Bis(oxazoline)-based coordination polymers: A recoverable system for enantioselective Henry reactions

Article abstract of DOI:10.1021/jo3005974

An efficient release-capture strategy for the recovery and reuse of enantioselective catalysts in the Henry reaction is described. This strategy is based on the precipitation of an insoluble coordination polymer at the end of each reaction, allowing easy

Full text of DOI:10.1021/jo3005974

Article
pubs.acs.org/joc
Bis(oxazoline)-Based Coordination Polymers: A Recoverable System
for Enantioselective Henry Reactions
Beatriz Angulo, Jose
́
I. García,* Clara I. Herrerías, Jose
́
A. Mayoral, and Ana C. Minana
̃
Instituto de Síntesis Química y Catal
Universidad de Zaragoza, E-50009 Zaragoza, Spain
́ ́ ́ ́
isis Homogenea (ISQCH) and Instituto Universitario de Catalisis Homogenea (IUCH), CSIC -
*
S Supporting Information
ABSTRACT: An efficient release−capture strategy for the
recovery and reuse of enantioselective catalysts in the Henry
reaction is described. This strategy is based on the
precipitation of an insoluble coordination polymer at the
end of each reaction, allowing easy separation from products.
The coordination polymer is formed through aggregation of
Cu(II) ions with ditopic bisoxazoline-based chiral ligands. Up
to 11 catalytic cycles have been conducted keeping good yields
and enantioselectivities.
INTRODUCTION
coordination polymers insoluble in low polar, noncoordinating
solvents. These polymers can therefore be selectively
precipitated at the end of the reaction simply by evaporation
of the reaction solvent and addition of hexane. The reaction
products are then extracted and the catalyst can be easily
reused. We have described the application of this release-
capture strategy to the cyclopropanation reaction of alkenes
■
In the past decade, a growing interest has been raised on the
recycling and reuse of enantioselective catalysts, using a variety
1
−6
of strategies.
Enantioselective catalysts are usually expensive
and often bear toxic metals. Therefore, easy separation from
reaction products and reuse in consecutive catalytic runs has
become an interesting goal from both academic and industrial
1
4,15
7
with ethyl diazoacetate,
and to the allylic oxidation of
points of view. Although immobilization through covalent
16
cycloalkenes with peroxoesters. As this is a relatively new
strategy, we are interested in testing its scope and applicability
to a variety of catalytic reactions, preferably with different
reaction mechanisms.
bonding has traditionally been one of the preferred strategies
for immobilization of chiral catalysts, it presents some
disadvantages, such as the necessity of functionalization of
the chiral ligand, which may affect to the activity and selectivity
8
The Henry or nitroaldol reaction is an attractive C−C bond-
forming reaction in which a nitroalkane compound is added to
an aldehyde or ketone to give primarily a 1,2-nitro alcohol,
which may be subsequently transformed in a variety of
functional groups, such as amine, carbonyl, carboxyl, azide,
etc. In many of these transformations, the stereogenic centers
formed in the addition reaction are kept, so enantioselective
versions of this reaction have been developed on the basis of
of the final catalysts, and the effect introduced by the support.
The latter often results in lower activities and enantioselectiv-
ities. Noncovalent immobilization strategies, like entrapment,
ionic exchange, or adsorption, avoid the chemical modification
of the chiral ligand but still introduce the perturbation of the
5
,6
support on the reactive system.
Liquid−liquid biphasic
systems lack, in principle, these drawbacks, but it is not always
possible to find two immiscible liquid phases compatible with
the catalytic requirements, i.e., lack of chemical interference
with catalyst and/or reagents, solubility of the catalyst only in
one phase, and reagents and products preferably in the other,
1
7,18
the use of chiral catalysts.1
Although the use of organo-
most of the chiral catalysts used
9,20
catalysts has been reported,
are based on coordination complexes of transition metals or
1
7,18,21
Recent examples can be
23,24 25,26
1
,9
lanthanides with chiral ligands.
etc. Clearly, the least perturbation would take place if the
reaction was carried out in the homogeneous phase, and then
the catalyst could be selectively precipitated. This would allow
an easy separation and further recycling. This goal can be
accomplished through different strategies, including among
22
found using lanthanides, zinc,
cobalt,
or chromi-
2
7,28
um,
but undoubtedly copper continues to be the preferred
Concerning the chiral ligands, since the pioneer-
ing work of Jørgensen and Evans, the bis(oxazoline)
family of ligands has become quite popular because of its easy
2
1,29−32
metal.
3
3,34
35
4
others linking of the chiral ligand to a soluble polymer, or self-
36−38
availability and the good results reported.
supporting of the catalytic complex through hydrogen
1
0,11
12,13
Several recoverable enantioselective catalysts have been
recently described for the Henry reaction, mainly based on
bonding
or coordination polymerization.
In most
cases, the selective precipitation of the catalyst takes place by
addition of a suitable cosolvent at the end of the reaction.
Our group has recently described a new family of polytopic
chiral ligands, based on the oxazoline moiety, able to form
Received: March 22, 2012
Published: June 5, 2012
©
2012 American Chemical Society
5525
dx.doi.org/10.1021/jo3005974 | J. Org. Chem. 2012, 77, 5525−5532

The Journal of Organic Chemistry
Article
three of the aforementioned strategies. Covalent bonding to a
synthesized following the methodology reported in our
previous work, by deprotonation of the corresponding
azabis(oxazoline) with butyllitium and subsequent reaction
2
2,39
support
has been described using silica as supports. Only
four catalytic cycles (so three recoveries) have been carried out
15
with these supported catalysts, with varied results. In the first
with α,α′-dibromo-p-xylene. Herein we also describe the
synthesis of a new ditopic chiral ligand, i-PrDiBox (2), which
was obtained by reaction of α,α′-dibromo-p-xylene with 2 equiv
of the bis(oxazoline) (3) previously deprotonated using
2
2
case, only ca. 65% ee could be achieved at −44 °C, whereas in
3
9
the second one, high enantioselectivities only could be
reached by using 10 mol % of catalyst and 25 mol % of a chiral
additive in the homogeneous phase, which makes the recovery
procedure less attractive. Biphasic catalytic systems described to
date are based on the use of ionic liquids. Thus, the use of a
46
potassium tert-butoxide as a base (Scheme 2).
It must be noted that a nonchiral DiBox ligand has been
recently described by Bellemin-Laponnaz and co-workers, as
4
0
47
chiral diamine−Cu(OAc) complex allowed up to five
well as their Zn(II), Ni(II), and Cu(II) complexes. However,
2
catalytic cycles with ca. 80% yield and 88% ee at 0 °C and
using 12 mol % of catalyst. Better results have recently been
described by using imidazolium-tagged bis(oxazolines).
these complexes have 2:1 metal:ligand stoichiometry, so no
coordination polymers were observed by these authors in any
case. Also, a ditopic arylidenIndaBox ligand has been described
by Yuriev and Liese and used in the homogeneous phase in the
4
1,42
Up
to 12 catalytic cycles were conducted in the best case, at 0 °C
and using 14 mol % catalyst, keeping enantioselectivities
around 90% ee and yields around 50%. The last strategy
described for catalyst recycling is a selective precipitation, either
48
prototypical Henry reaction, with rather modest results.
It is noteworthy to point out that the preparation of both
ditopic chiral ligands was very efficient, and high or even
quantitative yields were obtained in all cases. In addition, the
synthetic effort required for the preparation of the ditopic
ligands is equivalent to that needed for the preparation of the
corresponding monotopic ligands traditionally used in homo-
geneous catalysis.
43
by inclusion of fluorous ponytails in the chiral ligand or by
44
formation of charge-transfer adducts. In the first case, four
catalytic cycles were conducted with a continuous decrease in
enantioselectivity (from 90% ee in the first cycle to 72% ee in
the fourth cycle). In the second case, up to seven cycles could
be achieved, but also with a decrease in enantioselectivity and,
above all, in activity (from 81% yield at 24 h in the first cycle to
For the preparation of the coordination polymers, copper
was used as ligand-connecting metal; specifically, Cu(OAc)2
was chosen since this copper salt provided excellent results in
5
9% yield at 120 h in the last one).
In this work, we describe the application of copper
3
5
the nitroaldol reaction. The acetate ions facilitate the
deprotonation of nitroalkanes, making the use of an additional
base for the activation of the nitro derivative unnecessary and
avoiding possible interference, for example, in the final
precipitation of the catalyst. In addition, acetate as a counterion
fulfills an important requirement in the formation of
coordination polymers since it does not have a coordinating
ability as strong as chloride, which would prevent the
recruitment of two ligands by copper. For the same reason, a
weakly coordinating solvent or mixture of solvents would be
the perfect solvent for the formation of the self-supported
catalysts at the end of every reaction cycle.
complexes of ditopic ligands (i-PrDAX and i-PrDiBox, Figure
) as catalysts for the enantioselective nitroaldol (Henry)
1
Figure 1. Structure of the chiral ditopic ligands i-PrDAX and i-
PrDiBox.
Catalytic Results. In our previous works, the coordination
polymers were applied to the cyclopropanation reaction
between styrene and ethyl diazoacetate in dichloromethane
and the allylic oxidation of cycloalkenes with tert-butyl
perbenzoate. In the cyclopropanation, the polymers remained
solid in the absence of reactants and became soluble and
catalytically active in the presence of ethyl diazoacetate because
of the breakage of the polymer and formation of a carbenoid
reaction between several aromatic aldehydes and two different
nitro compounds (Scheme 1) and their efficient recovery and
reuse by selective precipitation of the corresponding coordina-
tion polymers.
RESULTS AND DISCUSSION
■
1
4,15
Synthesis of Ligands and Preparation of Coordination
Polymers. As we have previously commented, the ditopic
ligands selected to carry out this study were those based on
azabis(oxazoline) and bis(oxazoline) units, both with isopropyl
groups (Figure 1). Those units were chosen since their
homologous monotopic ligands have led to the best results in
intermediate species.
In the case of the Henry reaction, the
solvent is mainly responsible for the solubilization of the
coordination polymer since it is a coordinating aliphatic alcohol
that joins the coordination sphere of the copper leading to
16
monomeric and oligomeric species. The influence of the
solvent in the title reaction with chiral monotopic ligands (4
and 5) is shown in Table 1.
2
9,43,45
the nitroaldol reaction.
The i-PrDAX ligand (1) was
Scheme 1. Nitroaldol (Henry) Reactions Used in the Catalytic Tests
5
526
dx.doi.org/10.1021/jo3005974 | J. Org. Chem. 2012, 77, 5525−5532

The Journal of Organic Chemistry
Article
Scheme 2. Synthesis of the i-PrDiBox Ligand
Table 1. Influence of the Solvent on the Enantioselectivity of
the Henry Reaction
Table 2. Henry Reaction between Nitromethane and Several
Aromatic Aldehydes
a
a
b
c
entry
R
yield (%)
ee (%)
1
2
3
4
5
6
H
92
99
40
76
75
68
91
90
90
74
89
89
o-MeO
p-MeO
p-NO2
p-Cl
p-CH3
b
entry
ligand (L*)
solvent
ee (%)
a
Reagents and conditions: aromatic aldehyde (1 mmol), nitromethane
(10 mmol), L*-Cu(OAc) (0.05 mmol, 5 mol %), i-PrOH (2 mL),
room temperature, 24 h. Yields were determined by H NMR using
an internal standard. Enantiomeric excesses were determined by
HPLC using Chiralcel OD-H and Chiralpak IB columns. (S) product
1
2
3
4
5
6
i-PrAzaBox-Me (4a)
i-PrAzaBox-Me (4a)
i-PrAzaBox-Me (4a)
MeOH
EtOH
70
87
91
65
84
84
2
b
1
c
i-PrOH
MeOH
EtOH
i-PrBox-Me (5a)
2
was the major enantiomer in all cases. See the Supporting Information.
i-PrBox-Me (5a)
2
i-PrBox-Me (5a)
i-PrOH
2
a
Reagents and conditions: benzaldehyde (1 mmol), nitromethane (10
mmol), L*-Cu(OAc) (0.05 mmol, 5 mol %), solvent (2 mL), room
temperature, 24 h. Enantiomeric excesses were determined by HPLC
using a Chiralcel OD-H column. (S)-2-nitro-1-phenylethanol is the
major enantiomer. See the Supporting Information.
2
derivative showed a lower selectivity, a fact also observed in
similar studies.
b
29,49
In all cases, the (S) product was the major enantiomer, which
corresponds to a preferential approach of the nitro compound
to the Re face of the aldehyde. This is in line with previous
2
9,35,42
studies of similar catalysts
and can be rationalized
Independent of the chiral ligand utilized, the use of i-PrOH
as solvent produces a great increase in the enantioselectivity of
the reaction of Henry between benzaldehyde and nitromethane
assuming a pyramidal five-coordinated copper transition
35
structure following the first proposal made by Evans where
the transition structures in which steric interactions between
the nitro compound and the substituents of the oxazoline rings
are minimized (Figure 2). Approach by the Si face of the
aldehyde always results in steric hindrance either between the
aromatic group and the coordinated acetate ion or between one
(
8
up to 20%). A 91% ee with i-PrAzaBox ligand (4a) and an
4% ee with i-PrBox ligand (5a) in i-PrOH as solvent were
promising values for the application of their homologous
ditopic ligands in this reaction.
Before testing the coordination polymers as catalysts, other
aromatic aldehydes bearing either electron-withdrawing or
electron-donating groups were examined. For this study, the
complex 4a−Cu(OAc) was chosen as catalyst since it had
2
presented the best activity and selectivity in the previous
experiments. The results are gathered in Table 2. In general,
from moderate to high enantiomeric excesses were obtained in
all cases (74−91%). The position of the substituent in the
aromatic ring showed a clear influence in the activity of the
catalyst; the change of a methoxy group from an ortho to a para
position reduced the yield from 99% to 40% (Table 2, entries 2
and 3). However, as happens in similar studies carried out with
substituted aromatic aldehydes by other authors, there is not a
2
9,42,49
clear rationale for the fluctuations observed.
Clearly, it is
not an electronic effect, for p-NO , p-Me, and p-MeO
2
substituents lead to worse results than the parent compound,
benzaldehyde. Variations in enantioselectivity seem to be less
dependent on the aromatic substitution, and only the p-NO2
Figure 2. Stereochemical models to explain the enantiodiscrimination
observed.
5
527
dx.doi.org/10.1021/jo3005974 | J. Org. Chem. 2012, 77, 5525−5532

The Journal of Organic Chemistry
Article
of the oxygen atoms of the nitro group and one of the
substituents of the bis(oxazoline) (Figure 2).
Table 3. Henry Reaction between Nitromethane and o-
a
Anisaldehyde
In view of the results obtained in the study of the aldehyde
substrates, o-anisaldehyde was chosen as reference substrate in
order to carry out the study with the catalysts based on the
ditopic chiral ligands. In these experiments, the catalytic
complexes (1−Cu(OAc) or 2−Cu(OAc) ) (L*/Cu = 1/1
b
c
2
2
entry
L*
run
yield (%)
ee (%)
molar ratio) were prepared in i-PrOH, and subsequently, the
reagents were added to the solution. Once the reaction was
complete, the solvent and the excess of nitromethane were
removed under vacuum to avoid the presence of coordinating
molecules. A noncoordinating solvent was added to the crude
to extract the products, byproducts, and remaining starting
materials and at the same time allow the self-assembly of the
catalyst. At the beginning, hexane was chosen as a perfect
candidate because it could fulfill both requirements, that is, low
coordinating ability and a right solubilization of products.
However, part of the products remained together with the
catalyst after the extractions because of a partial solubilization
of them in hexane. Then, several solubility experiments were
1
2
3
4
5
6
7
8
9
0
1
2
i-PrAzaBox-Me (4a)
i-PrAzaBox-Bn (4b)
i-PrDAX (1)
1
1
1
2
1
1
1
2
3
4
5
6
99
78
25
5
90
89
95
98
84
63
76
62
75
73
72
73
i-PrBox-Me (5a)
82
82
47
30
28
30
30
14
2
i-PrBox-Bn (5b)
i-PrDiBox (2)
1
1
1
a
Reagents and conditions: o-anisaldehyde (1 mmol), nitromethane
(10 mmol), L*-Cu(OAc) (0.05 mmol, 5 mol %), i-PrOH (2 mL),
room temperature, 24 h. Yields were determined by H NMR using
carried out, and finally, a mixture of hexane/Et O = 1:1 was
2
2
b
1
chosen as an extraction solvent that maintains suitable
noncoordinating properties. After the addition of the solvents
to the reaction crude, a green solid precipitated as a
consequence of the assembly of the corresponding monomeric
catalytic species, in other words, as a result of the formation of
the coordination polymer, and accordingly, products could be
separated from the catalyst. Therefore, the latter could be
reused in successive reactions by adding new portions of
solvent and reagents. In the presence of i-PrOH, the polymer
disassembled again due to competitive coordination of solvent
molecules and thus became soluble and catalytically active
c
an internal standard. Enantiomeric excesses were determined by
HPLC using a Chiralcel OD-H column. (S)-1-(2-methoxyphenyl)-2-
nitroethanol was the major enantiomer in all cases. See the Supporting
Information.
have also been included for comparison, since they are closer to
the structure of the ditopic ligands used in this study (Figure
). It also must be pointed out that the dehydration product
4
(
Figure 3).
The results obtained with these kinds of complexes,
including their recovery and reuse, are collected in Table 3.
The results obtained with the corresponding monotopic ligands
with benzylic groups at the central nitrogen or carbon atom
Figure 4. Structure of the chiral monotopic ligands bearing benzylic
groups.
from the β-nitro alcohol was detected as a byproduct in some of
the Henry reactions described in this work. However, the yield
of this byproduct was considered insignificant (<5%), and it has
not been included in the following tables.
The presence of a benzyl group in the central nitrogen of the
azabis(oxazoline) leads to a decrease in the reaction yield
maintaining a similar enantioselectivity (Table 3, entries 1 and
2
). When the ditopic ligand 1 was tested in the Henry reaction
with o-anisaldehyde, an increase in the enantioselectivity was
achieved (Table 3, entry 3), even improving those obtained
with the monotopic ligands 4a and 4b. Nevertheless,
disappointingly, the activity of the catalyst was dramatically
reduced, and although the reuse of the catalyst retained the
enantioselectivity, the yield was only 5% (Table 3, entries 3 and
4
). The presence of the benzyl group in the bis(oxazoline) had
a different effect from that observed with azabis(oxazolines).
The use of the ligand 5b led to similar results in terms of yield,
but enantioselectivity was significantly reduced (Table 3, entries
5 and 6). The reaction with the ditopic ligand 2 led to an
Figure 3. Equilibrium of solvent-aided assembly disassembly.
5
528
dx.doi.org/10.1021/jo3005974 | J. Org. Chem. 2012, 77, 5525−5532

The Journal of Organic Chemistry
Article
intermediate enantioselectivity (76% ee) with a moderate yield,
and in this case, the recovery and reuse of the coordination
polymer was possible during five reaction cycles without a
remarkable loss of enantioselectivity (Table 3, entries 7−12).
The results of enantioselectivity obtained with the ditopic
chiral ligands were, in both cases, comparable to those obtained
with the monotopic ligands; however, the results in terms of
activity were still far from being acceptable. In order to rule out
a possible error in the experimental determination of the
results, a larger scale reaction was examined. Similar values for
yield and enantioselectivity were obtained in this experiment,
and the recoverability of the catalyst was not improved either.
On the other hand, it has been demonstrated that ligands
bis(oxazoline) and azabis(oxazoline) have different coordinat-
ing abilities toward copper. Azabis(oxazoline)−copper
complexes are considerably more stable than the analogous
bis(oxazoline)−copper complexes which, in our case, would
involve a greater difficulty for the disassembly of the
coordination polymer from ligand 1. Consequently, this could
explain the differences found in the activity of the catalyst i-
PrDAX−Cu (25% of yield) compared to that of the catalyst i-
PrDiBox−Cu (47% of yield).
Table 4. Henry Reaction between Nitromethane and o-
a
Anisaldehyde Using 4 mL of i-PrOH
c
d
entry
1
L*
run
yield (%)
ee (%)
i-PrAzaBox-Me (4a)
i-PrAzaBox-Bn (4b)
i-PrDAX (1)
1
1
93
78
83
77
89
84
80
90
81
92
90
95
96
82
82
65
70
78
88
75
75
72
61
53
40
44
94
89
85
85
89
75
92
91
90
90
88
87
83
86
63
60
70
70
62
79
73
60
57
58
57
53
b
2
3
4
5
6
7
8
9
1
2
3
4
5
0
5
6
7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
8
9
10
11
1
i-PrBox-Me (5a)
2
b
i-PrBox-Bn (5b)
i-PrDiBox (2)
1
In addition, when coordination polymers are used in a
release−capture strategy, the catalyst remains heterogeneous at
its resting state at the end of the reaction, but afterward, for its
reuse, its disassembly and solubilization in the reaction medium
are essential in order for it to become completely catalytically
active. In other words, a partial disassembly of the polymer
would produce dimeric or oligomeric species that are
catalytically active, but at the same time, they also include
blocked copper centers that are not able to catalyze the
reaction. It is worth mentioning that oligomeric species of this
1
2
3
4
5
6
7
8
9
10
11
1
kind have been previously detected in solution by DOSY H
16
NMR experiments. Therefore, the presence of dimeric or
oligomeric species misses part of the catalytic potential of the
coordination polymers. Taking into account these consider-
ations, we thought that the low activity of the polymeric
catalysts could be due to a lack of coordinating or enough
coordinating molecules in the reaction medium. It is evident
that in the previous experiments i-PrOH was able to
disassemble the coordination polymer, at least partially, but
possibly the number of available molecules of solvent was not
sufficient to shift the equilibrium of assembly disassembly and
thus release all potential monomeric catalytic complexes
a
Reagents and conditions: o-anisaldehyde (1 mmol), nitromethane
10 mmol), L*-Cu(OAc) (0.05 mmol, 5 mol %), i-PrOH (4 mL),
(
2
b
room temperature, 24 h. Reactions carried out with 2 mL of i-PrOH.
c
1
Yields were determined by H NMR using an internal standard.
d
Enantiomeric excesses were determined by HPLC using a Chiralcel
OD-H column. (S)-1-(2-methoxyphenyl)-2-nitroethanol was the
major enantiomer in all cases. See the Supporting Information.
(Table 4, entries 3−13), which could be used without a
significant loss of activity or selectivity during all reaction cycles.
In the case of the i-PrDiBox−Cu complex, there is an incipient
decrease in the yield and the enantioselectivity after runs 8 and
9 (Table 4, entries 23 and 24) which could be produced by a
progressive loss of small amounts of complex dissolved in the
extraction solvents. A copper analysis of the extracted reaction
crude was carried out by ICPMS showing a loss of metal
around 3% (0.0015 mmol Cu) with regard to the total amount
of copper added in the beginning. In some experiments carried
out in our laboratory using benzaldehyde as a substrate, the
addition of 20% more of i-PrDiBox−Cu after the ninth run
allowed us to restore the activity and the selectivity to their
initial values. A lower stability of the i-PrDiBox−Cu complexes
compared to the stability of their analogous i-PrDAX−Cu
complexes would be a possible explanation for the loss of
catalyst observed in the first case.
(
Figure 3). Accordingly, we decided to carry out the reaction
with 4 mL of i-PrOH instead of 2 mL in order to evaluate this
possibility (Table 4).
If we compare the results of Tables 3 and 4, it can be
concluded that an increase in the volume of solvent produces a
great improvement in the activity of catalysts including ditopic
ligands, that is, from 25% of yield (Table 3, entry 3) to 83%
(Table 4, entry 3) for ligand 1 and from 47% of yield (Table 3,
entry 7) to 65% (Table 4, entry 16) for ligand 2. These results
confirm our previous hypothesis; thus, a larger number of
coordinating solvent molecules in the reaction medium has
allowed the disassembly of the coordination polymers, releasing
all the latent catalytic complexes borne in that kind of
structures.
In addition to this, both i-PrDAX−Cu and i-PrDiBox−Cu
could be recovered and reused in 11 consecutive reactions with
similar results to those obtained with their corresponding
conventional complexes with ligands 4a and 4b. It is worth
highlighting the efficient recovery of the i-PrDAX−Cu complex
It should be mentioned that these results are not compatible
with the behavior recently described by Hong and co-workers
regarding the catalytic activity of the N-benzylazabis(oxazoline)
ligand 4b. These authors described that the ligand, without
copper, is able to catalyze efficiently the Henry reaction of p-
29
5
529
dx.doi.org/10.1021/jo3005974 | J. Org. Chem. 2012, 77, 5525−5532

The Journal of Organic Chemistry
Article
nitrobenzene with nitromethane (90% yield in 5 h at room
temperature, no enantioselectivity) even better than the same
Moreover, enantioselectivity in the anti product is lower than
that of syn diastereomer. This could be explained by a relative
destabilization of the lowest energy transition state due to steric
repulsion between the methyl groups of nitroethane and acetate
ion (Figure 5).
ligand with Cu(OAc) (95% yield in 16 h at room temperature,
2
75% ee). Under our reaction conditions with the i-PrDAX
ligand, when the polymer is completely disassembled, half of
the azabis(oxazoline) moieties bear a coordinate copper and
half do not. We could therefore observe an important decrease
of enantioselectivity because of the concurrence of the
nonenantioselective ligand-catalyzed reaction. As can be seen
in Table 4, this effect is not observed. We then tried out the
reaction of benzaldehyde with nitromethane in the presence of
ligand 4a, without copper, at room temperature. After 24 h, no
conversion of the reagents could be detected, so it seems that
under our reaction conditions the metal-free catalysis described
by Hong and co-workers does not take place and no
interference (nor cooperativity) with the free azabis(oxazoline)
moieties of the i-PrDAX ligand occurs.
Figure 5. Stereochemical models to explain the lower enantiodiscrimi-
nation observed in the anti product.
After the good results obtained with these coordination
polymers in both activity/selectivity and recoverability, we
decided to carry out a deeper evaluation of the Henry reaction
scope, and thus, another nitro derivative such as nitroethane
was also considered. The results with this component are
gathered in Table 5.
We decided to carry out the reactions with the ditopic ligand
(i-PrDAX) since the values for the enantioselectivity were
1
slightly better using the ligand with azabis(oxazoline) units. As
can be seen in Table 5, the i-PrDAX−Cu complex could again
be recovered and reused for up to seven reaction cycles with
good results (Table 5, entries 3−10). Despite the somewhat
erratic values in the enantioselectivity of the anti products, the
values of enantiomeric excess for the syn products were
moderately high during all the cycles and in addition, the yields
obtained using this catalyst were better than that obtained with
complex 4a−Cu. These results increase the application scope of
this release-capture methodology using coordination polymers.
Table 5. Henry Reaction between Nitroethane and o-
a
Anisaldehyde
CONCLUSIONS
■
c
d
,
e
d,f
yield
(%)
ee anti
(%)
ee syn
We have described the preparation of different coordination
polymers using copper(II) and two ditopic chiral ligands based
on bis(oxazoline) units, that is, i-PrDAX (1) and the new i-
PrDiBox ligand (2). The use of these coordination polymers in
the Henry reaction with different substrates has allowed the
recovery and reuse of the catalysts up to 11 reaction cycles with
very good results of activity and selectivity through a release−
capture strategy. The total disassembly and solubilization of the
coordination polymer in the reaction medium has proved to be
essential in order to take advantage of the full catalytic
properties of these kinds of structures. These results enlarge the
scope of applicability of this release−capture strategy through
the use of polytopic chiral ligands able to form insoluble
coordination polymers after reaction completion, thus joining
the best of homogeneous and heterogeneous catalysis worlds.
entry
L*
run
1
anti/syn
(%)
b
1
i-PrAzaBox-Me
47
57/43
77
91
(
4a)
b
2
i-PrBox-Me2
5a)
1
58
62/38
66
80
(
3
4
5
6
7
8
9
i-PrDAX (1)
1
2
3
4
5
6
7
8
70
60
74
71
66
51
40
27
55/45
49/51
52/48
54/46
48/52
54/46
59/41
58/42
75
40
67
64
61
50
72
55
91
80
87
86
89
86
92
90
10
a
Reagents and conditions: o-anisaldehyde (1 mmol), nitroethane (10
mmol), L*-Cu(OAc) (0.05 mmol, 5 mol %), i-PrOH (4 mL), room
temperature, 24 h. Reactions carried out with 2 mL of i-PrOH.
Yields were determined by H NMR using an internal standard.
Enantiomeric excesses were determined by HPLC using a Chiralpak
AD-H column. (1S,2R)-1-(2-Methoxyphenyl)-2-nitropropan-1-ol was
the major enantiomer. (1S,2S)-1-(2-Methoxyphenyl)-2-nitropropan-
2
b
c
1
EXPERIMENTAL SECTION
■
d
General Methods. All reactions were carried out under argon
atmosphere in oven-dried glassware. Anhydrous solvents such as
tetrahydrofuran, dichloromethane, and MeOH were obtained from a
SPS-device; however, EtOH and i-PrOH were distilled from calcium
hydride. Liquid benzaldehydes were obtained from different
commercial sources and distilled before use. EtNO2 was distilled
from potassium carbonate. The rest of purchased reagents were used
as received without further purification. The ligands i-PrAzaBox-Me
e
f
1
-ol was the major enantiomer. See the Supporting Information.
Preliminary reactions were carried out with ligands 4a and 5a
using nitroethane as starting material (Table 5, entries 1 and 2).
The results obtained are in line with those previously described
using similar ligands. Thus, low anti/syn selectivity is
observed, which can be explained by the low steric repulsion
expected between the nitro compound alkyl chain and the
aromatic ring of the aldehyde (Figure 2). This is clearly a
drawback for the reaction and constitutes an area of future
effort, although it is out of the scope of the present study.
5
1−53
51−53
54
(
4a),
i-PrAzaBox-Bn (4b),
i-PrBox-Bn (5b), and i-
1
4,15
4
4
PrDAX
were prepared according to literature procedures. The
1
chemical shifts were relative to TMS as an internal reference for H
NMR.
Synthesis of 1,4-Bis(2,2-bis((S)-4-isopropyl-4,5-dihydrooxa-
zol-2-yl)ethyl)benzene (i-PrDiBox) (2). A solution of 2,2′-
methylenebis[(4S)-4-isopropyl-4,5-dihydrooxazole] (3) (238 mg, 1
mmol) in anhydrous tetrahydrofuran (5 mL) was combined with
5
530
dx.doi.org/10.1021/jo3005974 | J. Org. Chem. 2012, 77, 5525−5532

The Journal of Organic Chemistry
Article
another solution of t-BuOK (112 mg, 1 mmol) also in anhydrous
tetrahydrofuran (5 mL). After the solution was stirred for 1 h at room
temperature, α,α′-dibromo-p-xylene (132 mg, 0.5 mmol) was added.
The mixture was heated under reflux overnight. The solvent was
evaporated under reduced pressure, and the residue was partitioned
between ethyl acetate (10 mL) and a saturated aqueous NaCl solution
CTQ2011-28124-C02-01). A.C.M. is indebted to the Consejo
Superior de Investigaciones Cientificas for a grant (CSIC JAE-
Pre program).
́
REFERENCES
■
(
10 mL). The aqueous phase was extracted with ethyl acetate (3 × 10
(
3
1) Dupont, J.; de Souza, R. F.; Suarez, P. A. Z. Chem. Rev. 2002, 102,
mL), and the combined organic phases were dried over anhydrous
667−3692.
MgSO . Evaporation of the solvent yielded the product as orange oil in
4
(
2) Fan, Q.-H.; Li, Y.-M.; Chan, A. S. C. Chem. Rev. 2002, 102,
25
1
almost quantitative yield: [α] = −57.5 (c 0.54, CHCl ); H NMR
D
3
3
(
(
385−3466.
(
3
4
400 MHz, CDCl , δ ppm) 7.10 (s, 4H), 4.28−4.06 (m, 4H), 4.01−
3
3) Song, C. E.; Lee, S.-G. Chem. Rev. 2002, 102, 3495−3524.
.79 (m, 8H), 3.76−3.63 (m, 2H), 3.25−3.05 (m, 4H), 1.79−155 (m,
H), 0.89 (d, 6H, J = 6.8 Hz), 0.82 (d, 6H, J = 6.8 Hz), 0.81 (d, 6H, J
4) Bergbreiter, D. E.; Tian, J.; Hongfa, C. Chem. Rev. 2009, 109,
6.8 Hz), 0.75 (d, 6H, J = 6.8 Hz); 13C NMR (100 MHz, CDCl , δ
530−582.
=
3
(5) Fraile, J. M.; García, J. I.; Mayoral, J. A. Chem. Rev. 2009, 109,
ppm) 164.0, 163.9, 136.4, 128.9, 71.82, 71.81, 70.05, 70.04, 41.3, 35.4,
−
1
360−417.
3
(
2.3, 32.2, 18.6, 18.5, 17.8, 17.7; IR (CN) ν 1665.2 cm ; HR-MS
+
(6) Fraile, J. M.; García, J. I.; Herrerias, C. I.; Mayoral, J. A.; Pires, E.
Chem. Soc. Rev. 2009, 38, 695−706.
ESI+) m/z = 579.3931 [MH ], calcd for C H N O + H 579.3910.
34
50
4
4
General Procedure for Henry Reactions Promoted by
(7) Chiral Catalyst Immobilization and Recycling; De Vos, D. E.,
Monotopic Azabis(oxazoline) or Bis(oxazoline)−copper Com-
plexes. A solution of Cu(AcO) (9.07 mg, 0.05 mmol) and the
Vankelecom, I. F. J., Jacobs, P. A., Eds. De Vos/Chiral; Wiley-VCH
2
corresponding monotopic ligand (4a,4b, 5a, or 5b) (0.055 mmol) in 2
mL of anhydrous dichloromethane was stirred at room temperature
for 30 min. Then the mixture was microfiltered to eliminate the
Verlag GmbH: Weinheim, 2007.
(8) Rechavi, D.; Lemaire, M. Chem. Rev. 2002, 102, 3467−3494.
(9) Hallett, J. P.; Welton, T. Chem. Rev. 2011, 111, 3508−3576.
(10) Gruijters, B. W. T.; Broeren, M. A. C.; van Delft, F. L.; Sijbesma,
R. P.; Hermkens, P. H. H.; Rutjes, F. P. J. T. Org. Lett. 2006, 8, 3163−
remaining Cu(OAc) . Afterward, the dichloromethane was removed
2
under vacuum, and 2 mL of anhydrous isopropyl alcohol together with
the corresponding aldehyde (1 mmol) and nitro derivative (10 mmol)
were added. The reaction solution was stirred at room temperature for
3
166.
11) Kim, D.-W.; Lim, S.-G.; Jun, C.-H. Org. Lett. 2006, 8, 2937−
940.
12) Shi, L.; Wang, X.; Sandoval, C. A.; Li, M.; Qi, Q.; Li, Z.; Ding, K.
Angew. Chem., Int. Ed. 2006, 118, 4214−4218.
13) Chen, S.-W.; Kim, J. H.; Song, C. E.; Lee, S.-G. Org. Lett. 2007,
, 3845−3848.
14) García, J. I.; Lopez-Sanchez, B.; Mayoral, J. A. Org. Lett. 2008,
0, 4995−4998.
15) García, J. I.; Herrerias, C. I.; Lopez-Sanchez, B.; Mayoral, J. A.;
Reiser, O. Adv. Synth. Catal. 2011, 353, 2691−2700.
16) Aldea, L.; Delso, I.; Hager, M.; Glos, M.; García, J. I.; Mayoral, J.
A.; Reiser, O. Tetrahedron 2012, 68, 3417−3422.
17) Palomo, C.; Oiarbide, M.; Mielgo, A. Angew. Chem., Int. Ed.
004, 43, 5442−5444.
18) Palomo, C.; Oiarbide, M.; Laso, A. Eur. J. Org. Chem. 2007,
(
2
(
2
4 h, and then it was filtered through a silica pad to eliminate the
catalyst. After that, the silica pad was washed with dichloromethane,
and the resulting solution was concentrated under vacuum. Yield was
determined by H NMR using mesitylene as internal standard.
1
(
9
(
1
Enantiomeric excesses were determined by HPLC using a chiral
column. Specific chromatographic conditions, retention times, and
some typical chromatograms are included in the Supporting
Information.
(
General Procedure for Henry Reactions Promoted by
Ditopic Azabis(oxazoline) or Bis(oxazoline)−copper Com-
(
plexes. A solution of Cu(AcO) (9.07 mg, 0.05 mmol) and the
2
corresponding ditopic ligand (1 or 2) (0.055 mmol) in 4 mL of
anhydrous isopropyl alcohol was stirred at room temperature for 30
min. Afterward, the corresponding aldehyde (1 mmol) and nitro
derivative (10 mmol) were added. The resulting solution was stirred at
room temperature for 24 h and then concentrated under vacuum. The
(
2
(
2007, 2561−2574.
residue was extracted with an anhydrous mixture of hexane/Et O (1:1)
(19) Takemoto, Y. Chem. Pharm. Bull. 2010, 58, 593−601.
(20) Alvarez-Casao, Y.; Marques-Lopez, E.; Herrera, R. P. Symmetry
2011, 3, 220−245.
2
(
3 × 2 mL) in order to separate the products from the solid polymer
which had already precipitated. The polymer was then dried under
argon atmosphere. Under these conditions, the catalyst was ready to
be used again in a new reaction by adding new portions of solvent and
reagents. The product solution was concentrated under vacuum, and
the determination of the results was carried out as described
previously.
(21) Blay, G.; Hernan
1211.
́
dez-Olmos, V.; Pedro, J. Synlett 2011, 1195−
(22) Bhatt, A. P.; Pathak, K.; Jasra, R. V.; Kureshy, R. I.; Khan, N.-U.
H.; Abdi, S. H. R. J. Mol. Catal. A 2006, 244, 110−117.
(23) Trost, B. M.; Yeh, V. S. C. Angew. Chem., Int. Ed. 2002, 41, 861−
8
63.
ASSOCIATED CONTENT
Supporting Information
NMR spectra of the i-PrDiBox (2) ligand and typical HPLC
conditions and chromatograms of selected Henry reactions.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(24) Bulut, A.; Aslan, A.; Dogan, O. J. Org. Chem. 2008, 73, 7373−
■
7
375.
*
S
(
2
(
1
25) Park, J.; Lang, K.; Abboud, K. A.; Hong, S. J. Am. Chem. Soc.
008, 130, 16484−16485.
26) Kogami, Y.; Nakajima, T.; Ikeno, T.; Yamada, T. Synthesis 2004,
947−1950.
27) Kowalczyk, R.; Kwiatkowski, P.; Skarze
(
̇
wski, J.; Jurczak, J. J. Org.
Chem. 2009, 74, 753−756.
(28) Zulauf, A.; Mellah, M.; Schulz, E. J. Org. Chem. 2009, 74, 2242−
2245.
AUTHOR INFORMATION
Corresponding Author
E-mail: jig@unizar.es.
■
*
(29) Lang, K.; Park, J.; Hong, S. J. Org. Chem. 2010, 75, 6424−6435.
(
(
30) Xu, H.; Wolf, C. Synlett 2010, 2765−2770.
Notes
31) Zhou, Y.; Dong, J.; Zhang, F.; Gong, Y. J. Org. Chem. 2011, 76,
The authors declare no competing financial interest.
5
(
88−600.
32) Panov, I.; Drabina, P.; Padel
Org. Chem. 2011, 76, 4787−4793.
̌
̊
̌ ́ ́
kova, Z.; Simunek, P.; Sedlak, M. J.
ACKNOWLEDGMENTS
This work was made possible by the financial support of the
Spanish Ministerio de Economia y Competitividad (project
■
(33) Christensen, C.; Juhl, K.; Jørgensen, K. A. Chem. Commun.
2001, 2222−2223.
́
5
531
dx.doi.org/10.1021/jo3005974 | J. Org. Chem. 2012, 77, 5525−5532

The Journal of Organic Chemistry
Article
(
34) Christensen, C.; Juhl, K.; Hazell, R. G.; Jørgensen, K. A. J. Org.
Chem. 2002, 67, 4875−4881.
35) Evans, D. A.; Seidel, D.; Rueping, M.; Lam, H. W.; Shaw, J. T.;
(
Downey, C. W. J. Am. Chem. Soc. 2003, 125, 12692−12693.
(
(
36) McManus, H. A.; Guiry, P. J. Chem. Rev. 2004, 104, 4151−4202.
37) Desimoni, G.; Faita, G.; Jørgensen, K. A. Chem. Rev. 2006, 106,
3
(
(
561−3651.
38) Hargaden, G. C.; Guiry, P. J. Chem. Rev. 2009, 109, 2505−2550.
39) Mayani, V. J.; Abdi, S. H. R.; Kureshy, R. I.; Khan, N.-U. H.;
Das, A.; Bajaj, H. C. J. Org. Chem. 2010, 75, 6191−6195.
40) Khan, N.-U. H.; Prasetyanto, E. A.; Kim, Y.-K.; Ansari, M. B.;
Park, S.-E. Catal. Lett. 2010, 140, 189−196.
41) Zhou, Z.-M.; Li, Z.-H.; Hao, X.-Y.; Zhang, J.; Dong, X.; Liu, Y.-
Q.; Sun, W.-W.; Cao, D.; Wang, J.-L. Org. Biomol. Chem. 2012, 10,
113.
42) Li, Z.-H.; Zhou, Z.-M.; Hao, X.-Y.; Zhang, J.; Dong, X.; Liu, Y.-
Q.; Sun, W.-W.; Cao, D. Appl. Catal., A 2012, 425−426, 28−34.
43) Rasappan, R.; Olbrich, T.; Reiser, O. Adv. Synth. Catal. 2009,
51, 1961−1967.
44) Didier, D.; Magnier-Bouvier, C.; Schulz, E. Adv. Synth. Catal.
(
(
2
(
(
3
(
2
(
011, 353, 1087−1095.
45) Ginotra, S. K.; Singh, V. K. Org. Biomol. Chem. 2007, 5, 3932.
(
46) Complete information about the procedure and characterization
of this new chiral ligand is included in the Experimental Section and in
the Supporting Information.
(
47) Nano, A.; Brelot, L.; Rogez, G.; Maisse-Franc o̧ is, A.; Bellemin-
Laponnaz, S. Inorg. Chim. Acta 2011, 376, 285−289.
(
(
(
48) Yuryev, R.; Liese, A. Synlett 2009, 2589−2592.
49) Spangler, K. Y.; Wolf, C. Org. Lett. 2009, 11, 4724−4727.
50) Fraile, J.; García, J.; Herrerías, C.; Mayoral, J.; Reiser, O.;
Socuel
51) Gissibl, A.; Finn, M. G.; Reiser, O. Org. Lett. 2005, 7, 2325−
328.
52) Gissibl, A.; Padie,
Cuevas-Yanez, E.; Turrin, C.-O.; Caminade, A.-M.; Majoral, J.-P.;
Reiser, O. Org. Lett. 2007, 9, 2895−2898.
53) Werner, H.; Vicha, R.; Gissibl, A.; Reiser, O. J. Org. Chem. 2003,
8, 10166−10168.
54) Castillo, M. R.; Castillon
Martín, M.; Mayoral, J. A.; Sola, E. Tetrahedron 2011, 67, 5402−5408.
́
lamos, A.; Werner, H. Chem.Eur. J. 2004, 10, 2997−3005.
(
2
(
́
C.; Hager, M.; Jaroschik, F.; Rasappan, R.;
̃
(
6
(
́
, S.; Claver, C.; Fraile, J. M.; Gual, A.;
5
532
dx.doi.org/10.1021/jo3005974 | J. Org. Chem. 2012, 77, 5525−5532