Rational design of sterically and electronically easily tunable chiral bisimidazolines and their applications in dual lewis acid/brensted base catalysis for highly enantioselective nitroaldol (Henry) reactions

Article abstract of DOI:10.1002/chem.200601220

A new addition to the rational design of sterically and electrically easily tunable chiral bis(imidazoline) ligands from chiral amino alcohols has been developed. Vast structural variation of chiral bis(imidazoline) ligands can be simply achieved by the choice of both the 1,2-amino alcohol and its N-1 R ¹ substituent. A small library of chiral bisimidazolines (1a-h) has been constructed. The method has provided an easy and simplified route to a diverse set of air-stable and water-tolerant chiral bis(imidazoline) ligands on 10 g scales. The dual Lewis Acid/Bronsted base catalytic system generated from the (5)-1a/Cu(OTf)₂ complex and Et3N was able to catalyze Henry reactions between aldehydes and nitromethane effectively at room temperature, and also to tolerate a wide scope of aldehydes with excellent enantiomeric excesses. Not only aromatic aldehydes but also aliphatic aldehydes afforded the nitroalcohol products, with enantiomeric excesses in the 93-98% range. This dual catalytic system is among the most effective systems so far reported for the asymmetric parent Henry reactions. This work also represents the first members of the class of chiral bisimidazolines to have been demonstrated to achieve excellent enantioselectivities.

Full text of DOI:10.1002/chem.200601220

FULL PAPER
DOI: 10.1002/chem.200601220
Rational Design of Sterically and Electronically Easily Tunable Chiral
Bisimidazolines and Their Applications in Dual Lewis Acid/Brønsted Base
Catalysis for Highly Enantioselective Nitroaldol (Henry) Reactions
ACHTREUNG
[
a]
Kuoyan Ma and Jingsong You*
Abstract: A new addition to the ration-
al design of sterically and electrically
easily tunable chiral bis(imidazoline) li-
ant chiral bis(imidazoline) ligands on
10 g scales. The dual Lewis Acid/
Brønsted base catalytic system generat-
of aldehydes with excellent enantio-
meric excesses. Not only aromatic alde-
hydes but also aliphatic aldehydes af-
forded the nitroalcohol products, with
enantiomeric excesses in the 93–98%
range. This dual catalytic system is
among the most effective systems so
far reported for the asymmetric parent
Henry reactions. This work also repre-
sents the first members of the class of
chiral bisimidazolines to have been
demonstrated to achieve excellent
enantioselectivities.
gands from chiral amino alcohols has
been developed. Vast structural varia-
tion of chiral bis(imidazoline) ligands
can be simply achieved by the choice
of both the 1,2-amino alcohol and its
ed from the (S)-1a/Cu(OTf) complex
A
H
R
U
G
2
and Et N was able to catalyze Henry
3
reactions between aldehydes and nitro-
methane effectively at room tempera-
ture, and also to tolerate a wide scope
1
N-1 R substituent. A small library of
chiral bisimidazolines (1a–h) has been
constructed. The method has provided
an easy and simplified route to a di-
verse set of air-stable and water-toler-
Keywords: asymmetric catalysis
·
dual acid/base catalysis · Henry
reaction · ligand design · N ligands
Introduction
great number of reactions with unparalleled enantioselectiv-
ity.
[2]
Asymmetric catalysis is undoubtedly a powerful and eco-
nomically viable tool for the synthesis of optically active or-
ganic compounds, and the design and synthesis of new and
improved chiral ligands and their application in asymmetric
catalysis continue to be research foci in both academic and
industrial laboratories. The ideal ligand should be easily pre-
pared, cheap, stable under the reaction conditions, and very
selective for a large number of processes, hence very flexi-
ble. It has not yet been discovered (and probably never
will be), but bis(oxazoline) ligands have proved to be a priv-
ileged class of chiral ligands, being capable of forming a
broad variety of metal complexes that are able to catalyze a
It is reasonable to assume that the oxazoline ring should
be structurally modifiable by the replacement of the O atom
with a substituted N atom to provide a new class of imidazo-
line ligands, which should maintain the basic framework of
oxazoline but also be further tunable electronically and con-
formationally over a wide range through judicious selection
1
of the substituted R group at the amine N atom (Figure 1).
The highly modular nature of the construction of these li-
gands should therefore easily allow better matching of chiral
ligand, metal ion, and substrate than are attainable with the
analogous oxazoline ligands, and thus facilitate the achieve-
ment of excellent catalytic performance (activity and enan-
tioselectivity). Surprisingly, however, imidazolines have not
[
1]
[
a] K. Ma, Prof. J. You
Key Laboratory of Green Chemistry and Technology of Ministry of
Education College of Chemistry
and State Key Laboratory of Biotherapy
West China Hospital, West China Medical School
Sichuan University, 29 Wangjiang Road, Chengdu 610064 (P.R.
China)
Fax : (+86)28-8541-2203
E-mail: jsyou@scu.edu.cn
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
Figure 1. Chiral bis(imidazoline) ligands.
Chem. Eur. J. 2007, 13, 1863 – 1871
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1863

[3]
yet been applied as ligands in catalysis to any great extent.
Results and Discussion
More surprisingly, as far as we know, only a very few exam-
ples describing asymmetric catalysis versions have appeared
Although several different approaches to imidazoline rings
[4]
[14]
in the literature to date. Can this class of chiral imidazo-
lines serve as efficient and useful ligands in asymmetric cat-
alysis? This possibility prompted us to explore the synthesis
of a series of chiral bis(imidazoline) ligands and to evaluate
their scope as ligands in asymmetric reactions.
Nitroaldol (Henry) reactions
are conceivable,
are generally prepared from chiral 1,2-diamines.
enantiopure 4-substituted imidazolines
[4,15,16]
Very
recently, Beller and his co-workers reported the first synthe-
sis of chiral pyridinebisimidazolines derived from pyridine-
[17]
2,6-dinitrile and chiral diamines (Scheme 1), the obtained
between nitroalkanes and car-
bonyl compounds are important
carbonyl addition processes
that afford an atom-economic
approach to a-hydroxynitroal-
kanes, which serve as valuable
synthetic intermediates. It is
therefore not surprising that
great efforts have recently been
directed towards the develop-
ment of a catalytic asymmetric
version of this reaction. In com-
parison with their closely relat-
ed catalytic asymmetric aldol
reaction counterparts, however, Scheme 1. Bellerꢁs synthetic route to chiral bis(imidazoline) ligands.[17]
very few examples of efficient
catalytic enantioselective Henry
[
5–6]
reactions are known to date,
ligands being used in ruthenium-catalyzed asymmetric epox-
idations with hydrogen peroxide as oxidant, albeit with only
moderate ee values (up to 71% ee). Notably, the methodolo-
gy based on chiral 1,2-diamines somewhat restricts conven-
and no significant success had been achieved until the cur-
rent contributions to this area illustrated by the recent stud-
ies by the groups of Shibasaki, Trost, Evans, and
Palomo. Typically, among the most outstanding examples,
an aldehyde is treated with a nitroalkane (usually nitrome-
thane) in the presence of a chiral metal complex and other
additives to produce nitroalcohols with good to excellent op-
tical purities. Some limitations have normally been encoun-
tered, however: a high catalyst loading (30 mol%) may be
[7]
[8]
[9]
[10]
2
ient variation of the C-4-substituent R groups on the imid-
A
H
R
U
G
commercially available in the forms of enantiomerically
pure building blocks is currently not wide. Moreover, steric
and electronic circumstances around the C-4-position should
be one of the crucial factors governing the enantioselectivity
and activity in the catalytic asymmetric reaction. In other
words, further steric and electronic regulation solely through
remote functionalization at the N-1 position is difficult. Fur-
thermore, variations in the substituent at the N-1 position
on the imidazoline ring might be easily carried out by N-al-
kylation and N-acylation, whereas analogous N-aryl groups
were not introduced at the N-1 position.
[10]
required, for example, or the substrate scope may be lim-
[
6d–e,7–8]
ited to aromatic or aliphatic aldehydes,
or a lower
Environmentally friend-
ly organocatalysts have notably begun to surface in the liter-
[6d,7–8,10]
temperature may be needed.
[11]
ature, but these methods have also featured some impor-
tant restrictions in terms of substrate scope, selectivity, and/
or the need for a low reaction temperature. Very recently,
two excellent examples of highly enantioselective Henry re-
actions between nitromethane and aromatic aldehydes or
relatively active a-keto esters in the presence of modified cin-
In contrast, our strategy to develop enantiopure bis(im
A
C
H
T
R
E
U
N
G
id-
[
12]
ACHTREUNG
azoline) ligands from chiral amino alcohols is based on prac-
tical considerations, because the ease with which natural a-
amino acids, the least expensive chiral starting materials,
can be transformed into 1,2-amino alcohols opens up a wide
variety of building blocks available in the chiral pool, so a
vast structural variety of chiral bis(imidazoline) ligands can
be simply achieved through the choice of both 1,2-amino al-
[13]
chona alkaloids have appeared.
Because Henry reactions are still among the more chal-
lenging enantioselective reactions, and controlling the ste-
reochemistry is still difficult, a catalyst that would overcome
those limitations discussed above would be advantageous.
We therefore chose to focus initial studies on evaluation of
the behavior of chiral bis(imidazoline) ligands in asymmetric
Henry reactions and wish to describe here our results in the
application of new chiral bis(imidazoline) ligands for highly
enantioselective parent Henry reactions with a broad range
of substrates.
1
cohol and the N-1 R substituent.
To illustrate our concept, a small library of chiral bisim-
[18]
A
H
R
U
G
As shown in
Scheme 2, the synthesis of bisimidazolines 1 is a straightfor-
ward process, starting from the bis-amido alcohols 3, which
[19]
are easily prepared from readily available amino alcohols.
1864
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1863 – 1871

Easily Tunable Chiral Bisimidazolines in Enantioselective Henry Reactions
FULL PAPER
Henry reaction between nitro-
methane and benzaldehyde, but
ethanol was found to be the
best one. In addition to solvent
effects, we also tried to opti-
mize other reaction conditions
such as catalyst loading, ratio of
ligand to Cu(OTf) , reaction
A
H
R
U
G
2
time, and temperature, but
these factors had little effect on
improving of yields. Attempts
to raise the product yields
through the use of different
copper salts in the presence of
(
S)-1a in ethanol were also un-
successful: Cu(OAc) ·H O and
Cu(ClO ) were capable of af-
fording good yields (84% and
0%, respectively), but very
low ee values (12% and 20%,
AHCTREUNG
2
2
AHCTREUNG
4
2
9
respectively),
whereas
Cu-
A
H
R
U
G
3
2
2
2
2
provided very low yields (less
than 10%).
Scheme 2. Synthetic route to chiral bis(imidazoline) ligands.
Treatment of bis-amido alcohols 3 with SOCl and/or PCl5
Table 1. The effect of the quantity of Et
3
N on the enantioselective Henry
2
reaction between benzaldehyde and nitromethane catalyzed by (S)-1a/
ACHTUNG-
afforded chloroethyl imidoyl chlorides 4, and this was fol-
lowed by two chloride displacements with amines or ani-
lines. A wide variety of sterically and electronically different
ligands could thus be prepared by structural variation of the
bridging group and the imidazoline ring. It is noteworthy
that the preparations of chiral bisimidazolines have been
performed without problems on 10 g scales. These com-
pounds were isolated in analytically pure form as white or
pale yellow solids in moderate to good yields, depending on
the natures of the substituents on the imidazoline rings and
the bridging groups. These ligands are stable enough in air
to be purified by means of column chromatography on silica
gel without special precautions against water or air.
[
a]
T
E
N
2
(OTf) .
Entry
(S)-1a/
A
H
R
U
G
2
Et
3
N
Yield
[%]
ee
[%]
[
b]
[c]
A
H
R
U
G
A
H
R
U
1
2
–
–
–
10
–
2
5
10
20
50
100
trace
98
40
43
52
94
76
76
95
–
–
[
d]
3
4
5
6
7
8
9
10
10
10
10
10
10
10
97
95
95
96
83
74
53
Our initial exploration of catalytic asymmetric Henry re-
actions focused on couplings between aldehydes and nitro-
methane in the presence of bis(imidazoline) ligands in com-
[
(
a] Reactions were performed on
a
0.5 mmol scale with Cu
10 mol%) and ligand (S)-1a (10.5 mol%) in the presence of Et N with
at room tempera-
A
C
H
T
R
E
U
N
G
(OTf)
2
3
bination with Cu(OTf) (OTf: trifluoromethanesulfonyl).
A
C
H
T
R
E
U
N
G
2
nitromethane (10 equiv) in ethanol (1.5 mL) under N
2
Unfortunately, though, the reactions were often low-yield-
ing: benzaldehyde, for example, was treated with nitrome-
ture for 24 h. [b] Yield of isolated product based on aldehyde after chro-
matographic purification. [c] Enantiomeric excess was determined by
HPLC on Chiracel OD-H. [d] This reaction was run for four days under
identical conditions.
thane in the presence of the (S)-1a/Cu
generated in situ from 10.5 mol% of (S)-1a and 10 mol% of
Cu(OTf) in ethanol at room temperature—to afford the
A
H
R
U
G
ACHTREUNG
2
nitro alcohol product in only 40% yield even after four
days, but, excitingly, with an excellent enantioselectivity of
up to 97% ee (Table 1, entry 3). Efforts to optimize asym-
metric Henry reactions to improve the product yields under
these conditions proved to be frustrating. For example, a
series of solvents (e.g., methanol, ethanol, isopropanol, n-
butanol, nitromethane, THF, dichloromethane, DMF, and
DMSO) were first tested in the catalytic enantioselective
Recently, the concept of dual acid/base catalysis has been
[20–21]
introduced for asymmetric catalytic reactions.
Base-cat-
alyzed, nonselective Henry reactions are long known, so we
reasoned that Et N might meet the requirement as a Brønst-
3
ed base to facilitate the deprotonation of nitromethane as a
prelude to the aldol addition process to improve the yields
of enantioselective reactions, as we had observed that benz-
A
H
R
U
G
Chem. Eur. J. 2007, 13, 1863 – 1871
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1865

J. You and K. Ma
the assistance of 10 mol% of Et N at room temperature
Table 2. Enantioselective Henry reaction between benzaldehyde and ni-
3
tromethane catalyzed by (S)-1a/Cu
A
H
R
U
G
2
(10 mol%) and base
with the formation of the racemic nitroaldol product in over
[
a]
[6a,b,9–10]
(10 mol%).
98% yield (Table 1, entry 2).
As expected, the yields
of nitroaldol products were improved dramatically when
Et N was added to generate a binary Lewis acid/Brønsted
3
base catalytic system, although the quantity of Et N was cru-
cial to the outcome of the reaction, the effect of the chiral
Lewis acid/Brønsted base ratio being described in Table 1,
3
[
b]
[c]
Entry
Base
Et
pyridine
DBU
Yield [%]
ee [%]
1
2
3
4
5
6
7
3
N
94
5
54
80
6
96
77
86
86
22
28
22
which shows a clear maximum at 10 mol% of Et N, indicat-
3
ing that equimolar amounts of chiral Lewis acid and Et N
3
A
H
R
U
G
2
Et(iPr) N
constitute the best catalytic system in terms of both enantio-
selectivity and yield (Table 1, entry 6). This shows that our
optimized catalytic system significantly overcomes troubles
potentially arising from the chemical incompatibility of
1-methyl-1H-imidazole
DMAP
K CO
2 3
52
82
[a] Reactions were performed on
10 mol%) and ligand (S)-1a (10.5 mol%) in the presence of base with
use of nitromethane (10 equiv) in ethanol (1.5 mL) under N at room
a
0.5 mmol scale with Cu
A
C
H
T
R
E
U
N
G
(OTf)
2
(
Lewis acids and Brøsted bases and suppresses the Et N-initi-
3
2
ated background reaction. When Et N is in excess in rela-
3
2
temperature for 24 h. [b] Yield of isolated product based on aldehyde
after chromatographic purification. [c] Enantiomeric excess was deter-
mined by HPLC on Chiracel OD-H.
+
tion to the chiral Lewis acid, the Cu center is likely to be
coordinated by Et N to give the inactive Cu(Et N)(OTf) -bi-
A
H
R
U
G
ACHTREUNG
3
3
2
simidazoline complex, hence trapping the chiral Lewis acid
and resulting both in diminished yields and in poorer enan-
tioselectivities. Moreover, the remaining uncoordinated
bisimidazoline complex. However, the enantioselectivity and
yield were significantly reduced in relation to those achieved
Et N further promotes the racemic reaction path, affording
3
only racemic nitroaldol product, so the combination of the
with Et N (Table 2, entry 4). Furthermore, the bulkier and
3
Cu
A
C
H
T
R
E
U
N
G
(OTf) /1a complex (10 mol%) and Et N (100 mol%)
more basic DBU afforded a lower yield of the nitroaldol
2
3
gave a 95% yield but only 53% ee (Table 1, entry 9).
adduct than obtained with Et(iPr )N, with similar enantiose-
A
H
R
U
G
2
The proposed transition state model is consistent with ex-
perimental observations and accounts for the absolute con-
figurations of some selected products. A complex that simul-
taneously binds the two reaction partners should position
the nucleophile perpendicular to the ligand plane, while the
electrophile, for maximum activation, should be positioned
in one of the more Lewis acidic equatorial sites in the ligand
plane, which is in accord with previously reported steric and
lectivity (Table 2, entry 3). Pyridine, DMAP, and 1-methyl-
1H-imidazole were also tested as bases, but lower yields of
nitroaldol product were again formed, probably as a result
of strong coordination of pyridine and imidazole to the
copper complex (Table 2, entries 2, 5, and 6). The inorganic
base K CO was a good base in terms of yield, but the enan-
2
3
tioselectivity was disappointing (Table 2, entry 7).
In addition, a series of reaction solvents (e.g., methanol,
ethanol, isopropanol, n-butanol, nitromethane, THF, tolu-
ene, dichloromethane, acetonitrile, diethyl ether, DMF, and
DMSO) were also tested in the catalytic enantioselective
Henry reaction between benzaldehyde and nitromethane in
[6b,9,10]
electronic considerations (Figure 2).
A series of Brønsted bases were also tested in the reaction
combination with ligand (S)-1a, Cu
A
H
R
U
G
(
Table 3). The protonic solvents were found to be superior
to the nonprotonic ones, and of the different solvents tested,
ethanol was clearly the best choice for this reaction, with
9
4% yield and 96% ee.
Table 4 illustrates the impact of a variety of reaction pa-
rameters on the course of the asymmetric catalysis process.
A series of copper salts were first screened in combination
with (S)-1a and Et N in ethanol. In each instance the reac-
tion was performed at ambient temperature over 24 h
3
Figure 2. Proposed transition state model for the Henry reaction.
(
Table 4, entries 1, 4–8), and as the data collected in Table 4
show, the enantioselectivity was dependent on the counter-
ions. Cu(OTf) proved to be the best choice, affording the
nitroaldol adduct both in high yield and with an excellent ee
value. Although Cu(NO ) ·3H O and Cu(ClO ) ·6H O,
when used in place of Cu
between benzaldehyde and nitromethane in the presence of
1
0 mol% of (S)-1a/Cu
shown in Table 2. After screening a variety of bases (e.g.,
Et N, pyridine, DBU, Et(iPr) N, 1-methyl-1H-imidazole,
A
H
R
U
G
ACHTREUNG
2
2
A
H
E
N
ACHTREUNG
3
3
2
2
4
2
2
DMAP, and K CO ), we found that Et N gave the best
A
H
R
U
G
2
3
3
result (Table 2, entry 1). Et
A
H
R
U
G
selective results (94% and 94% ee, respectively), the reac-
2
Et N but has the same basicity, and it was expected that the
3
tions were notably sluggish (Table 4, entries 6 and 8). In
bulkiness of the base would retard coordination to the chiral
contrast, Cu
A
H
R
U
G
Lewis acid, reducing levels of the inactive Cu
A
H
R
U
G
A
T
E
N
providing very high yields, but the enantiomeric excesses
3
1866
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1863 – 1871

Easily Tunable Chiral Bisimidazolines in Enantioselective Henry Reactions
FULL PAPER
Table 3. Effect of solvents on the enantioselective Henry reaction be-
ing the tridentate valine-derived (S)-1a–c, we found that the
[
a]
A
H
R
U
G
2
tween benzaldehyde and nitromethane catalyzed by (S)-1a/Cu(OTf) .
enantioselectivities and reactivities of the ligands were
1
largely dependent on the R substituents at the N-1 position
of the imidazoline ring. Our results indicate that the catalyt-
ic performances of chiral imidazolines with N-aryl groups at
their N-1 positions may be superior to those obtained with
N-alkyl groups (Table 4, entries 1, 9, and 10). The (S)-1a/Cu-
[
b]
[c]
Entry
Solvent
Yield [%]
ee [%]
1
2
3
4
5
6
7
8
9
methanol
ethanol
2-propanol
n-butanol
nitromethane
THF
60
94
56
50
85
24
20
34
25
40
–
92
96
92
95
91
80
81
78
81
90
–
A
H
R
U
G
and with 96% ee. Notably, with chiral pyridinebisimidazo-
lines derived from pyridine-2,6-dinitrile and chiral diamines
as reported by Beller and his co-workers, analogous N-aryl
groups could not easily be introduced at the N-1 position in
toluene
[17]
2
dichloromethane
acetonitrile
diethyl ether
DMF
the imidazoline ring. On the other hand, the R substitu-
ents at the C-4-positions in the imidazoline rings have a dis-
tinct influence on catalytic performance (Table 4, entries 1,
1
1
1
0
1
2
1
1, and 12). Excellent enantioselectivities could be obtained
DMSO
–
–
with the phenylalanine-derived (S)-1d or the phenylglycine-
derived (R)-1e, but the lower yields were only in the 20–
30% range. These are clear validations of our strategy, in
that the catalytic performances of chiral bisimidazolines
may be sterically and electrically easily tuned both by
remote functionalizations at their N-1 positions and through
the associated steric and electronic conditions at their C-4-
positions. Additionally, the bidentate ligands 1 f, 1g, or 1h
gave lower yields and enantioselectivities of nitroalcohol
product (Table 4, entries 13–15), this observation clearly in-
dicating that the stereochemical course of this Henry reac-
tion depends not only on the stereo- and electronic situation
of the bisimidazoline backbone but also on that of the bridg-
ing scaffold. Of the eight chiral bisimidazolines with the il-
lustrated absolute configurations, the tridentate valine-de-
rived (S)-1a with a phenyl group at the N-1 position proved
to be the ligand of choice. In contrast, excellent enantiose-
lectivity could in most cases be achieved only when tert-
butyl bisoxazoline derived from high-priced, non-natural
tert-leucine or from indabox—prepared from the quite ex-
[
(
a] Reactions were performed on
10 mol%) and ligand (S)-1a (10.5 mol%) in the presence of Et
use of nitromethane (10 equiv) in solvent (1.5 mL) under N at room
a
0.5 mmol scale with Cu
A
H
R
U
G
3
N with
2
temperature for 24 h. [b] Yield of isolated product based on aldehyde
after chromatographic purification. [c] Enantiomeric excess was deter-
mined by HPLC on Chiracel OD-H.
Table 4. Some representative results from the screening of reaction con-
ditions for the catalytic enantioselective Henry reaction between benzal-
dehyde and nitromethane in the presence of Et
[
a]
3
N (10 mol%) as base.
2
+
[b]
[c]
Entry
1
1/Cu (10 mol%)
Yield [%]
ee [%]
(S)-1a/Cu
(S)-1a/Cu(OTf)
(S)-1a/Cu(OTf)
(S)-1a/Cu
(S)-1a/Cu(ClO
(S)-1a/Cu(NO
(S)-1a/CuCl ·2H
(S)-1a/Cu(ClO
(S)-1b/Cu(OTf)
(S)-1c/Cu(OTf)
(S)-1d/Cu(OTf)
(R)-1e/Cu(OTf)
(S)-1 f/Cu(OTf)
(S)-1g/Cu(OTf)
(R)-1h/Cu(OTf)
A
C
H
T
R
E
U
N
G
(OTf)
2
2
2
94
91
46
98
95
51
94
50
64
85
22
23
48
34
46
96
98
91
11
2
94
43
94
58
93
91
À91
À8
À14
7
[
d]
2
3
4
5
6
7
8
9
1
1
1
1
1
1
ACHTREUNG
[
e]
ACHTREUNG
A
C
H
T
R
E
U
N
G
2 2
(OAc) ·H O
A
C
H
T
R
E
U
N
G
4 2
)
A
C
H
T
R
E
U
N
G
3
)
2
·3H
2
O
2
O
[6a–b,9,22]
pensive cis-1-amino-2-indanol—was used.
2
[7–10]
In the most outstanding examples,
the reaction tem-
A
C
H
T
R
E
U
N
G
4 2 2
) ·6H O
A
C
H
T
R
E
U
N
G
2
perature typically played a significant role in determining
the ee values of the nitroalcohol products. Lowering the
temperature of the reaction could significantly increase the
ee value, and the catalytic systems reported by Shibasaki,
Trost, and Palomo and their co-workers even needed much
lower temperatures (À30/À788C) to achieve high enantiose-
0
1
2
3
4
5
A
C
H
T
R
E
U
N
G
2
A
C
H
T
R
E
U
N
G
2
A
C
H
T
R
E
U
N
G
2
A
C
H
T
R
E
U
N
G
2
A
C
H
T
R
E
U
N
G
2
A
C
H
T
R
E
U
N
G
2
[
7–8,10]
lectivities.
In sharp contrast, in our catalytic system
(OTf) , and Et N the enantiose-
[
(
[
a] Reactions were performed on a 0.5 mmol scale with nitromethane
10 equiv) in ethanol (1.5 mL) under N at room temperature for 24 h.
b] Yield of isolated product based on aldehyde after chromatographic
2
generated from (S)-1a, Cu
ACHTREUNG
2
3
lectivity was not very sensitive to the temperature, although
the temperature could significantly effect the yield of nitro-
aldol adduct. When carried out at 08C, for example, the re-
action afforded the corresponding aldol adduct in 91% yield
and with 98% ee within 24 h (Table 4, entry 2), whereas
when the same reaction was performed at 408C, a large rate
acceleration was observed but with an accompanying in-
crease in elimination byproduct, resulting in a drop in the
yield of the corresponding nitroaldol adduct to 46%, though
without any significant loss of enantioselectivity (Table 4,
entry 3).
purification. [c] Enantiomeric excess was determined by HPLC on Chira-
cel OD-H. [d] The reaction was performed at 08C. [e] The reaction was
performed at 408C.
were significantly lower (Table 4, entries 4 and 7). Notably,
Cu(ClO ) gave a very high yield, but of almost racemic
product (Table 4, entry 5).
A
C
H
T
R
E
U
N
G
4 2
The results from the ligand survey in combination with
Cu(OTf) are also summarized in Table 4. First, after screen-
A
C
H
T
R
E
U
N
G
2
Chem. Eur. J. 2007, 13, 1863 – 1871
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1867

J. You and K. Ma
With optimized conditions now to hand, the scope of the
catalytic enantioselective Henry reaction was demonstrated
by treatment of aldehydes with nitromethane in the pres-
Table 6. Range of aldehydes used in Henry reactions with nitromethane
[
a]
in the presence of (S)-1a/Cu
A
H
R
U
G
2
(2 mol%) either under N
2
or under
[
b]
air.
ence of 10 mol% of the (S)-1a/Cu
A
H
R
U
G
1
0 mol% of Et N in ethanol at room temperature for 24 h.
3
We were delighted to find that our catalytic system was able
to tolerate a wide scope of aldehydes. Not only aromatic alde-
hydes but also aliphatic aldehydes afforded the nitroalcohol
products with enantiomeric excesses in the 93–98% range. As
is evident in Table 5, regardless of whether the aromatic al-
dehydes were electron-rich (Table 5, entries 3 and 7), elec-
[c]
[d]
Entry
Aldehyde (R)
Ph
Yield [%]
ee [%]
[a]
1
80
82
63
53
61
78
80
65
56
59
94
91
90
97
90
95
91
94
95
91
[
[
[
[
[
[
[
[
a]
a]
a]
a]
b]
b]
b]
b]
2
3
4
5
6
7
8
9
1
4-ClC
1-naphthyl
2-MeOC H
4
6 4
H
6
cyclohexyl
Ph
4-ClC
1-naphthyl
2-MeOC
cyclohexyl
6
H
4
Table 5. Range of aldehydes used in Henry reactions with nitromethane
[
a]
A
H
R
U
G
2
in the presence of (S)-1a/Cu(OTf) (10 mol%).
6
H
4
[
b]
0
[
(
a] Reactions were performed on
2 mol%), ligand (S)-1a (2.1 mol%), and Et
tromethane (10 equiv) in ethanol (3.0 mL) under N
a
1.0 mmol scale with Cu
N (2 mol%) with use of ni-
at room temperature
A
C
H
T
R
E
U
N
G
(OTf)
2
3
[
b]
[c]
2
Entry
Aldehyde (R)
Yield [%]
ee [%]
for 24 h. [b] Reactions run under air at room temperature for 24 h.
[c] Yield of isolated product based on aldehyde after chromatographic
purification. [d] Enantiomeric excess was determined by HPLC on Chira-
cel OD-H or AD-H columns. The absolute configurations of nitroaldol
adducts were assigned by comparison with literature compounds.
1
2
3
4
5
6
7
8
9
Ph
2-MeC
94
78
65
92
74
60
45
82
76
95
91
88
85
98
96
97
96
95
97
95
94
98
97
93
96
96
94
95
6
H
4
2-MeOC
2-ClC
1-naphthyl
4-MeC
4-MeOC
4-ClC
4-FC
PhCH
iPr
iBu
6
H
4
4
6
H
4
6
H
4
6
H
6
H
4
range, although slightly low yields of products were ob-
served.
In addition, we were delighted to find that the (S)-1a/Cu-
6
H
4
10
11
12
13
14
2 2
CH
A
H
R
U
G
nBu
cyclohexyl
we routinely conduct the reactions under inert atmospheres,
we have determined that the reactions are not highly
oxygen- or moisture-sensitive; under identical conditions,
[
(
a] Reactions were performed on
10 mol%), ligand (S)-1a (10.5 mol%), and Et
nitromethane (10 equiv) in ethanol (1.5 mL) under N
a
0.5 mmol scale with Cu
N (10 mol%) with use of
at room tempera-
A
H
R
U
G
3
even with 2 mol% of (S)-1a/Cu(OTf) , the reactions run
A
H
R
U
G
2
2
ture for 24 h. [b] Yield of isolated product based on aldehyde after chro-
matographic purification. [c] Enantiomeric excess was determined by
HPLC on Chiracel OD-H, OD, OJ, or AD-H columns. The absolute con-
figurations of nitroaldol adducts were assigned by comparison with litera-
ture compounds.
under an atmosphere of air proceed in comparable enantio-
meric excesses and yields (Table 6, entries 6–10).
Conclusion
tron-poor (Table 5, entries 4, 8, and 9), electron-neutral
In summary, we have developed a new addition to the ra-
tional design of sterically and electrically easily tunable
chiral bisimidazoline ligands, which are readily synthesized
in a straightforward fashion from commercially available, in-
expensive a-amino acids. Moreover, our method provides an
easy and simplified route to a diverse set of air-stable and
water-tolerant chiral bis(imidazoline) ligands on 10 g scales.
Our methodology is more versatile than the route reported
(
Table 5, entries 2, 5, and 6), or sterically hindered (Table 5,
entries 2–5), all of them smoothly underwent catalytic enan-
tioselective Henry reactions with reasonable product yields
and excellent enantioselectivities under our conditions. Most
remarkably, the examined aliphatic unbranched and even
branched or sterically hindered aldehydes also reacted with
nitromethane to give the optically active Henry adducts in
excellent yields (85–98%) and enantiomeric excesses (93–
[17]
by Beller and co-workers,
offering fewer restrictions in
2
96% ee) (Table 5, entries 10–14).
the R group substituent at C-4 and in the substituent at the
N-1 position in the imidazoline ring. We have further dem-
onstrated the use of (S)-1a in asymmetric Henry reactions,
in which it is the first member of the class of chiral bisimid-
Although not yet investigated in detail, the asymmetric
Henry reactions also took place with substantially lower cat-
alyst loadings than under our standard conditions of
10.5 mol% of (S)-1a. As shown in Table 6, treatment of
A
H
R
U
G
both aromatic and aliphatic aldehydes with nitromethane at
room temperature in the presence of 2 mol% of (S)-1a/Cu-
enantioselectivities. To the best of our knowledge, our dual
catalytic system is among the most effective systems so far
reported for the asymmetric parent Henry reactions. Further
investigations into other versions of asymmetric catalysis are
A
C
H
T
R
E
U
N
G
(OTf) and 2 mol% of Et N gave the corresponding nitroal-
2
3
cohol products with enantiomeric excesses in the 90–97%
1868
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1863 – 1871

Easily Tunable Chiral Bisimidazolines in Enantioselective Henry Reactions
FULL PAPER
currently underway and will be reported in due course. In
particular, because their modular natures enable easy tuning
of steric and electronic properties, it is reasonable to suggest
that this class of ligands should be widely applicable in
asymmetric catalysis.
70.33, 122.30, 122.97, 124.42, 128.24, 136.63, 142.68, 150.03, 159.66 ppm;
+
HRMS (FAB) calcd for [C29
H
33
N
5
+H] : 452.2814; found: 452.2817; ele-
mental analysis calcd (%) for C29
C 76.98, H 7.47, N 15.37.
H N : C 77.13, H 7.37, N 15.51; found:
33 5
2
,6-Bis[(S)-4-isopropyl-1-benzyl-4,5-dihydro-1H-imidazol-2-yl]pyridine
[
(S)-1b]: This compound was prepared by the same procedure as de-
scribed above (for (S)-1a). Starting materials were (S)-2-amino-3-methyl-
butan-1-ol and benzylamine. Compound (S)-1b was obtained in 78%
yield as a pale yellow semi-solid after purification by column chromatog-
Experimental Section
3
raphy on silica gel with elution with ethyl acetate/methanol/Et N
2
5
1
(
5:1:0.02). [a] = À133.5 (c = 0.39 in CHCl
3
); H NMR (600 MHz,
D
1
3
CD OD): d = 0.86 (2”d, J = 7.2 Hz, 7.2 Hz, 12H), 1.80–1.83 (m, 2H),
General remarks: H NMR spectra were obtained with a Bruker AV-300
3
.19 (t, J = 9.0 Hz, 2H), 3.51 (t, J = 11.4 Hz, 2H), 3.95–3.99 (m, 2H),
(
300 MHz) or
a
Varian Inova-600 (600 MHz) spectrometer, while
1
3
4.42 (2”d, J = 15.6 Hz, 15.6 Hz, 4H), 7.19–7.26 (m, 10H), 7.95 (d, J =
C NMR spectra were also recorded with a Bruker AV-300 (75 MHz) or
1
3
1
7.8 Hz, 2H), 8.09 ppm (t,
J = 7.2 Hz ,1H); C NMR (150 MHz,
a Varian Inova-600 (150 MHz). The H chemical shifts were measured
1
3
CD OD): d = 18.07, 18.76, 34.20, 51.97, 53.17, 69.89, 127.22, 128.58,
relative to tetramethylsilane as the internal reference, while the C NMR
chemical shifts were recorded with CDCl as the internal standard. Ele-
3
1
28.81, 129.63, 138.46, 139.76, 150.41, 164.65 ppm; HRMS (FAB) calcd
+
3
for [C31
%) for C23
3.97.
2,6-Bis[(S)-1,4-diisopropyl-4,5-dihydro-1H-imidazol-2-yl]pyridine
H
37
N
5
+H] : 480.3127; found: 480.3133; elemental analysis calcd
mental analyses were performed with a Heraeus CHN-O-RAPID instru-
ment. The high-resolution FAB-mass spectra and EI-mass spectra were
obtained with a JEOL JMS-SX/SX 102 A spectrometer, and the GC-MS
spectra were obtained with an Agilent 6890—5973 machine. Optical rota-
tions were determined on a WZZ-2B polarimeter. Melting points were
determined and are uncorrected.
(
1
37 5
H N : C 77.62, H 7.78, N 14.60; found: C 76.61, H 8.04, N
[(S)-
1c]: This compound was prepared by the same procedure as described
above ((S)-1a). Starting materials were (S)-2-amino-3-methylbutan-1-ol
and isopropylamine. Compound (S)-1c was obtained in 94% yield as a
white solid after purification by column chromatography on silica gel
Materials: l-Phenylalanine and l-valine were purchased from ChengDu
Chempep New Technology Co., Ltd, while d-phenylglycine was pur-
chased from Lancaster. Aldehydes were purchased from Aldrich, Lancas-
ter, Acros, and AstatTech in China: all liquid aldehydes were distilled
before use, while the others were used without further purification. Sol-
3
with elution with ethyl acetate/methanol/Et N (1:1:0.02) to methanol/
2
D
5
Et
3
N (1:0.01). m.p. 107–1088C; [a] = À144.1 (c = 0.51 in CHCl
3
);
1
H NMR (600 MHz, CDCl ): d = 0.94 (d, J = 6.6 Hz, 6H), 0.99 (d, J =
3
vents were dried by heating at reflux for at least 24 h over CaH
2
(di-
6.0 Hz, 6H), 1.03 (d, J = 7.2 Hz, 6H), 1.13 (d, J = 6.0 Hz, 6H), 1.84–
chloromethane or chloroform), sodium/benzophenone (THF), or sodium
1.90 (m, 2H), 3.18 (dd, J = 9.0 Hz, 9.6 Hz, 2H), 3.20 (dd, J = 9.6 Hz,
(
(
C
Et
2
H
5
OH or CH
N) was dried over P
and was freshly distilled prior to use. Except as noted, com-
3
OH) and freshly distilled prior to use. Triethylamine
9.6 Hz, 2H), 3.88–3.92 (m, 2H), 4.40–4.44 (m, 2H), 7.77–7.86 ppm (m,
1
3
3
2
O
5
, while nitromethane (CH NO ) was dried
3
2
3H); C NMR (150 MHz, CDCl ): d = 17.99, 18.92, 19.44, 20.79, 33.41,
3
over CaCl
2
45.76, 45.86, 69.41, 125.04, 137.18, 150.20, 162.77 ppm; HRMS (FAB)
mercial reagents were used as received without further purification.
Unless otherwise indicated, all syntheses and manipulations were carried
out under dry dinitrogen. Amino alcohols were prepared by the proce-
dure described by Meyers and McKennon. Isophthaloyl dichloride and
pyridine-2,6-dicarbonyl dichloride were prepared by the procedure de-
+
calcd for [C23
%) for C23
H
37
N
5
] : 383.3049; found: 383.3047; elemental analysis calcd
(
H
37
N
5
: C 72.02, H 9.72, N 18.26; found: C 71.29, H 9.85, N
1
8.05.
[
19]
2
1
,6-Bis[(S)-4-benzyl-1-phenyl-4,5-dihydro-1H-imidazol-2-yl]pyridine [(S)-
d)]: This compound was prepared by the same procedure as described
[
23]
scribed by Cram et al.
Procedure for the preparation of chiral bis(imidazoline) ligands 1a–h—
,6-bis[(S)-4-isopropyl-1-phenyl-4,5-dihydro-1H-imidazol-2-yl]pyridine
(S)-1a)]: solution of pyridine-2,6-dicarbonyl dichloride (8.16 g,
0 mmol) in CH Cl (50 mL) was added dropwise at 08C to a stirred solu-
above ((S)-1a). Starting materials were (S)-2-amino-3-phenylpropan-1-ol
and aniline. Compound (S)-1d was obtained in 85% yield as a pale
yellow powder after purification by column chromatography on silica gel
2
[
4
A
25
D
with elution with ethyl acetate/methanol (5:1). m.p. 55–578C; [a]
=
2
2
1
2
9
7
2
7
=
4.5 (c = 0.5 in CHCl
3
); H NMR (600 MHz, CDCl
3
): d = 2.75 (dd, J =
tion of l-valinol (8.24 g, 80 mmol) and triethylamine (11.0 mL, 80 mmol)
in CH Cl (150 mL). The reaction mixture was then allowed to warm to
room temperature and stirring was continued for 12 h, after which water
200 mL) was added. The layers were separated, the organic layer was
dried over Na SO , and the removal of the solvent in vacuo gave a white
solid, which was recrystallized from CH Cl to afford 3a as white crystals
.0 Hz, 4.8 Hz, 2H), 2.78 (dd, J = 9.0 Hz, 5.4 Hz, 2H), 3.65 (dd, J =
.8 Hz, 9.6 Hz, 2H), 3.67 (dd, J = 7.8 Hz, 9.6 Hz, 2H), 4.50–4.55 (m,
H), 6.50 (d, J = 7.8 Hz, 4H), 6.92 (t, J = 7.8 Hz, 2H), 7.05 (t, J =
.8 Hz, 4H), 7.18–7.27 (m, 10H), 7.52 (d, J = 7.8 Hz, 2H), 7.62 ppm (t, J
2
2
(
2
4
1
3
7.8 Hz, 1H); C NMR (150 MHz, CDCl
3
): d = 41.94, 57.83, 65.67,
2
2
1
1
5
8
22.10, 123.15, 124.58, 126.23, 128.28, 128.35, 129.31, 136.74, 138.17,
(
12.4 g, 92%).
+
42.29, 150.14, 160.07 ppm; HRMS (FAB) calcd for [C37
H
33
N
H
5
+H] :
33 5
N : C
SOCl
1
PCl
2
(4.20 mL, 57.5 mmol) was added at 08C to a mixture of 3a (3.37 g,
(70 mL), which was then stirred at reflux for 5 h.
5
(4.37 g, 21 mmol) was then added, the mixture was stirred at reflux
48.2814; found: 548.2806; elemental analysis calcd (%) for C37
1.14, H 6.07, N 12.79; found: C 80.31, H 6.37, N 12.37.
0.0 mmol) and CHCl
3
2
,6-Bis[(R)-1,4-diphenyl-4,5-dihydro-1H-imidazol-2-yl]pyridine [(R)-1e]:
for a further 6 h, and volatiles were removed under reduced pressure to
afford 4a. CHCl (70 mL), Et (10.2 mL, 72.9 mmol), and aniline
2.0 mL, 21.8 mmol) were added to the residue at 08C and the resulting
This compound was prepared by the same procedure as described above
(S)-1a). Starting materials were (R)-2-amino-2-phenylethanol and ani-
3
3
N
(
(
line. Compound (R)-1e was obtained in 75% yield as a brown powder
mixture was stirred at room temperature for 2 h, followed by heating at
reflux for 24 h. The solution was then washed with NaOH (20%, 50 mL)
and the aqueous layer was extracted with CH
bined organic layers were dried over Mg SO
moved under reduced pressure to give a brown solid, which could be pu-
rified by column chromatography on silica gel with elution with ethyl
after purification by column chromatography on silica gel with elution
2
5
with ethyl acetate/methanol/Et
3
N (15:1:0.02). m.p. 76–798C; [a]
D
=
2
Cl
2
(3”50 mL). The com-
1
2
44.0 (c = 0.5 in CHCl
3
); H NMR (600 MHz, CDCl
3
): d = 3.85 (dd, J
2
4
, and the solvent was re-
=
8.4 Hz, 9.6 Hz, 2H), 3.87 (dd, J = 8.4 Hz, 9.6 Hz, 2H), 5.29–5.32 (m,
2H), 6.66–6.68 (m, 4H), 7.01–7.03 (m, 2H), 7.14–7.17 (m, 4H), 7.26–7.30
acetate/methanol/Et
3
N (9:1:0.01) to afford (S)-1a as a pale yellow solid
(m, 6H), 7.34–7.36 (m, 4H), 7.70–7.73 (m, 1H), 7.76–7.78 ppm (m, 2H);
C NMR (150 MHz, CDCl ): d = 61.46, 67.67, 122.91, 123.60, 124.94,
3
2
5
13
(
3.97 g, 88%). m.p.: 121–1238C; [a] = À20.0 (c = 0.5 in CHCl
3
);
): d = 0.88 (d, J = 6.6 Hz, 6H), 0.98 (d, J =
D
1
H NMR (600 MHz, CDCl
3
126.77, 127.23, 128.39, 128.56, 136.97, 142.35, 143.57, 149.46, 160.72 ppm;
+
6
6
.6 Hz, 6H), 1.87–1.92 (m, 2H), 3.59–3.61 (m, 2H), 3.99–4.06 (m, 4H),
.60–6.62 (m, 4H), 6.92–6.94 (m, 2H), 7.08–7.11 (m, 4H), 7.58–7.64 ppm
HRMS (FAB) calcd for [C35
H
29
N
5
+H] : 520.2501; found: 520.2511; ele-
mental analysis calcd (%) for C35
C 80.68, H 5.90, N 13.20.
H N : C 80.90, H 5.63, N 13.48; found:
29 5
1
3
(
m, 3H); C NMR (150 MHz, CDCl
3
): d = 17.71, 18.79, 32.81, 55.82,
Chem. Eur. J. 2007, 13, 1863 – 1871
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1869

J. You and K. Ma
(
2
S)-4-Isopropyl-2-[3-((S)-4-isopropyl-1-phenyl-4,5-dihydro-1H-imidazol-
-yl)phenyl]-1-phenyl-4,5-dihydro-1H-imidazole [(S)-1 f]: A solution of
isophthaloyl dichloride (4.06 g, 20 mmol) in CH Cl (40 mL) was added
dropwise at 08C to a stirred solution of l-valinol (4.12 g, 40 mmol) and
triethylamine (5.5 mL, 40 mmol) in CH Cl (70 mL). The reaction mix-
phy on silica gel with elution with ethyl acetate/petrol ether to afford the
racemic nitroaldol adduct.
2
2
General procedure for the catalytic enantioselective Henry reaction: Cu-
A
H
R
U
G
2
(OTf) (18.1 mg, 0.05 mmol), ligand (S)-1 (0.0525 mmol), and ethanol
2
2
(1.5 mL) were placed in a Schlenk tube fitted with a magnetic stirrer
under N2 at room temperature. After the mixture had been allowed to
stir for 2 h, aldehyde (0.5 mmol) was added, and this was followed by the
addition of dry freshly distilled nitromethane (0.26 mL, 5 mmol) and tri-
ture was then allowed to warm to room temperature, stirring was contin-
ued for 12 h, and then water (100 mL) was added. The layers were sepa-
rated, the organic layer was dried over Na SO , and the removal of the
2 4
solvent in vacuo gave a white solid, which was recrystallized from CH
2
Cl
2
A
H
R
U
G
to afford 3 f as white crystals (6.4 g, 95%).
same temperature for 24 h, volatiles were then removed in vacuo, and
the residue was purified by column chromatography on silica gel with
elution with ethyl acetate/hexane to afford the nitroaldol adduct.
A solution of 3 f (3.67 g, 10.9 mmol) in SOCl
for 5 h, excess thionyl chloride was then removed under reduced pressure
to afford 4 f, and CH Cl (60 mL), Et N (9.0 mL, 66 mmol), and aniline
2.2 mL, 24 mmol) were added to the residue at 08C. The mixture was
2
(5 mL) was stirred at reflux
The reported yields of asymmetric Henry reactions are isolated yields
and are the averages of at least two runs. The Henry reaction products
2
2
3
(
1
13
then allowed to warm to room temperature and stirred for 24 h and then
washed with NaOH (10%, 50 mL), and the aqueous fraction was extract-
are known and their H and C NMR spectra agreed with those in the
[
6d,9]
literature cited below.
Enantiomeric excesses were determined by
HPLC (LabTech 600 series) with Chiracel OD-H, OD, OJ, or AD-H col-
umns. The absolute configurations of nitroaldol adducts were assigned by
comparison with literature compounds.
ed with CH
over Mg SO
2
Cl
2
(3”50 mL). The combined organic layers were dried
2
4
and the solvent was removed in vacuo to give a brown
[
6d,9]
solid, which could be purified by column chromatography on silica gel
with elution with ethyl acetate/methanol (9:1) to afford (S)-1 f as a pale
2
D
5
yellow semi-solid (4.07 g, 83%). [a]
3
= 91.5 (c = 0.55 in CHCl );
1
3
H NMR (600 MHz, CDCl ): d = 0.91 (d, J = 6.6 Hz, 6H), 0.99 (d, J =
Acknowledgements
6
4
6
2
1
1
.6 Hz, 6H), 1.86–1.91 (m, 2H), 3.57 (dd, J = 7.2 Hz, 9.6 Hz, 2H), 4.02–
.06 (m, 2H), 3.59 (dd, J = 7.2 Hz, 9.6 Hz, 2H), 6.71–6.72, (m, 4H),
.94–6.97 (m, 2H), 7.09–7.14 (m, 5H), 7.35 (2”d, J = 1.8 Hz, 1.8 Hz,
H), 7.89 ppm (t, J = 1.2 Hz, 1H); C NMR (150 MHz, CDCl ): d =
3
7.76, 18.73, 32.94, 56.11, 70.03, 122.59, 123.18, 127.71, 128.55, 129.18,
This work was supported by grants from the National Natural Science
Foundation of China (no. 20572074), the Program for New Century Ex-
cellent Talents in University (NCET-04-0881), the Outstanding Young
Scientist Award of Sichuan Province, the Setup Foundation of Sichuan
University, and the Foundation of the Education Ministry of China for
Returnees.
1
3
30.00, 131.59, 142.91, 160.67 ppm; HRMS (FAB) calcd for
+
[
C
30
H
34
N
4
+H] : 451.2862; found: 451.2852; elemental analysis calcd (%)
for C30
34 4
H N : C 79.96, H 7.61, N 12.43; found: C 79.85, H 7.74, N 12.23.
(
S)-4-Benzyl-2-[3-((S)-4-benzyl-1-phenyl-4,5-dihydro-1H-imidazol-2-yl)-
phenyl]-1-phenyl-4,5-dihydro-1H-imidazole [(S)-1g]: This compound was
prepared by the same procedure as described above ((S)-1 f). Starting
materials were (S)-2-amino-3-phenylpropan-1-ol and aniline. Compound
[
[
1] T. P. Yoon, E. N. Jacobsen, Science 2003, 299, 1691–1693.
2] For recent reviews, see: a) M. P. Sibi, S. Manyem, J. Zimmerman,
Chem. Rev. 2004, 104, 3263–3295; b) G. Desimoni, G. Faita, P.
Quadrelli, Chem. Rev. 2003, 103, 3119–3154; c) P. Braunstein, F.
Naud, Angew. Chem. 2001, 113, 702–722; Angew. Chem. Int. Ed.
2001, 40, 680–699; d) K. A. Jørgensen, M. Johannsen, S. Yao, H.
Audrain, J. Thornauge, Acc. Chem. Res. 1999, 32, 605–613; e) A. K.
Ghosh, P. Mathivanan, J. Cappiello, Tetrahedron: Asymmetry 1998,
9, 1–45.
(
S)-1g was obtained in 90% yield as a pale yellow powder after purifica-
tion by column chromatography on silica gel with elution with ethyl ace-
2
D
5
tate/methanol (9:1). m.p. 49–518C; [a] = 164.4 (c = 0.5 in CHCl
3
);
1
H NMR (600 MHz, CDCl
dd, J = 9.0 Hz, 4.8 Hz, 2H), 3.64 (dd, J = 6.6 Hz, 9.6 Hz, 2H), 3.65 (dd,
J = 6.6 Hz, 9.6 Hz, 2H), 4.48–4.53 (m, 2H), 6.57 (2”d, J = 0.6 Hz,
3
): d = 2.77 (dd, J = 9.0 Hz, 5.4 Hz, 2H), 2.79
(
1
2
.2 Hz, 4H), 6.93 (t, J = 7.2 Hz, 2H), 7.08–7.13 (m, 5H), 7.17–7.20 (m,
H), 7.22–7.26 (m, 8H), 7.33 (2”d, J = 1.2 Hz, 1.2 Hz, 2H), 7.90 ppm (t,
[3] For imidazolines as ligands in catalytic reactions, see: a) B. C¸ etin-
kaya, B. Alici, I. Özdemir, C. Bruneau, P. H. Dixneuf, J. Organomet.
Chem. 1999, 575, 187–192; b) B. C¸ etinkaya, E. C¸ etinkaya, P. B.
Hitchcock, M. F. Lappert, I. Özdemir, J. Chem. Soc. Dalton Trans.
1997, 1359–1362; c) A. Bastero, A. Ruiz, C. Claver, S. Castillón,
Eur. J. Inorg. Chem. 2001, 3009–3011; d) A. A. Ruiz, C. Claver, Or-
ganometallics 2002, 21, 5820–5829; e) A. Bastero, C. Claver, A. Rui,
S. Castillón, E. Daura, C. Bo, E. Zangrando, Chem. Eur. J. 2004, 10,
1
3
J = 1.2 Hz, 1H); C NMR (150 MHz, CDCl
3
): d = 42.12, 57.96, 65.14,
22.40, 123.33, 126.27, 127.81, 128.28, 128.55, 129.10, 129.35, 130.10,
31.39, 138.03, 142.44, 161.05 ppm; HRMS (FAB) calcd for
1
1
+
[
C
38
H
34
N
4
+H] : 547.2862; found: 547.2865; elemental analysis calcd (%)
: C 83.48, H 6.27, N 10.25; found: C 83.32, H 6.37, N 10.07.
R)-2-[3-((R)-1,4-Diphenyl-4,5-dihydro-1H-imidazol-2-yl)phenyl]-1,4-di-
for C38
34 4
H N
(
3
747–3760.
phenyl-4,5-dihydro-1H-imidazole [(R)-1h]: This compound was prepared
by the same procedure as described above ((S)-1 f). Starting materials
were (R)-2-amino-2-phenylethanol and aniline. Compound (R)-1h was
obtained in 78% yield as a pale yellow powder after purification by
column chromatography on silica gel with elution with ethyl acetate/
[
4] For imidazolines as ligands in enantioselective catalysis, see: a) C.
Botteghi, A. Schionato, G. Chelucci, H. Brunner, A. Kürzinger, U.
Obermann, J. Organomet. Chem. 1989, 370, 17–31; b) A. J. Davin-
port, D. L. Davies, J. Fawcett, D. R. Russell, J. Chem. Soc. Perkin
Trans. 1 2001, 1500–1503; c) J. Dupont, G. Ebeling, M. R. Delgado,
C. S. Consorti, R. Burrow, D. H. Farrar, A. J. Lough, Inorg. Chem.
Commun. 2001, 4, 471–474; d) C. A. Busacca, US Patent 6316620,
2
D
5
petrol ether (2:1). m.p. 83–858C; [a] = À42.0 (c = 0.5 in CHCl
3
);
1
3
H NMR (600 MHz, CDCl ): d = 3.83 (dd, J = 7.8 Hz, 9.6 Hz, 2H), 3.84
(
dd, J = 7.8 Hz, 9.6 Hz, 2H), 5.29 (2”d, J = 8.4 Hz, 7.8 Hz, 2H), 6.77–
2
001; e) F. Menges, M. Neuburger, A. Pfaltz, Org. Lett. 2002, 4,
6
.79 (m, 4H), 6.99 (t, J = 7.2 Hz, 2H), 7.15–7.17 (m, 5H), 7.24–7.34 (m,
0H), 7.46 (2”d, J = 1.2 Hz, 1.2 Hz, 2H), 8.06 ppm (t, J = 1.8 Hz, 1H);
4
713–4716.
1
[
5] For an excellent short review, see: C. Palomo, M. Oiarbide, A.
Mielgo, Angew. Chem. 2004, 116, 5558–5560; Angew. Chem. Int.
Ed. 2004, 43, 5442–5444.
1
3
C NMR (150 MHz, CDCl
3
): d = 61.58, 67.57, 122.84, 123.65, 126.67,
1
1
5
1
27.18, 127.84, 128.57, 128.70, 129.48, 130.37, 131.31, 142.56, 143.55,
61.77 ppm; HRMS (FAB) calcd for [C36
19.2538; elemental analysis calcd (%) for C36
0.80; found: C 83.11, H 6.01, N 10.59.
+
H
30
N
H
4
+H] : 519.2549; found:
: C 83.37, H 5.83, N
[
6] a) C. Christensen, K. Juhl, K. A. Jørgensen, Chem. Commun. 2001,
30 4
N
2
222–2223; b) C. Christensen, K. Juhl, R. G. Hazell, K. A. Jørgen-
sen, J. Org. Chem. 2002, 67, 4875–4881; c) D.-M. Du, S.-F. Lu, T.
Fang, J. Xu, J. Org. Chem. 2005, 70, 3712–3715; d) B. M. Choudary,
K. V. S. Ranganath, U. Pal, M. L. Kantam, B. Sreedhar, J. Am.
Chem. Soc. 2005, 127, 13167–13171; e) T. Purkarthofer, K. Gruber,
M. G. Khadjawi, K. Waich, W. Skranc, D. Mink, H. Griengl, Angew.
[
24]
Preparation of the racemic nitroaldols:
A
mixture of aldehyde
(
1.0 mmol), MeNO
2
(2.0 mmol), and Et N (3 drops) in ethanol
3
(
2.0 mmoL) was stirred at room temperature for 10 h. Volatiles were re-
moved in vacuo and the residue was purification by column chromatogra-
1870
www.chemeurj.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2007, 13, 1863 – 1871

Easily Tunable Chiral Bisimidazolines in Enantioselective Henry Reactions
FULL PAPER
Chem. 2006, 118, 3532–3535; Angew. Chem. Int. Ed. 2006, 45, 3454–
456.
7] a) H. Sasai, T. Suzuki, S. Arai, T. Arai, M. Shibasaki, J. Am. Chem.
Soc. 1992, 114, 4418–4420; b) T. Arai, Y. M. A. Yamada, N. Yama-
moto, H. Sasai, M. Shibasaki, Chem. Eur. J. 1996, 2, 1368–1372;
c) H. Sasai, S. Watanabe, T. Suzuki, M. Shibasaki, Org. Synth. 2004,
[18] The configuration of (S)-1a was confirmed by X-ray structure analy-
sis and these data are reported in: K. Ma, L. Cheng, H.-M. Gau, J.
You, Acta Crystallogr. Sect. E 2006, 62, o1595–o1597). The crystal
structure of 1a shows no symmetry relationship between the two
imidazoline rings, while the corresponding bond lengths reveal only
3
[
small variations. In comparison with oxazoline compounds, the im
azoline imine CÀN bond distances are slightly longer, by about
0.02 ꢂ.
A
C
H
T
R
E
U
N
G
id-
1
0, 571–577.
ACHTREUNG
[
8] a) B. M. Trost, V. S. C. Yeh, Angew. Chem. 2002, 114, 889–891;
Angew. Chem. Int. Ed. 2002, 41, 861–863; b) B. M. Trost, V. S. C.
Yeh, H. Ito, N. Bremeyer, Org. Lett. 2002, 4, 2621–2623.
[19] For the preparation of a-chiral amino alcohols, see: M. J. McKen-
non, A. I. Meyers, J. Org. Chem. 1993, 58, 3568–3571.
[
9] D. A. Evans, D. Seidel, M. Rueping, H. W. Lam, J. T. Shaw, C. W.
Downey, J. Am. Chem. Soc. 2003, 125, 12692–12693.
[20] For reviews on the concept of dual acid/base catalysis, see: a) M.
Shibasaki, N. Yoshikawa, Chem. Rev. 2002, 102, 2187–2209; b) M.
Shibasaki, M. Kanai, K. Funabashi, Chem. Commun. 2002, 1989–
1999; c) G. J. Rowlands, Tetrahedron 2001, 57, 1865–1882; d) J.-A.
Ma, D. Cahard, Angew. Chem. 2004, 116, 4666–4683; Angew. Chem.
Int. Ed. 2004, 43, 4566–4583; e) H. Grçger, Chem. Eur. J. 2001, 7,
5247–5251.
[21] For selected dual acid/base catalytic systems, see: a) Y. Hamashima,
D. Sawada, M. Kanai, M. Shibasaki, J. Am. Chem. Soc. 1999, 121,
2641–2642; b) E. Ichikawa, M. Suzuki, K. Yabu, M. Albert, M.
Kanai, M. Shibasaki, J. Am. Chem. Soc. 2004, 126, 11808–11809;
c) R. Takita, K. Yakura, T. Ohshima, M. Shibasaki, J. Am. Chem.
Soc. 2005, 127, 13760–13761; d) S. Lundgren, E. Wingstrand, M.
Penhoat, C. Moberg, J. Am. Chem. Soc. 2005, 127, 11592–11593;
e) C. Palomo, M. Oiarbide, R. Halder, A. Laso, R. López, Angew.
Chem. 2006, 118, 113–116; Angew. Chem. Int. Ed. 2006, 45, 117–
120..
[22] Selected recent excellent examples of catalysis with chiral bisoxazo-
lines, see: a) J. S. Johnson, D. A. Evans, Acc. Chem. Res. 2000, 33,
325–335; b) D. M. Barnes, J. Ji, M. G. Fickes, M. A. Fitzgerald, S. A.
King, H. E. Morton, F. A. Plagge, M. Preskill, S. H. Wagaw, S. J. Wit-
tenberger, J. Zhang, J. Am. Chem. Soc. 2002, 124, 13097–13105;
c) D. A. Evans, K. A. Scheidt, K. R. Fandrick, H. W. Lam, J. Wu, J.
Am. Chem. Soc. 2003, 125, 10780–10781; d) M. P. Sibi, Z. Ma, C. P.
Jasperse, J. Am. Chem. Soc. 2004, 126, 718–719; e) C. Palomo, M.
Oiarbide, R. Halder, M. Kelso, E.-G. Bengoa, J. M. Garcia, J. Am.
Chem. Soc. 2004, 126, 9188–9189; f) D. A. Evans, K. R. Fandrick,
H.-J. Song, J. Am. Chem. Soc. 2005, 127, 8942–8943; g) D. A. Evans,
J. Wu, J. Am. Chem. Soc. 2005, 127, 8006–8007; h) C. Palomo, M.
Oiarbide, E. Arceo, J. M. Garcꢃa, R. López, A. Gonzµlez, A.
Linden, Angew. Chem. 2005, 117, 6343–6345; Angew. Chem. Int.
Ed. 2005, 44, 6187–6190.
[
10] C. Palomo, M. Oiarbide, A. Laso, Angew. Chem. 2005, 117, 3949–
952; Angew. Chem. Int. Ed. 2005, 44, 3881–3884.
11] a) T. Ooi, K. Doda, K. Maruoka, J. Am. Chem. Soc. 2003, 125,
054–2055; b) Y. Sohtome, Y. Hashimoto, K. Nagasawaa, Adv.
3
[
2
Synth. Catal. 2005, 347, 1643–1648; c) T. Marcelli, R. N. S. van der
Haas, J. H. van Maarseveen, H. Hiemstra, Synlett 2005, 2817–2819;
d) T. Okino, S. Nakamura, T. Furukawa, Y. Takemoto, Org. Lett.
2
1
004, 6, 625–627; e) T. P. Yoon, E. N. Jacobsen, Angew. Chem. 2005,
17, 470–472; Angew. Chem. Int. Ed. 2005, 44, 466–468.
[
12] T. Marcelli, R. N. S. van der Haas, J. H. van Maarseveen, H. Hiem-
stra, Angew. Chem. 2006, 118, 943–945; Angew. Chem. Int. Ed.
2
006, 45, 929–931.
[
[
13] H. Li, B. Wang, L. Deng, J. Am. Chem. Soc. 2006, 128, 732–733.
14] a) T. Hayashi, E. Kishi, V. A. Soloshonok, Y. Uozumi, Tetrahedron
Lett. 1996, 37, 4969–4972; b) Y.-R. Lin, X.-T. Zhou, L.-X. Dai, J.
Sun, J. Org. Chem. 1997, 62, 1799–1834; c) D. H. Hunter, S. K. Sim,
Can. J. Chem. 1972, 50, 669–677; d) E. Y. Borisova, N. V. Kropoto-
va, E. M. Cherkasova, USSR Patent 466231, 1975; Chem. Abstr.
1
976, 83, 79277; e) R. Oi, K. B. Sharpless, Tetrahedron Lett. 1991,
3
2, 999–1002; f) J. Legters, J. G. H. Willems, L. Thijs, B. Zwanen-
burg, Recl. Trav. Chim. Pays-Bas 1992, 111, 59–68; g) P. Molina, I.
Dïaz, A. Tµrraga, Synlett 1995, 1031–1032; h) C. Hulme, L. Ma, J.
Romano, M. Morrissette, Tetrahedron Lett. 1999, 40, 7925–7928.
15] a) R. C. F. Jones, I. Turner, K. J. Howard, Tetrahedron Lett. 1993, 34,
[
6
329–6332; b) R. C. F. Jones, K. J. Howard, J. S. Snaith, Tetrahedron
Lett. 1996, 37, 1707–1710; c) Y. Langlois, P. I. Dalko, J. Org. Chem.
1
1
998, 63, 8107–8117; d) E. J. Corey, M. J. Grogan, Org. Lett. 1999, 1,
57–160; e) T. Isobe, K. Fukuda, Y. Araki, T. Ishikawa, Chem.
Commun. 2001, 243–244.
[
[
16] A procedure providing enantiopure 4-substituted imidazoline rings
from chiral 1,2-amino alcohols has recently been reported: N. A.
Boland, M. Casey, S. J. Hynes, J. W. Matthews, M. P. Smyth, J. Org.
Chem. 2002, 67, 3919–3922.
17] While we were collecting data, Beller and his co-workers reported
the first synthesis of chiral pyridinebisimidazolines derived from pyr-
idine-2,6-dinitrile and chiral diamines: S. Bhor, G. Anilkumar, M. K.
Tse, M. Klawonn, C. Dçbler, B. Bitterlich, A. Grotevendt, M.
Beller, Org. Lett. 2005, 7, 3393–3396.
[23] M. Newcomb, J. M. Timko, D. M. Walba, D. J. Cram, J. Am. Chem.
Soc. 1977, 99, 6392–6398.
[24] X. Beebe, N. E. Schore, M. J. Kurth, J. Org. Chem. 1995, 60, 4196–
4203.
Received: August 24, 2006
Revised: October 8, 2006
Published online: November 22, 2006
Chem. Eur. J. 2007, 13, 1863 – 1871
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
1871