Immobilisation of bis(oxazoline)-copper complexes on clays and nanocomposites. Influence of different parameters on activity and selectivity

Article abstract of DOI:10.1039/b203849h

Bis(oxazoline)-copper complexes, immobilised by electrostatic interactions on Laponite and Nafion/silica nanocomposites, have been tested as catalysts in the cyclopropanation of styrene with ethyl diazoacetate. Several factors play a decisive role in the

Full text of DOI:10.1039/b203849h

View Article Online / Journal Homepage / Table of Contents for this issue
Immobilisation of bis(oxazoline)–copper complexes on clays and
nanocomposites. Influence of different parameters on activity and
selectivity{
a
Jos e´ M. Fraile, Jos e´ I. Garc ´ı a, Mark A. Harmer, Clara I. Herrer ´ı as,
a
b
a
a
Jos e´ A. Mayoral,* Oliver Reiser and Heiko Werner
c
c
a
Departamento de Qu ´ı mica Org a´ nica, Instituto de Ciencia de Materiales de Arag o´ n, Facultad
de Ciencias, Universidad de Zaragoza-C.S.I.C., E-50009 Zaragoza, Spain
Central Research and Development, DuPont Company, Wilmington, Delaware, USA
b
c
Department of Chemistry, Institut f u¨ r Organische Chemie, Regensburg University, Germany
Received 19th April 2002, Accepted 7th June 2002
First published as an Advance Article on the web 3rd October 2002
Bis(oxazoline)–copper complexes, immobilised by electrostatic interactions on Laponite and Nafion/silica
nanocomposites, have been tested as catalysts in the cyclopropanation of styrene with ethyl diazoacetate.
Several factors play a decisive role in the outcome of the reaction. Firstly, the nature of the solid counterion is
important, with perfluorosulfonic solids being the best in this regard. Secondly, the nature of the solvent used
has a marked influence as some solvents cause the solid support to modify the stereochemical course of the
reaction. Finally, the nature of the chiral ligand is also very important, in particular the strength of its copper
complex, in order to avoid the formation of non-chiral catalytic copper centres. This particular effect has been
highlighted by using a ligand that is able to give strong complexation, namely an iminobis(oxazoline) ligand,
and this leads to the best enantioselectivities being obtained for reactions using electrostatically immobilised
catalysts.
4
,5
5–7
inorganic
and organic
supports. Bis(oxazoline)–copper
3
Introduction
complexes, which are useful catalysts for several reactions, are
cationic in nature and, therefore, they can also be immobilised
by formation of ion pairs with anionic supports.
Homogeneous metal complexes with chiral ligands are widely
used as enantioselective catalysts. However, from an industrial
point of view, catalysts that are not soluble in the reaction
medium are easier to handle and separate. Furthermore, in
some cases they can be recovered and reused. One of the
strategies most frequently used to prepare chiral heterogeneous
7
–9
In this
paper we analyse the influence of several factors on the catalytic
performance of the resulting solids. The factors investigated are
the nature of the support, the structure of the chiral ligand, and
the reaction conditions. In order to perform this investigation
we selected the cyclopropanation of styrene (1) with ethyl
diazoacetate (2) as the benchmark reaction (Scheme 1).
1
catalysts is the immobilisation of the homogeneous complex.
Depending on the nature of the support, either an organic
polymer or an inorganic solid, new organic or hybrid chiral
materials can be obtained and these can be useful catalysts
in enantioselective organic reactions.
Results and discussion
In most cases, immobilisation is carried out by covalent
bonding of the chiral ligand to an organic or inorganic support.
The method used to form this covalent bond has a decisive
influence on the catalytic activity and selectivity of the system.
In particular, marked effects on the enantioselectivity have
Although bis(oxazoline) 7, bearing tert-butyl groups, is the best
ligand for Cu-catalysed cyclopropanation reactions, bis(ox-
azoline)s bearing other substituents (5 and 6) were tested in
1
been described. These changes are generally accounted for in
terms of the conformational changes experienced by the ligand
through its interaction with the support. In this regard, immo-
2
bilisation without a covalent bond, a process that avoids the
chemical modification of the ligand, is an interesting alter-
native. However, very few studies have been undertaken to
analyse the influence of different factors on the results of
reactions promoted by these catalysts.
Bis(oxazoline)s are a well-known family of chiral ligands that
are able to form stable complexes with a large variety of metals,
and these complexes efficiently promote a large number of
3
enantioselective organic reactions in the homogeneous phase.
Due to this fact, immobilised bis(oxazoline) complexes are, in
principle, versatile chiral heterogeneous catalysts, which has
led to increased interest in the immobilisation of such systems.
Bis(oxazoline)s have been covalently bonded to insoluble
{Basis of a presentation given at Materials Discussion No. 5, 22–25
September 2002, Madrid, Spain.
Scheme 1
3
290
J. Mater. Chem., 2002, 12, 3290–3295
This journal is # The Royal Society of Chemistry 2002
DOI: 10.1039/b203849h

View Article Online
Table 1 Characterisation of the bis(oxazoline)–Cu(II
immobilized on Laponite
)
complexes
a
Surface
area/
d001/A m g
2
1
21
Cu/mmol g
Ligand as prepared
Cu/mmol g
after 1 run
2
21
˚
2N : Cu
5
6
7
0.35
0.28
0.11
0.72
0.22
0.25
0.10
—
1.10
1.20
1.30
—
17.3
17.8
15.6
13.8
207
225
218
290
b
c
—
a
Prepared by cation exchange of the bis(oxazoline)–Cu(OTf)
2
com-
c
plex in methanol with the sodium form of the Laponite. Non-chiral
b
2
solid obtained by cation exchange of CuCl in water. Broad signal.
order to assess the generality of the immobilisation method
prior to its application to different enantioselective reactions.
Previous studies using several clays as supports showed
8
that Laponite leads to the best results. Laponite is a poorly
laminated clay and has a larger proportion of external
exchange sites in comparison with other clays that may
account for the better accessibility of the catalytic sites and
hence for its better performance.
Fig. 2 IR spectra of the 6–Cu(II) complex (a) in solution, (b) exchanged
in Laponite, and (c) exchanged in SAC-40 Nafion/silica.
almost square planar structure. In this arrangement the surface
of the clay acts as the counterion of the cationic complex, thus
confirming that the complex is immobilised by electrostatic
Bis(oxazoline)–Cu(II) complexes were immobilised on Lapo-
nite by cation exchange in methanol using the sodium form
of the clay. Elemental analysis (Table 1) indicates that the
amount of exchanged complex depends on the structure of the
10
interactions with the anionic support.
The Laponites containing the three different bis(oxazoline)–
Cu(II) complexes were tested as catalysts in the benchmark
cyclopropanation reaction. Table 2 compares the results
obtained in these reactions with those observed using the
respective homogeneous catalysts. In the homogeneous phase
the 5–Cu(II) complex (with triflate as the counterion) leads to
2
1
bis(oxazoline) and ranges from about 0.3 mmol g with 5
2
1
and 6 to about 0.1 mmol g with 7. As a consequence of the
Laponite structure, XRD patterns (Fig. 1) show very broad
reflections and it is not easy to obtain conclusions from them.
In spite of this, it seems that 5– and 6–Cu(II) show expanded
basal spacings, which indicates that part of the complex is
accommodated in the interlamellar space. The broad X-ray
signal obtained after the exchange of 7–Cu(II) does not allow us
to draw any conclusions, although the lack of intercalation is
also in agreement with the low copper content in 7–Cu(Lap).
The introduction of large cations accounts for the experimen-
tally observed decrease in the surface area from values of
5
8% yield with a 2.1 : 1 trans : cis ratio and 59% and 53%
enantiomeric excess (ee) in the trans and cis products, respec-
tively. The same complex, when immobilised on Laponite,
yields 25% of cyclopropane with equimolecular amounts of
trans and cis products, and with only 24% and 19% ee,
respectively. These values are even worse when the catalyst is
recovered by filtration and reused. When bis(oxazoline) 6 is
used as the chiral ligand in the homogeneous phase, a yield of
2
21
around 300 m g in the starting and Cu-exchanged Laponites
2
21
to 200–225 m
Laponites (Table 1).
g
for the bis(oxazoline)–Cu exchanged
3
3% is obtained in a 2 : 1 trans : cis ratio and in 60% and 51%
ee, respectively. The yield, trans/cis selectivity and ee in the
trans product decrease only slightly in the reaction promoted
by the corresponding Laponite-immobilised complex. On the
other hand, the cis-cyclopropanes are obtained with only 24%
ee. In this case the recovered catalyst leads to virtually the same
results. In the homogeneous phase, the best results (72% yield,
and 94% and 91% ee in the trans- and cis-cyclopropanes,
respectively) are obtained with the 7–Cu(II) complex. However,
The structures of the exchanged complexes were confirmed
by IR spectroscopy (Fig. 2). The spectra show the CLN band
2
1
at around 1640 cm for the 6–Cu(II) complex. The structure
of the 5–Cu(II) complex in Laponite was also studied by EPR
and EXAFS. The results of these experiments indicate that
exchanged copper complexes are isolated species with an
Table 2 Results obtained from the cyclopropanation reaction at 25 uC
with homogeneous and Laponite-supported catalysts
a
d
Ee (%)
trans : cis
(3 : 4)
b
c
c
Catalyst
Cu (%)
t/h Yield (%)
trans cis
5
5
5
6
6
6
7
7
7
a
–Cu–(OTf)
–Cu–(Lap)e
–Cu–(Lap)
2
10
1
—
7
58
2.1
1.0
1.1
2.1
1.6
1.4
2.4
1.8
1.4
59
24
14
60
49
43
94
69
43
53
19
12
51
24
19
91
64
37
22 25
22 32
5
21 30
24 28
7
e
–Cu–(OTf)
–Cu–(Lap)
2
10
33
0.84
e
e
—
–Cu–(Lap)
–Cu–(OTf)
–Cu–(Lap)
2
10
72
0.33
39 30
88 26
e
e
—
–Cu–(Lap)
Reactions carried out in dichloromethane using a 1 : 1 styrene :
b
diazoacetate ratio. Proportion of Cu used in the reaction with
c
regard to the reagents. Determined by GC at total conversion of
d
ethyl diazoacetate (ref. 9b). Determined by GC using a Cyclodex-B
Fig. 1 Low-angle (2h) X-ray diffraction patterns of (a) Laponite, (b)
7–Cu(II) complex exchanged in Laponite, and (c) 6–Cu(II) complex
exchanged in Laponite.
e
column (ref. 9b); 3R and 4R are the major enantiomers. Using
recovered catalyst.
J. Mater. Chem., 2002, 12, 3290–3295
3291

View Article Online
a
Table 3 Results obtained from the cyclopropanation reaction at 25 uC with homogeneous and Laponite-supported catalysts in different solvents
c
DDG{d
Catalyst
Solvent
Yield (%)
trans : cis (3 : 4)
Ee (%)
trans
cis
trans
cis
6
6
6
6
6
6
6
6
7
7
7
a
–Cu–(OTf)
–Cu–(OTf)
2
2
CH
Styrene
CH Cl
Styrene
CH Cl
Styrene
CH Cl
n-Hexane
2
Cl
2
33
41
19
59
30
40
27
10
72
30
9
2.1
2.2
1.6
1.6
1.6
0.4
1.5
0.4
2.4
1.8
1.0
60 (3R)
55 (3R)
17 (3R)
11 (3R)
49 (3R)
7 (3R)
45 (3R)
3 (3R)
94 (3R)
69 (3R)
27 (3R)
51 (4R)
42 (4R)
13 (4R)
9 (4R)
24 (4R)
34 (4S)
22 (4R)
33 (4S)
91 (4R)
64 (4R)
8 (4R)
—
—
—
—
0.63
0.08
—
0.02
—
—
—
—
—
0.29
–0.42
—
–0.41
—
0.90
0.10
–Cu–Cl
–Cu–Cl
2
2
2
2
–Cu–(Lap)
–Cu–(Lap)
–Cu–(Lap)
–Cu–(Lap)
2
2
b
2
2
–Cu–(OTf)
–Cu–(Lap)
–Cu–(Lap)
2
CH
CH
2
Cl
Cl
2
2
2
1.00
0.33
b
n-Hexane
Reactions carried out in dichloromethane using a 1 : 1 styrene : diazoacetate ratio, except when styrene is used as the solvent. Catalyst recovered
c
from the reaction carried out in styrene. Determined by GC using a Cyclodex-B column (ref. 9b). DG
d
{
{
21
S R
2 DG ; values in kcal mol .
the results are noticeably worse (30% yield, and 69% and
4% ee) when this complex is immobilised on Laponite. These
Cu(I)–carbene intermediate. This intermediate can either react
6
with a diazoacetate molecule, leading to dimeric by-products,
or the carbene can be directly inserted into the styrene double
bond to yield the cyclopropanes. The stereochemical course of
the reaction, i.e. the trans–cis distribution and enantioselec-
values are even lower (26% yield, and 43% and 37% ee) on using
the recovered catalyst. Analysis of the recovered catalysts
shows that only in the case of 5–Cu(II) is there a noticeable
decrease in the copper content. After the addition of diazo-
acetate, the catalyst was separated by filtration and the reaction
in the homogeneous phase did not continue, even when an
additional amount of diazoacetate was added. This experiment
clearly shows the heterogeneous character of the reaction.
The low yield obtained in the reaction is related to the
concurrence of the carbene intermediate dimerization process,
which affords maleate and fumarate. For this reason styrene is
normally used either in excess or, in some cases, even as the
solvent. In the homogeneous phase the use of styrene, rather
than dichloromethane, as the solvent in the reaction catalysed
1
2
tivities, is determined in this insertion step. The four possible
transition states, leading to four cyclopropanes, are schema-
tically represented in Fig. 3 for the bis(oxazoline) 6. In this
projection, the carbene–Cu(
I
) complex (CuLCH–COOEt) is
represented by the central circle bonded to the ester group,
E. The two oxazoline rings are almost coplanar and they have
been represented as a line nearly perpendicular to the E–Ca
bond. The phenyl groups correspond to the substituents of the
1
2
oxazoline rings, with the relative stereochemistry calculated
for the intermediate. Finally, the figure represents the incoming
styrene and shows the more important steric interactions either
with the ester group (always present) or with the Laponite
surface (present only in the heterogeneous phase). The selec-
tivities reflect the relative energies of these transition states. In
the homogeneous phase, the trans-preference (3) is explained
by the higher energy of the cis-transition states as a conse-
quence of the steric interaction between the phenyl group
of styrene and the ester group of the carbene (Fig. 3). The
preference for the 3R and 4R cyclopropanes over their
enantiomers 3S and 4S, is due to the steric interaction between
2
by 6–Cu(OTf) increases the final yield without modifying
either the trans–cis product distribution or the enantioselec-
tivities (Table 3). The same effect is observed in the reaction
catalysed by 6–CuCl
are lower as a result of the previously described effect of the
2
; in this case the enantioselectivities
1
1
counterion. When styrene is used as the solvent in the
presence of the Laponite catalyst, the yield increases from 30%
to 40%, although a dramatic effect on the selectivities is also
observed. Under these conditions the less stable cis-cyclopro-
pane is obtained preferentially and the enantioselectivity in the
trans products is noticeably reduced (from 49% to 7% ee).
In addition, the direction of the asymmetric induction in cis-
cyclopropanes is reversed in that it changes from 24% ee in 4R
to 34% ee in 4S. When the catalyst used in styrene is recovered
and reused in dichloromethane, the results are the same as
those obtained originally, showing that the changes in selec-
tivity are not due to the catalyst being permanently affected by
the reaction.
In an attempt to analyse the origin of this effect, the reaction
was carried out in hexane—a solvent of low polarity that is
more similar to styrene than to dichloromethane. Under these
conditions the yield decreases to 10%, but the trans–cis distri-
bution and the enantioselectivities are the same as those
obtained using styrene. Finally, we wondered if this effect was
exclusive to ligand 6 and we therefore compared the results for
the Laponite-containing complex 7–Cu(II) in reactions carried
out in dichloromethane and hexane. The results show the
same trend as before; i.e. an increase in the amount of cis-
cyclopropanes and in products 3S and 4S, but the greater
steric interaction of the tert-butyl group precludes inversion of
configuration in the major products.
The cyclopropanation reaction is catalysed by the bis-
oxazoline)–Cu(I) complex, which is obtained by in situ
(
reduction of the Cu(II) complex with the diazoacetate. From
Fig. 3 Proposed model of the steric interactions responsible for the
stereoselectivities observed both in homogeneous and heterogeneous
phase reactions catalysed by the 6–Cu(II) complex.
the Cu(
I
) complex the rate limiting step is the formation of a
3
292
J. Mater. Chem., 2002, 12, 3290–3295

View Article Online
the ester group of the diazoacetate and the R-substituent of the
bis(oxazoline), an interaction that is more important in the
transition states leading to the 3S and 4S products. Similar
behaviour is also observed with the Laponite-supported
catalysts in dichloromethane. However, when the reaction is
carried out in a solvent with a low relative permittivity, the
electrostatic attraction between the cationic catalytic species
and the clay support increases, meaning that the clay sheet
is closer to the intermediate, playing an active role in the
Table 4 Characterisation of the bis(oxazoline)–Cu(II
immobilized on Nafion/silica
)
complexes
Cu/
mmol g
as prepared
Cu/
mmol g
after 1 run
Surface
area/
m g
21
21
a
2
21
Catalyst
2N : Cu
b
5
6
6
7
7
–Cu–(SAC-13)
–Cu–(SAC-13)
–Cu–(SAC-40)
–Cu–(SAC-13)
–Cu–(SAC-40)
0.05
0.08
0.37
0.07
0.27
0.31
0.03
0.08
0.35
0.03
0.26
0.29
—
—
82
84
377
88
395
370
b
1.00
b
—
stereochemical course of the reaction. When the Cu(I)–carbene
0.37
1.43
intermediate is close to the clay surface, the orientation in
which the ester group points away from the surface will be
favoured. As styrene approaches this intermediate, not only the
previously mentioned steric interactions are important, but the
interactions with the clay surface will also play a role. These
latter interactions disfavour the formation of trans-cyclopro-
panes and they work in the opposite sense to the former
interactions in asymmetric induction. Under these conditions
the selectivity will depend on the balance between several sets
of interactions. When R ~ Ph, the intermolecular interactions
are weaker and the interactions with the surface dominate
the stereochemical course of the reaction, leading to 4S as
8–Cu–(SAC-40)
a
Prepared by cation exchange of the bis(oxazoline)–Cu(OTf)
2
b
com-
plex in methanol with the sodium form of the Nafion/silica. The N
content is too low for a reliable determination.
Table 5 Results obtained from the cyclopropanation reaction at 25 uC
with the Nafion/silica catalysts
a
b
Ee (%)
trans : cis
Catalyst
Cu (%) t/h Yield (%) (3 : 4)
trans cis
t
5–Cu–(SAC-13)
0.30
0.30
2.2
21 34
46 23
1.6
1.6
1.9
1.9
1.8
1.8
1.5
1.6
1.5
1.5
2.1
1.9
1.9
49
37
57
56
62
57
23
14
18
9
47
36
46
45
50
49
19
14
16
11
84
80
88
the major product. On the other hand, when R ~ Bu, the
c
5–Cu–(SAC-13)
6–Cu–(SAC-13)
6–Cu–(SAC-13)
intermolecular interactions are stronger and the interactions
with the surface cannot reverse the stereochemical course of
the reaction, so that although the amount of 4S is increased,
it is not the major product. In order to compare the order of
magnitude of these interactions, the difference in activation free
9
9
9
9
2
2
2
2
27
27
28
30
28
30
37
31
c
6–Cu–(SAC-40)c
6–Cu–(SAC-40)
7–Cu–(SAC-13)
7–Cu–(SAC-13)
0.42
1.6
c
{
energies (DDG ) between the two trans and the two cis transi-
tion states are given in Table 3. When R ~ Ph, a change of
7
7
8
–Cu–(SAC-40)
–Cu–(SAC-40)
–Cu–(SAC-40)
c
{
solvent leads to a modification of the DDG values in the range
0
d
0.93
0.42
24 20
24 21
24 20
90
88
91
2
.55–0.70 kcal mol . These changes are even more important
1
c,d
e
8–Cu–(SAC-40)
7–Cu–(SAC-40)
t
21
when R ~ Bu (0.67–0.80 kcal mol ), in spite of the fact
that the stereochemical course of the reaction is not reversed.
These results prove that the support plays an active role in the
reaction and that the stereochemistry is modified when the
reaction is carried out in the confined space of the clay.
Another important point to consider is the influence of the
nature of the support on the course of the reaction. In the
homogeneous phase, the nature of the counterion has a
a
Reactions carried out in 1,2-dichloroethane (SAC-13) or dichloro-
methane (SAC-40) using a 1 : 1 styrene : diazoacetate ratio. Deter-
b
mined by GC using a Cyclodex-B column (ref. 9b); 3R and 4R are
the major enantiomers. Recovered catalyst. It is necessary to heat
c
d
e
the reaction until the reduction of Cu(II) takes place. After adding
0
.042 eq. of 7.
1
1
decisive influence on the enantioselectivity and, in this
regard, Laponite seems to be worse than triflate{. Nafion, a
perfluorosulfonic resin, has anionic groups similar to triflate
but it also has a very low surface area. To overcome this
drawback we used Nafion/silica nanocomposites—solids with
when 5–Cu(II) is supported on SAC-13, the recovered catalyst
is less active and less enantioselective. Unfortunately, the use
of ligand 7, which gives the best results in the homogeneous
phase, leads to very low enantioselectivities (23% and 19% ee).
These results are even worse when the catalyst is recovered and
reused. Elemental analyses of the mother liquors after catalyst
synthesis suggest that the unexpectedly low enantioselectivity
is due to the presence of non-chiral copper sites, which are
able to promote a concurrent non-enantioselective reaction.
The use of SAC-40 as a support allowed us to confirm this
hypothesis. Whereas the complex 6–Cu(II), when immobilised
on SAC-40, has the N : Cu ratio expected for a bis(oxazoline)–
Cu complex, the immobilisation of 7–Cu(II) on the same sup-
port led to a solid with a very low nitrogen content and a
lower N : Cu ratio, a situation consistent with the presence of
exchanged Cu(II) that is not complexed by the chiral ligand
(Table 4). When these solids were tested as catalysts in the
cyclopropanation reaction, the results obtained were found to
be in agreement with this observation. The 6–Cu(II) complex
immobilised on SAC-40 gave the same yield and enantioselec-
tivities (62% and 50% ee) as obtained in the homogeneous
phase and, furthermore, the catalyst could be recovered and
reused with the same results. In comparison with the same
complex immobilised on SAC-13, the only advantage of the
nanocomposite (with the higher degree of functionalisation and
1
3
acidic properties that have also been used as supports for
4
1
organometallic compounds. These supports have a lower
degree of functionalisation than Nafion, but they do have
1
5
higher surface areas. We used two different Nafion/silica
nanocomposites with different degrees of functionalisation and
surface areas. Catalysts were modified by cation exchange
in methanol of the bis(oxazoline)–Cu(II) complex with the
sodium form of Nafion.
Analyses of the resulting solids (Table 4) indicate that the
immobilisation in SAC-40 (40 wt% Nafion/silica nanocompo-
site) leads to a catalyst with a higher copper content and sur-
face area, a situation consistent with the higher degree of
functionalisation and surface area of the support. The presence
of the complex was demonstrated by IR spectroscopy (Fig. 2).
From our previously gained knowledge of the influence of
the counterion, it is not unexpected that the use of SAC-13 (13
wt% Nafion/silica nanocomposite) increases the enantioselec-
tivities of the reactions promoted by 5–Cu(II) and 6–Cu(II
(
)
Table 5). In particular, the latter catalyst gives rise to enantio-
selectivities (57% and 46% ee) that are almost the same as those
obtained in the homogeneous phase, and furthermore, it can be
recovered and reused with the same results. On the other hand,
surface area) is a slightly better enantioselectivity. The 7–Cu(II
)
complex immobilised on SAC-40 also led to low enantioselec-
tivities (18% and 16% ee), again in agreement with the presence
of more than 50% of non-chiral copper centres.
{The IUPAC name for triflate is trifluoromethanesulfonate.
J. Mater. Chem., 2002, 12, 3290–3295
3293

View Article Online
namely the iminobis(oxazoline), led to the best enantioselec-
tivities found for this reaction with a supported catalyst.
Finally, the nature of the solvent can modify the strength of
the catalyst–support interaction, which not only modifies the
catalytic activity but also the stereochemical course of the
reaction as the support becomes closer to the catalytic centres.
In summary, we have shown that ion-pair formation is
a suitable strategy to immobilise cationic chiral catalysts,
provided that certain key factors are taken into account.
Scheme 2
Analyses of the SAC-immobilised catalysts after recovery
do not show detectable leaching of copper. Furthermore, the
above-described filtration experiment shows the absence of a
homogeneous reaction.
The results obtained show that it is possible to successfully
immobilise the 6–Cu(II) complex by electrostatic interactions
with anionic supports, but also that this strategy fails for the
immobilisation of the 7–Cu(II) complex. One possible explana-
tion for this surprising result is that the formation constant
Experimental
2
(
,2’-Isopropylidenebis[(4S)-4-benzyl-4,5-dihydro-1,3-oxazole]
5) and 2,2’-isopropylidenebis[(4S)-4-phenyl-4,5-dihydro-1,3-
oxazole] (6) were obtained from (S)-phenylalaninol and
1
7
(S)-phenylglycinol, respectively, as previously described. 2,2’-
Isopropylidenebis[(4S)-4-tert-butyl-4,5-dihydro-1,3-oxazole] (7)
was purchased from Aldrich. Bis[(4S)-4-tert-butyl-4,5-dihy-
dro-1,3-oxazol-2-yl]methylamine was prepared following the
method previously described. Laponite was obtained from
Laporte Adsorbents. 13 wt% Nafion/silica nanocomposite
of the 6–Cu(II) complex is greater than that of the 7–Cu(II
)
1
6
complex, meaning that all the chiral ligand remains bonded to
the copper after cation exchange in the case of 6. Scheme 2
shows a possible way in which the chiral ligand can leave the
complex and pass into the solution. The equilibria will be
shifted to right in the case of complexes with low formation
constants, such as 7–Cu(II).
9
b
(SAC-13) was prepared following a slight modification of the
1
original method.
5
Preparation of the 40 wt% Nafion/silica nanocomposite
SAC-40)
Bearing this possibility in mind, we decided to explore the
use of the iminobis(oxazoline) 8. Given the donor character of
the nitrogen atom in the central bridge, this ligand is expected
to form a stronger complex with Cu(II) and it is known that
these kinds of complexes efficiently promote the enantioselec-
(
A mixture of tetramethoxysilane (204 g), distilled water (33 g)
and 0.04 M HCl solution (3 g) was stirred for 45 min to give a
clear solution. To a Nafion solution (containing 5 wt% Nafion
in a mixed alcohol–water solution, from Aldrich) (1200 ml)
was added 0.4 M NaOH (150 ml) with stirring. The addition
was performed over 15 min and the resulting solution was
stirred for an additional 15 min period. The silicon-containing
solution was added rapidly to the stirred Nafion-containing
solution and, after 10–15 min, the resulting solution gelled. The
gel was placed in an oven and dried under a nitrogen flow
for about 2 d. The hard glass-like product was ground slightly
and passed through a 10-mesh screen. The material was then
stirred with 3.5 M HCl (1 l), filtered and washed with water.
This last process was repeated three times. The final solid was
1
6
tive cyclopropanation reaction in the homogeneous phase.
When the 8–Cu(II) triflate complex was exchanged in methanol
with the sodium form of SAC-40, the solid obtained had a 2N :
Cu ratio near the theoretical value of 1.5 (Table 4). When this
solid was used to promote the cyclopropanation reaction it was
found that the rate was quite slow and it was necessary to heat
the mixture to achieve the reduction of Cu(II) to Cu(I) by the
diazoacetate. This characteristic of the reaction is consistent
with strong complexation of the cation. Under these conditions
the reaction yield is rather low (20%), but the enantioselec-
tivities are very high (90% and 84% ee) and the recovered
catalyst exhibits the same behaviour (Table 5).
treated in HNO (20 wt%) at 80 uC for 3 h, washed with water
3
and then dried at 100 uC for 12 h.
The surface area, pore volume, and pore diameter were
Taking this result into account, it may be expected that the
enantioselectivity obtained with the 7–Cu(II) complex immo-
bilised on SAC-13 and SAC-40 would increase if the reaction
were carried out in the presence of an additional amount of
chiral ligand in solution, in order to shift the equilibrium to the
complex form. In fact, under these conditions the enantio-
selectivities observed are the same as those obtained in the
homogeneous phase (91% and 88% ee), although the chemical
yield decreases to 20%, a fact indicating that the non-
complexed Cu(II) is more active than the bis(oxazoline)–
Cu(II) complex. In order to confirm that this reaction is truly
heterogeneous, the filtration experiment described above was
carried out and further progress of the reaction was not
observed in solution.
2
21
3
21
˚
, and 8.9 A,
determined to be 468 m
respectively.
g
, 1.05 cm
g
Preparation of Laponite-exchanged catalysts
To a solution of bis(oxazoline) (1 mmol) in methanol (20 ml)
was added Cu(II) triflate (1 mmol ). Laponite (1 g) was slowly
added to the above solution and the resulting suspension was
stirred at room temperature for 24 h. The solid was separated
by filtration, thoroughly washed with methanol and then with
dichloromethane, and dried under a flow of air before use.
Preparation of Nafion/silica-exchanged catalysts
Prior to exchange, Nafion/silica nanocomposites were trans-
formed into their sodium forms by passing a solution of 2 M
NaCl through a column containing the solid until neutral pH
was obtained. The solids were then washed with deionised
water and dried under vacuum at 150 uC for about 4 h.
To a solution of bis(oxazoline) (1 mmol) in methanol (10 ml)
was added Cu(II) triflate (1 mmol). To this solution was
added the corresponding Nafion/silica (1.5 g) and the resulting
suspension was stirred at room temperature for 24 h. The solid
was separated by filtration, thoroughly washed with methanol
and then with dichloromethane, and dried under vacuum.
Conclusions
The results described show that bis(oxazoline)–copper catalysts
can be successfully immobilised by the formation of ion-pairs
with anionic supports. The results obtained depend on several
factors. Firstly, the coordinating ability of the anionic solid is
an important property and, in this regard, anions with low
basicity seem to be better. Another significant factor is the
nature of the chiral ligand, with the strength of complexation
being particularly important to avoid the formation of non-
chiral catalytic centres. The use of a more coordinating ligand,
3
294
J. Mater. Chem., 2002, 12, 3290–3295

View Article Online
Characterisation of the materials
3
(a) A. K. Gosh, P. Mathivanan and J. Capiello, Tetrahedron:
Asymmetry, 1998, 9, 1–45; (b) K. A. Jørgensen, M. Johannsen,
S. L. Yao, H. Audrain and J. Thorhauge, Acc. Chem. Res., 1999,
32, 605–613; (c) A. Pfaltz, Synlett, 1999, 835–842; (d) A. Pfaltz,
J. Heterocycl. Chem., 1999, 36, 1437–1451; (e) J. S. Johnson and
D. A. Evans, Acc. Chem. Res., 2000, 32, 325–335; (f) F. Fache,
E. Schultz, M. L. Tommasino and M. Lemaire, Chem. Rev., 2000,
Copper analyses were carried out by plasma emission spectro-
scopy on a Perkin-Elmer Plasma 40 emission spectrophoto-
meter. Nitrogen analyses were carried out on a Perkin-Elmer
2
40 elemental analyser. Step-scanned X-ray diffraction pat-
terns of oriented samples were collected at room temperature
from 3u in 2h up to 60u, using a D-max Rigaku system with a
rotating anode. The diffractometer was operated at 40 kV and
1
00, 2159–2231.
(a) D. Rechavi and M. Lemaire, Org. Lett., 2001, 3, 2493–2496;
b) J. K. Park, S.-W. Kim, T. Hyeon and M. Kim, Tetrahedron:
4
(
8
monochromator. Transmission FT-IR spectra of self-sup-
ported wafers evacuated (v10 Torr) at 50 uC were recorded
with a Mattson Genesis Series spectrophotometer. Surface
areas were determined by N2 adsorption (BET) using a
Micromeritics ASAP 2000 apparatus.
0 mA, and the Cu-Ka radiation was selected using a graphite
Asymmetry, 2001, 12, 2391–2395; (c) R. J. Clarke and J. J. Shannon,
Chem. Commun., 2001, 1936–1937.
M. I. Burguete, J. M. Fraile, J. I. Garc ´ı a, E. Garc ´ı a-Verdugo,
C. I. Herrer ´ı as, S. V. Luis and J. A. Mayoral, J. Org. Chem., 2001,
2
4
5
6
6
6, 8893–8901.
(a) M. I. Burguete, J. M. Fraile, J. I. Garc ´ı a, E. Garc ´ı a-Verdugo,
S. V. Luis and J. A. Mayoral, Org. Lett., 2000, 2, 3905–3908;
(
b) S. Orlandi, A. Mandoli, D. Pini and A. Salvadori, Angew.
Cyclopropanation reactions
Chem., Int. Ed., 2001, 113, 2587–2589; (c) K. Hallman and
C. Moberg, Tetrahedron: Asymmetry, 2001, 12, 1475–1478.
M. J. Fern a´ ndez, J. M. Fraile, J. I. Garc ´ı a, J. A. Mayoral,
M. I. Burguete, E. Garc ´ı a-Verdugo, S. V. Luis and M. A. Harmer,
Topics Catal., 2000, 13, 303–309.
J. M. Fraile, J. I. Garc ´ı a, J. A. Mayoral and T. Tarnai,
Tetrahedron: Asymmetry, 1998, 9, 3997–4008.
(a) J. M. Fraile, J. I. Garc ´ı a, J. A. Mayoral and T. Tarnai,
Tetrahedron: Asymmetry, 1997, 8, 2089–2092; (b) J. M. Fraile,
J. I. Garc ´ı a, J. A. Mayoral, T. Tarnai and M. A. Harmer, J. Catal.,
To a suspension of the appropriate catalyst (amounts of Cu
given in Tables 2 and 5) in a solution of styrene (5 mmol) and
n-decane (100 mg, internal standard) in dichloromethane or
n-hexane (5 ml), under an argon atmosphere, was added a
solution ethyl diazoacetate (2.5 mmol) in the same solvent
7
8
9
(
0.5 ml) during 2 h using a syringe pump. The reaction was
monitored by GC and, after complete consumption of the ethyl
diazoacetate, a second portion of this reagent was added in the
same way. After completion of the reaction, the catalyst was
separated by filtration, washed, dried, and reused following the
same method. The products of the reactions were identified by
1
999, 186, 214–221; (c) C. Langham, S. Taylor, D. Bethell,
P. McMorn, P. C. Bullman-Page, D. J. Willock, C. Sly,
F. E. Hancock, F. King and G. J. Hutchings, J. Chem. Soc.,
Perkin Trans. 2, 1999, 1043–1049; (d) G. J. Hutchings,
C. Langham, P. Piaggio, S. Taylor, M. McMorn, D. J. Willock,
D. Bethell, P. C. Bulman-Page, C. Sly, F. E. Hancock and F. King,
Stud. Surf. Sci. Catal., 2000, 120, 521–526; (e) S. Taylor, J. Gullick,
P. McMorn, D. Bethell, P. C. Bulman Page, F. E. Hanckock,
F. King and G. J. Hutchings, J. Chem. Soc., Perkin Trans. 2, 2001,
9
b
GC as described previously.
Acknowledgements
This work was made possible by the generous financial support
of the CICYT (project MAT99-1176) and the MCYT (Acci o´ n
Integrada HA2001-0096). C.I.H. is also indebted to the MCYT
for a grant.
1
714–1723; (f) S. Taylor, J. Gullick, P. McMorn, D. Bethell,
P. C. Bulman Page, F. E. Hanckock, F. King and G. J. Hutchings,
J. Chem. Soc., Perkin Trans. 2, 2001, 1724–1728.
P. J. Alonso, J. M. Fraile, J. Garc ´ı a, J. I. Garc ´ı a, J. I. Mart ´ı nez,
J. A. Mayoral and M. C. S a´ nchez, Langmuir, 2000, 16, 5607–5612.
J. M. Fraile, J. I. Garc ´ı a, J. A. Mayoral and T. Tarnai, J. Mol.
Catal. A, 1999, 144, 85–89.
J. M. Fraile, J. I. Garc ´ı a, V. Mart ´ı nez-Merino, J. A. Mayoral and
L. Salvatella, J. Am. Chem. Soc., 2001, 123, 7616–7625.
(a) M. Harmer, A. J. Vega, Q. Sun, W. E. Farneth, A. Heidekum
and W. F. Holderich, Green Chem., 2000, 7–14; (b) M. A. Harmer
and Q. Sun, Appl. Catal. A, 2001, 221, 45–62.
1
1
1
1
0
1
2
3
References
1
Chiral Catalysts Immobilization and Recycling, eds. D. E. De Vos,
I. F. J. Vankelecom and P. A. Jacobs, Wiley-VCH, Weinheim,
2
000.
(a) K. T. Wan and M. E. Davies, Nature, 1994, 370, 449–450;
b) M. J. Sabater, A. Corma, A. Domenech, V. Forn e´ s and
2
(
1
1
4
5
A. J. Seem, J. Mol. Catal. A, 2001, 177, 108–112.
M. A. Harmer, W. E. Farneth and Q. Sung, J. Am. Chem. Soc.,
H. Garc ´ı a, Chem. Commun., 1997, 1285–1286; (c) K. B. M. Jansen,
I. Laquiere, W. Dehaen, R. F. Parton, I. F. J. Vankelecom and
P. A. Jacobs, Tetrahedron: Asymmetry, 1997, 8, 3481–3487;
1
996, 118, 7708–7715.
1
1
6
7
M. Glos and O. Reiser, Org. Lett., 2000, 2, 2045–2048.
D. M u¨ ller, G. Umbritch, B. Weber and A. Pfaltz, Helv. Chim.
Acta, 1991, 74, 232–240.
(
d) J. M. Fraile, J. I. Garc ´ı a, J. Massam and J. A. Mayoral, J. Mol.
Catal. A, 1998, 136, 47–57; (e) T. Sento, S. Shimazu, N. Ichikumi
and T. Uematsu, J. Mol. Catal. A, 1999, 137, 263–267.
J. Mater. Chem., 2002, 12, 3290–3295
3295