonyl group
might retard formation of iminium ion with primary aminocatalyst
4, thereby leading to poor reactivity. To our delight, cyclic enones,
such as 2-cyclohexene-1-one and 2-cyclopenten-1-one, were all
highly active acceptors (entries 17 and 18). Reactions of 1q and 1r
went to completion within 24 h and provided the desired adducts
with good levels of enantioselectivity (96% ee for 1aq, 81% ee for
1ar). Gratifyingly, trace double Michael addition products were
detected for these two cyclic acceptors and the desired mono-
addition adducts were isolated in 98% and 92% yields, re-
spectively. As mentioned by previous approach, double addition
was found to be prone to occur in the case of cyclic enones.13l Be-
sides nitromethane, nitroethane and 2-nitropropane were favor-
ably applicable to this asymmetric process (entries 19 and 20).
When nitroethane reacted with cinnamone 1a, excellent enantio-
selectivities were obtained for both diastereoisomers, but the di-
astereomeric ratio was close to 1:1, presumably due to facile
epimerization of the adduct in the presence of proton-sponge.
More importantly, this Michael addition also occurred efficiently
when the amount of donor was reduced. The model reaction
completed within 96 h in the case of nine equivalent of nitro-
methane and afforded the desired adduct 3aa in satisfactory yield
and excellent enantiomeric excess (entry 21).
To our delight, our organocatalytic protocol was also effective
with chalcones, a class of challenging substrates for iminium ion
activation.3b,12a As demonstrated in Table 4, the Michael reactions
of chalcone 5a and its analogues 5bei proceeded smoothly, gen-
erating the desired adducts 6aei with good to excellent enantio-
selectivities (65e97% ee) (Table 4, entries 1e9). Owning to the
steric hindrance imposed by the aromatic ring adjacent to the
carbonyl group, chalcone and its analogues displayed relatively
lower reactivity in comparison with cinnamones. Nevertheless,
good conversions could be achieved when reaction time was pro-
longed to 168 h. Gratifyingly, both electron-neutral 5a, electron-
rich 5bec, 5f and electron-poor chalcones 5d, 5g delivered the
corresponding adducts in good yields and with satisfactory
cinchona-derived primary amine
4 (20 mol %), benzoic acid
(40 mol %) and proton-sponge (20 mol %) as additive at 40 ꢀC.
Table 2
Optimization of reaction conditions for Michael addition of nitromethane 2a to
cinnamone 1aa
Entry
Additive
Yield (%)b
ee (%)c
1
2
3
4
5
6
7
8
BA
96
91
82
86
93
70
NR
NR
NR
93
80
98
80
99
97
96
96
94
97
95
/
PNBA
ONBA
SA
OFBA
HOAc
TFA
TsOH
TfOH
BA
BA
BA
BA
BA
/
/
9
10d
11e
12f
13g
14h
96
96
96
96
96
a
Unless otherwise noted, the reaction was performed with 0.1 mmol of 1a,
20 mol % 4, 40 mol % benzoic acid and 20 mol % proton-sponge in 1 mL of nitro-
methane at 30 ꢀC for 96 h. BA¼benzoic acid, PNBA¼p-nitrobenzoic acid, ONBA¼o-
nitrobenzoic acid, SA¼salicylic acid, OFBA¼o-fluorobenzoic acid.
b
Isolated yield after flash chromatography on silica gel.
Determined by HPLC on chiral AS-H column.
With 20 mol % benzoic acid.
With 10 mol % proton-sponge.
With 40 mol % proton-sponge.
The reaction was performed with 10 mol % 4, 20 mol % benzoic acid and 20 mol %
c
d
e
f
g
proton-sponge.
h
The reaction was conducted at 40 ꢀC for 48 h.
With the optimal reaction conditions in hand, we subsequently
examined a range of a,b-unsaturated enones and nitroalkanes to
explore the generality of this novel catalytic system. As illustrated
in Table 3, cinnamone derivatives 1beh possessing various