Supported heteropolyanions as solid counterions for the electrostatic immobilization of chiral copper complexes

Article abstract of DOI:10.1016/j.jcat.2010.07.017

Supported (PW₁₂O₄₀^3-) is a suitable counterion to immobilize chiral azabis(oxazoline)-copper complexes through electrostatic interactions (ion pair formation). Catalytic results are good to excellent in cyclopropanation reactions (up to 97% ee). Partial leaching during recycling when mono-exchanged species are present, favored by strong coordination of by-products, can be minimized by a proper solvent choice and by optimization of the recycling system, allowing the recovery and reuse of the catalyst at least six times without loss of activity and enantioselectivities in the range of 82-96% ee for trans-cyclopropanes.

Full text of DOI:10.1016/j.jcat.2010.07.017

Journal of Catalysis 275 (2010) 70–77
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Supported heteropolyanions as solid counterions for the electrostatic
immobilization of chiral copper complexes
c,
M. Rosario Torviso a, Mirta N. Blanco b, Carmen V. Cáceres b, José M. Fraile c, , José A. Mayoral
*
**
a Departamento de Química Orgánica, Facultad de Farmacia y Bioquímica, Universidad de Buenos Aires, Junín 956, 1113 Buenos Aires, Argentina
b Centro de Investigación y Desarrollo en Ciencias Aplicadas ‘‘Dr. Jorge Ronco” (CINDECA), Departamento de Química, Facultad de Ciencias Exactas,
Universidad Nacional de La Plata, CCT-La Plata, CONICET, Calle 47 N° 257, 1900 La Plata, Argentina
c Departamento de Química Orgánica, Instituto de Ciencia de Materiales de Aragón and Instituto Universitario de Catálisis Homogénea, Facultad de Ciencias,
Universidad de Zaragoza – C.S.I.C., E-50009 Zaragoza, Spain
a r t i c l e i n f o
a b s t r a c t
Supported (PW12O340À) is a suitable counterion to immobilize chiral azabis(oxazoline)–copper complexes
through electrostatic interactions (ion pair formation). Catalytic results are good to excellent in
cyclopropanation reactions (up to 97% ee). Partial leaching during recycling when mono-exchanged
species are present, favored by strong coordination of by-products, can be minimized by a proper solvent
choice and by optimization of the recycling system, allowing the recovery and reuse of the catalyst at
least six times without loss of activity and enantioselectivities in the range of 82–96% ee for trans-
cyclopropanes.
Article history:
Received 7 June 2010
Revised 16 July 2010
Accepted 16 July 2010
Available online 17 August 2010
Keywords:
Heteropolyacids
Enantioselective catalysis
Immobilized catalysts
Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction
solids with a high acidity such as nafion-like solids [5]. However,
nafion itself has an extremely low surface area, whereas
The inherent practical advantages of heterogeneous over
homogeneous catalysts have boosted the research about immobili-
zation of chiral metal complexes to obtain heterogeneous catalysts
for enantioselective processes [1,2]. Non-covalent methods
represent an interesting alternative to covalent immobilization
[3], given that in many cases the same homogeneous complex
can be used for immobilization without any chemical modification,
and hence the preparation of the supported catalyst is made much
easier.
In immobilization methods based on electrostatic interactions,
the support acts as a counterion for the chiral complex, forming
a support–complex ion pair. It is well known the influence of the
counterion on the reaction results in different enantioselective
reactions. As an example, the stereochemistry of cyclopropanation
is modified due to the pyramidalization of the transition states
caused by the coordination of anions with strong coordinating abil-
ity [4]. For this reason, the search of new solid counterions, to be
used as supports for chiral catalysts, is an area of interest but
unfortunately very few explored. In most of the cases, anions with
a poor coordinating ability are needed, therefore coming from
composites are expensive and, in general, display a very low
functionalization.
In order to broaden the range of applicability of electrostatically
immobilized chiral catalysts, it is important to test other solids
with superacidic properties as supports. One example is the
Keggin-type heteropolyacid [6], that can be immobilized into
several supports, silica for instance, acting as acid catalysts in
several organic reactions [7–10]. The examples about the use of
these solids as supports of chiral catalysts are scarce. Pioneering
work by Augustine et al. [11–13] described the immobilization of
Rh-diphosphine complexes on different alumina-supported het-
eropolyanions and their use in hydrogenation reactions, with im-
proved activity and enantioselectivity [14,15], mainly upon
recycling. Ru-Binap [16] and Rh-prolinamide [17] complexes were
also immobilized by the same procedure. This methodology was
compared with the direct cation exchange on nafion-like solids,
with better activity and stability against leaching [18], in agree-
ment with a strong support–complex interaction. However, the
examples in the literature dealing with the immobilization of other
types of complexes or the application in enantioselective reactions
different from hydrogenation are scarce. Only in one case, the
immobilization of bis(oxazoline)–copper complexes has been de-
scribed with no particular success [19].
* Corresponding author. Fax: +34 976 76 20 77.
** Corresponding author.
E-mail address: jmfraile@unizar.es (J.M. Fraile).
Bis(oxazoline)–metal complexes constitute one family of
versatile enantioselective catalysts [20] widely studied in aspects
0021-9517/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2010.07.017

M. Rosario Torviso et al. / Journal of Catalysis 275 (2010) 70–77
71
concerning to immobilization [21,22] and specially with
non-covalent immobilization methods. In fact, bis(oxazoline)–cop-
per complexes immobilized by non-covalent methods have been
used as catalysts in cyclopropanation [5,23], Diels–Alder [24], azir-
idination [25], Mukaiyama aldol reactions [26], and carbene inser-
tions into C–H bonds [27]. Moreover, the support surface may play
an important role in the stereochemical course of the enantioselec-
tive reactions [28], demonstrated in different examples with cop-
per complexes immobilized by electrostatic interactions [29–31].
The aim of this work is to study the possible use of supported het-
eropolyanions, in this case phosphotungstate, as carriers for chiral
copper complexes.
repeated twice. Finally, the solid was dried and calcined as before
washing.
The immobilized heteropolyacid (PW-SR/SR2/SG/A, 1 g) was neu-
tralized by stirring in
a solution of triethylamine (0.05 ml,
0.34 meq) in dichloromethane (5 ml). Then, the solid (PWn-SR/
SR2/SG/A) was separated by filtration, thoroughly washed with
dichloromethane, and dried under vacuum.
2.4. Preparation of the immobilized catalyst
The chiral ligand (0.02 mmol) and Cu(OTf)2 (7.3 mg, 0.02 mmol)
were dissolved in the minimum amount of anhydrous dichloro-
methane under an argon atmosphere. The solution was stirred
for 15 min and filtered through a PTFE microfilter, and the solvent
was removed under an argon flow. The complex was redissolved in
nitroethane (5 ml, pretreated with Na2CO3 and distilled to elimi-
nate all the acid traces), and the required amount of support
(PWn, PWn-SR, PWn-SR2, PWn-SG, PWn-A) was added. The suspen-
sion was stirred under argon for 24 h at rt. The solid was separated
by filtration, thoroughly washed with nitroethane and dichloro-
methane, and dried under vacuum.
2. Experimental
2.1. Preparation of ligands
Azabox ligands (Fig. 1) were prepared by the method described
by Reiser and coworkers [32].
2.2. Neutralization of the unsupported heteropolyacid
H3PW12O40ÁnH2O (PW, Fluka) was calcined at 200 °C for 3 h. Tri-
ethylamine (0.2 ml, 1.2 meq) was added to a suspension of calcined
PW (1 g, 0.347 mmol) in dichloromethane (5 ml), and the mixture
was stirred for 24 h at room temperature (rt). After concentration
under vacuum, the solid was filtered off, washed with dichloro-
methane, and dried to obtain the neutralized heteropolyacid PWn.
2.5. Characterization of supports and catalysts
Elemental analysis was performed in a Perkin-Elmer 2400 ele-
mental analyzer. Copper and tungsten analysis was carried out
by ICP in a Perkin-Elmer Plasma 40 emission spectrometer. FT-IR
spectra of the solids in KBr wafers treated at 200 °C under vacuum
(10À4 torr) were recorded in a Nicolet Avatar 360 FT-IR spectro-
photometer. 31P- and 13C-CP-MAS-NMR spectra were measured
at room temperature on a Bruker Avance III WB400 spectrometer
at 161.99 and 100.63 MHz, respectively. Solid samples were
2.3. Preparation of the immobilized heteropolyacid
PW was immobilized on Ralt-Chemie silica (surface area 253
m2/g, pore volume 1.1 cm3/g, mean pore diameter 11.3 nm) with
a particle size 0.15–0.21 mm (SR) or a particle size 0.42–0.84 mm
(SR2), Grace–Davison silica (SG, surface area 311 m2/g, pore volume
1.7 cm3/g, mean pore diameter 5–10 nm), or Alfa Aesar alumina (A,
surface area 220 m2/g, pore volume 0.62 cm3/g, mean pore diame-
ter 7 nm).
packed in 4-mm ZrO2 rotors and spun at 5 kHz.
tact time were, respectively, 5
s and 3 ms for 31P and 2.45
2 ms for 13C. XPS measurements were taken on an Axis Ultra DLD
(Kratos Tech.) with monochromatized Al K source (1486.6 eV)
p
/2 pulse and con-
l
ls and
a
at 15 kV and 10 mA. Samples were evacuated at room temperature.
In the case of SR, SR2, and A, immobilization was carried out by
the pore-filling impregnation method [33]. In a typical preparation,
the support (1 g) was immersed into a solution of PW (1.1 ml,
0.19 g PW/ml) in EtOH/H2O = 1:1 v/v, at room temperature for
72 h. After this time, the solid was dried at rt for 24 h and calcined
at 200° C for 3 h. The calcined solid was stirred in dichloromethane
(10 ml) for 24 h, separated by filtration, and this process was
repeated twice. Finally, the solid was again calcined at 200 °C for
3 h.
In the case of SG, immobilization was carried out by the equilib-
rium impregnation method [34]. In a typical preparation, silica SG
(1 g) was immersed into a solution of PW (4 ml, 0.12 g PW/ml) in
EtOH/H2O = 1:1 v/v, under constant stirring at room temperature
for 72 h. After this time, the solid was separated from the solution
by centrifugation, dried for 24 h at room temperature, and calcined
at 200 °C for 3 h. The calcined solid was stirred in dichloromethane
(10 ml) for 24 h, separated by filtration, and this process was
2.6. Typical cyclopropanation reaction
Ethyl diazoacetate (5 mmol) was slowly added (2 h) with a syr-
inge pump over a stirred solution or suspension formed by styrene
(5 mmol), n-decane (100 mg, internal standard), and catalyst (see
Tables 1–3 for amounts) in dichloromethane (4 ml) at rt. The
reaction mixture was stirred for 24 h. In some cases, styrene
(4 ml) instead of dichloromethane was used as solvent. The reac-
tion was monitored by GC: FID from Hewlett–Packard 5890II;
cross-linked methyl silicone column: 25 m  0.2 mm  0.33
lm;
helium as carrier gas, 20 p.s.i.; injector temperature: 230 °C; detec-
tor temperature: 250 °C; oven temperature program: 70 °C
(3 min), 15 °C minÀ1 to 200 °C (5 min); retention times: ethyl
diazoacetate 4.28 min, styrene 5.03 min, n-decane (internal
standard) 6.93 min, diethyl fumarate 8.73 min, diethyl maleate
9.04 min, cis-cyclopropanes 11.84 min, trans-cyclopropanes
12.35 min. Enantioselectivities were also determined by gas chro-
matography: FID from Hewlett–Packard 5890II, Cyclodex B col-
umn: 30 m  0.25 mm  0.25
lm; helium as carrier gas, 20 p.s.i.;
N
O
O
injector temperature: 230 °C; detector temperature: 250 °C; oven
temperature program: 125 °C isotherm; retention times: (1S, 2R)-
cyclopropane 28.9 min, (1R, 2S)-cyclopropane 29.8 min, (1R, 2R)-
cyclopropane 34.3 min, (1S, 2S)-cyclopropane 34.9 min.
In the case of heterogeneous catalysts, the solid was recovered
by different procedures:
AzaiPr: R = iPr
AzatBu: R = tBu
N
N
R
R
Fig. 1. Azabis(oxazoline) ligands with isopropyl and tert-butyl substituents.

72
M. Rosario Torviso et al. / Journal of Catalysis 275 (2010) 70–77
Table 1
Analysis of the catalysts.
Entry
Catalyst
Initial Cu amounta (mmol/g)
W (mmol/g)
Cu (mmol/g)
Molar ratio (Cu/W)
1
2
3
4
5
6
7
8
9
10
11
12
13
PWn
–
3.51
3.28
3.22
1.13
0.99
–
0.96
0.79
1.01
1.26
1.21
1.23
1.19
–
–
AzaiPr–Cu–PWn
AzatBu–Cu–PWn
PWn-SR
0.50
0.82
–
0.47
0.50
0.50
0.50
0.20
–
0.23
0.60
–
0.18
0.17
0.25
0.27
0.12
–
0.070
0.186
–
0.182
–
0.260
0.342
0.119
–
0.264
–
AzaiPr–Cu–PWn-SR
AzatBu–Cu–SR
b
AzatBu–Cu0.5–PWn-SR
AzatBu–Cu–PWn-SR2
AzatBu–Cu0.2–PWn-SR
PWn-SG
AzatBu–Cu–PWn-SG
PWn-A
0.50
–
0.50
0.32
–
0.08
AzatBu–Cu–PWn-A
0.067
a
Amount of copper used in the preparation of the immobilized catalysts.
Complex adsorbed on silica without PW.
b
Table 2
Results obtained in the cyclopropanation reaction between styrene and ethyl diazoacetate in dichloromethane catalyzed by Azabox–Cu(II) complexes.a
Entry
Catalyst
% Cat
Run
% Yieldb
Trans/cisb
% ee transb
% ee cisb
AzaiPr–Cu(OTf)2
1
1
1
1
1
1
1
2
1
2
1
1
2
1
2
3
1
2
1
2
3
1
2
3
1
78
39
17
47
82
46
45
30
21
50
57
6
66:34
62:38
73:27
68:32
73:27
71:29
71:29
68:32
66:34
66:34
73:27
62:38
73:27
70:30
70:30
72:28
67:33
74:26
73:27
62:38
72:28
65:35
60:40
74:26
66
54
62
70
92
83
81
88
84
75
82
45
96
88
62
93
77
97
90
40
90
59
10
94
44
43
52
54
84
76
74
81
77
55
83
41
85
72
39
80
65
87
78
46
79
45
11
80
c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
AzaiPr–Cu–Laponited
AzaiPr–Cu–PWn
0.45
0.22
0.22
1
AzaiPr–Cu–PWn-SR
c
AzatBu–Cu(OTf)2
AzatBu–Cu–Laponited
AzatBu–Cu–Nafion/SiO2
0.45
d
0.9
AzatBu–Cu–PWn
AzatBu–Cu–SR
0.7
0.26
AzatBu–Cu0.5–PWn-SR
0.5
55
20
3
AzatBu–Cu–PWn-SR2
AzatBu–Cu–PWn-SG
0.5
38
20
28
30
25
48
28
2
0.66
AzatBu–Cu–PWn-SG
AzatBu–Cu–PWn-A
0.33
0.5
17
a
b
c
Recovery by method A (see Section 2 for details).
Total conversion of ethyl diazoacetate is observed with freshly prepared catalysts. Determined by gas chromatography; 1R cyclopropanes are the major products.
Ref. [43]. Results obtained with methyl diazoacetate, using phenylhydrazine as reductant.
Ref. [37].
d
Table 3
Comparison of the recycling of immobilized Azabox–Cu(II) complexes after cyclopropanation reactions in dichloromethane or styrene as solventa.
Catalyst
% Cat
Run
CH2Cl2
% Yieldb
82c
Styrene
% yieldb
85
trans/cisb
% ee transb
% ee cisb
trans/cisb
% ee transb
% ee cisb
AzatBu–Cu(OTf)2
1
1
73:27c
92c
84c
72:28
93
85
AzatBu–Cu0.5–PWn-SR
0.5
1
2
3
55
20
3
73:27
70:30
70:30
96
88
62
85
72
39
60
60
10
70:30
77:23
66:34
96
90
20
84
81
21
AzatBu–Cu–PWn-SR2
0.5
1
2
3
4
38
20
–
72:28
67:33
–
93
77
–
80
65
–
65
50
48
40
74:26
72:28
72:28
71:29
96
96
85
87
79
82
64
64
–
–
–
–
a
b
c
Recovery by method A (see Section 2 for details).
Total conversion of ethyl diazoacetate is observed with freshly prepared catalysts. Determined by gas chromatography; 1R cyclopropanes are the major products.
Ref. [43]. Results obtained with methyl diazoacetate, using phenylhydrazine as reductant.
– method A: filtration, washing twice with 5 ml of dichlorometh-
ane, drying, and reuse under the same conditions, after drying
under vacuum.
– method B: centrifugation, washing twice with 5 ml of hexane,
and solid reuse under the same conditions after drying under
vacuum (B1) or without drying (B2).

M. Rosario Torviso et al. / Journal of Catalysis 275 (2010) 70–77
73
– method C: solvent evaporation, suspension in hexane (3 ml),
centrifugation, washing twice with 3 ml of hexane, drying,
and reuse under the same conditions.
observed from the IR spectra (Fig. 2), the treatment of PW with
Et3N produces clear changes. First, the characteristic band at
982 cmÀ1, assigned to the asymmetric stretching of the W@Od
3À
bond of the [PW12O40
]
anion, is shifted to 952 cmÀ1 and is
accompanied by a shoulder at 978 cmÀ1. Also, in addition to the
band at 1080 cmÀ1 attributed to the asymmetric stretching of the
P–Oa bond of the tungstophosphate anion, two new bands appear
3. Results and discussion
3.1. Preparation and characterization of catalysts
at 1100 and 1048 cmÀ1, which can be assigned to the splitting of
7À
such band produced when the lacunary [PW11O39
]
species is
Azabis(oxazolines) are structural analogs of bis(oxazolines) and
have shown excellent behavior as chiral ligands in different asym-
metric reactions in homogeneous phase [35,36]. Furthermore, they
lead to stronger complexes, which is advantageous from the point
of view of electrostatic immobilization [37]. Two azabox ligands
with different substituents (AzaiPr, AzatBu) have been used in this
work (Fig. 1).
present [39,40]. So, the observed features are compatible with a
partial transformation of the primary structure of PW into new
species, as the lacunary anion.
On the other hand, the 31P-CP-MAS-NMR spectrum of PW
(Fig. 3) showed the presence of a sharp signal at À15.2 ppm char-
acteristic of H3PW12O40. The neutralized PWn showed in addition
two broad bands at À12.2 and À13.4 ppm, compatible with the
presence of dimeric species, probably formed by aggregation with
Et3N and favored in solid state. The band at À15.2 ppm in the sup-
ported PW is wider than in the unsupported one, in agreement
with the lower mobility (hindered rotation) of the supported
species. Finally, the spectrum of the neutralized supported PW pre-
sents an important band at À10.1 ppm [40], probably due to the
presence of the lacunary species detected by IR. The relative inten-
sity of that broad band with respect to that of PW is highly depen-
dent on the hydration level of the sample, indicating the efficiency
of the crosspolarization by the closer proximity of water molecules,
in agreement with the lacunary character of those species. More-
over, the 13C-CP-MAS-NMR spectrum shows the expected signals
at 48.2 and 9.2 ppm, corresponding to the triethylammonium
cation. The W content (64.5%) is slightly lower than the theoretical
value of the expected anhydrous species (Et3N)3PW12O40 (69.3%),
and it may be due to partial hydration and the formation of the
spectroscopically detected lacunary species.
H3PW12O40 (PW) was used for immobilization due to its higher
acidity in comparison with related heteropolyacids [9], which
should confer to the anion a lower coordinating character. In
addition, it has been the most used heteropolyacid to immobilize
chiral Rh complexes [11,13]. PW was supported on Ralt-Chemie
(SR) silica, as well as Alfa Aesar alumina (A), using the pore-filling
impregnation method followed by drying and calcination, in a typ-
ical process to obtain supported acid catalysts [33]. In the case of
Grace–Davison silica, PW was supported by the equilibrium
impregnation method [34]. In this regard, this methodology is
different from that used to immobilize other chiral complexes
[11–19], because in those cases PW was supported by the diffusive
impregnation method [38], without separation and calcination of
the supported PW. This may be the reason why the silica was
not a suitable support in those cases, as a very low linkage between
PW and the support takes place. Due to the sensitivity of oxazo-
line-containing ligands to acid-catalyzed hydrolysis, supported
and unsupported PW was neutralized (PWn) by stirring in a solu-
tion of triethylamine (Et3N) in dichloromethane prior to exchange
with copper complexes. This step was not necessary in the case of
Rh-diphosphine complexes and their analogs [11–18], which were
immobilized without separation of the supported PW. Our method
is also different from that used by Telalovic and Hanefeld [19],
because in that case supported PW was not neutralized, copper
was first immobilized and then it was complexed with the
bis(oxazoline) ligand.
Cation exchange was carried out in nitroethane in order to pre-
vent leaching of PW in methanol, which is the most commonly
used solvent with other supports [41]. The required amount of sup-
port (PWn, PWn-SR, PWn-SG, PWn-A) was stirred with a solution of
the complex in nitroethane under argon for 24 h at rt. The solid
was separated by filtration, thoroughly washed with nitroethane,
and dried under vacuum.
The catalysts were analyzed (Table 1) to determine the amount
of copper per PW unit and to compare it with the maximum theo-
retical cation exchange capacity (Cu/PW molar ratio = 1.5, or Cu/W
molar ratio = 0.125). In the case of unsupported PW, the use of an
excess of complex during the exchange process leads to a maxi-
The supported and unsupported PWn was characterized by
chemical analysis (Table 1), FT-IR, and CP-MAS-NMR. As can be
Fig. 2. IR spectra of: PW (a), PWn (b), PW-SR (c, dashed line), PWn-SR (d, dashed
line), and Et3N (e, dotted line).
Fig. 3. 31P-CP-MAS-NMR spectra of: PW (a), PWn (b), PW-SR (c), PWn-SR (d).

74
M. Rosario Torviso et al. / Journal of Catalysis 275 (2010) 70–77
-
-
3
(azabox-Cu)2+(OTf )2 + (Et3N+H)3[PW12O40
] -Silica
-
(Et3N+H)(OTf )
-
3
[(azabox-Cu)2+]1.5[PW12O40
]
-Silica
-
[(OTf )(azabox-Cu)+] (solution)
and/or
[(OTf )(azabox-Cu)2+]3[PW12O40
reduction
-
-
3
or
]
-Silica
-
3
[(azabox-Cu)+]3[PW12O40
]
-Silica (solid)
Fig. 4. Mono- and di-exchanged species and the possible effect of Cu(II)–Cu(I) reduction.
mum content of 0.186 Cu/W (entry 3), which seems to be the limit
for this system given that it is not exceeded by the use of an initial
larger copper excess. The existence of an equilibrium is shown by
the lower content, Cu/W molar ratio = 0.07 (entry 2), obtained
when the theoretical stoichiometric amount of complex is used
in the exchange process. In any case, it can be seen that the Cu/
W molar ratio can go beyond the theoretical value necessary to
compensate the negative charge of the intact heteropolyanion.
An excess of complex was used in the exchange process with sup-
ported PW. Values of Cu/W molar ratio ranging from 0.182 to 0.342
can be obtained under such conditions (entries 5, 7, 8, 11), whereas
the theoretical Cu/W molar ratio can be obtained when the initial
amount of complex is close to this value (entry 9). Again in most
cases, the charge corresponding to the amount of immobilized
complex exceeds the theoretical exchange capacity of the solid.
This may be due to the presence of mono-exchanged species
(Fig. 4). The presence of sulfur was detected by XPS in all the sam-
ples with an excess of complex, in agreement with the exchange of
only one of the triflates of the initial complex. On the contrary, sul-
fur was not observable in the samples with Cu/W values below the
theoretical value. However, the mono-exchanged species would
explain Cu/W values up to 0.25 (1 Cu per charge of PW). Higher
values must be due to the presence of lacunary heteropolyanion
species with higher charge, as shown by the spectroscopic
techniques, and/or a contribution of complex adsorption on silica
surface. In fact, SR support is able to adsorb 0.17 mmol of com-
plex/g (entry 6), and it cannot be discarded the presence of a small
amount of complex directly adsorbed on the bare silica or alumina
surfaces in all the solids.
As above-mentioned, the presence of sulfur coming from triflate
was detected by XPS. Values of S/Cu molar ratio ranging between
0.8 and 1.1 were obtained for all the solids exceeding the theoret-
ical Cu/W molar ratio. On the contrary, sulfur was not detected in
AzaiPr–Cu–PWn and AzatBu–Cu–PWn-A, the two catalysts with
Cu/W ratio far below the theoretical amount. These results point
to the existence of mono-exchanged species, as proposed in Fig. 4.
The intact structure of the copper complexes was also con-
firmed by IR spectroscopy (Fig. 5). The presence of all the bands
in the skeletal range of the spectrum, mainly the imine bands at
1652 and 1678 cmÀ1, and the difference from the spectrum of
the free ligand (1636 cmÀ1) are in agreement with the presence
of the ligand forming a copper complex.
Regarding the structure of the supported heteropolyanion, the
spectrum is significantly modified after exchange of the copper
complex (Fig. 6). The intensity of the band at 978 cmÀ1 increases,
leading to different relative intensities with respect to the band
at 952 cmÀ1, which depends on the loading of complex. This result
seems to indicate the coexistence of two types of W@O sites, one
compatible with the lacunary species (952 cmÀ1) and the other
one (978 cmÀ1) compatible with the intact Keggin heteropolyan-
ion. The increase in this signal after complex exchange seems to
indicate that copper occupies the W vacant in the lacunary species
[42].
3.2. Cyclopropanation reactions
The catalysts were tested in the benchmark reaction between
styrene and ethyl diazoacetate (Fig. 7), and the results, compared
Fig. 5. IR spectra of AzatBu–Cu(OTf)2 (a) and two different samples of AzatBu–Cu–
Fig. 6. IR spectra of PW-SR (dashed line), PWn-SR (dotted line), and two different
PWn-SR (b and c) in comparison with the spectrum of free AzatBu (dotted line).
samples of AzatBu–Cu–PWn-SR (solid lines).

M. Rosario Torviso et al. / Journal of Catalysis 275 (2010) 70–77
75
alysts. As the W/Si ratio is kept constant, the leaching of the whole
PW species from the support can be discarded, in agreement with
the low solubility of PW in low polarity solvents, such as the reac-
tion medium [34]. However, the drop in the Cu/W ratio in the cat-
alysts used once, in agreement with a copper leaching in the range
of 15–40%, and the absence of sulfur in them point to a loss of
mono-exchanged species in the Cu(II)–Cu(I) reduction process pre-
vious to the catalytic reaction (Fig. 4). In some experiments, Azat-
Bu–Cu0.5-PWn-SR was treated with ethyl diazoacetate in
dichloromethane to reduce Cu(II), and the solid was filtered. The
activity of the solution gave only 2% yield with trans/cis = 80:20
and lower enantioselectivity (77% e.e. trans), whereas the solid pre-
served more than 95% of the activity observed in a normal reaction,
in spite of the loss of copper, demonstrating the absence of signif-
icant leaching of active species. After the first reuse, the Cu/W ratio
is kept stable and further deactivation is due to poisoning. This ef-
fect has been observed in several cases, and the copper loss is high-
er when the Cu/W ratio of the catalyst is higher, in agreement with
the hypothesis of mono-exchanged species (Fig. 4). On the other
hand, this copper loss is not observed with catalysts having a Cu/
W ratio below the maximum theoretical cation exchange capacity,
in which the presence of sulfur is not detected.
In a second set of experiments, the most promising catalysts
were tested in the same reaction but with a large excess of styrene,
in the absence of any additional solvent. The comparison with the
results in dichloromethane is shown in Table 3. As can be observed,
the results are much better in styrene, mainly regarding reuse.
Now the solid with larger particle size (SR2) showed a better perfor-
mance, with similar yield and selectivities in four consecutive runs.
The recoverability of the catalyst is much better in styrene, proba-
bly due to the higher chemoselectivity of the reaction, mainly in
the case of SR2 support. This recoverability drastically increases
the productivity of the catalyst. In fact using the homogeneous
complex, 85 mol of cyclopropanes per mol of Cu is obtained, with
93% e.e. (trans), whereas the heterogeneous catalyst is able to reach
406 mol of cyclopropanes per mol of Cu, with a cumulative 91.7%
e.e. (trans). An additional detrimental factor for recovery is the
attrition of the catalyst particles, leading to a mass loss during
the filtration, which reduces the effective amount of catalyst in
the consecutive runs, with the consequent negative effect on the
catalytic activity. In view of that, a study of different recovery pro-
cesses was performed.
Trying to prevent the mass loss by filtration, catalyst and solu-
tion were separated by centrifugation, and the catalyst was
washed with hexane instead of dichloromethane (method B). As
can be seen in Table 4, this method is especially suitable for reac-
tions carried out in styrene. Three consecutive runs were tried with
AzaiPr–Cu–PWn-SR, even with very low amount of catalyst,
whereas up to six consecutive runs were tested with AzatBu–
Cu0.5-PWn-SR and higher amount of catalyst. In view of the detri-
mental effect of dichloromethane, it was eliminated by evapora-
tion before the centrifugation in the presence of hexane (method
C). With this method, it is also possible to recover the unsupported
azabox–Cu–PWn catalysts, although with the limitations in either
activity or enantioselectivity described above. This method also al-
lows the recovery of AzaiPr–Cu–PWn-SR without any drop in
enantioselectivity for four runs, although with a drop in activity
after the first run. The detrimental effect of dichloromethane in
the separation process was again demonstrated by the deactivation
in the fifth run when method B2 (direct centrifugation) was ap-
plied instead of method C (evaporation prior centrifugation).
The use of supported heteropolyacids as supports for chiral cat-
alysts requires the combination of a suitable preparation methods
of the chiral heterogeneous catalyst (support, PW impregnation
method, chiral ligand, cation exchange method) with an optimized
reaction-recovery procedure (solvent, separation method, wash-
S
S
R
R
Ph
Ph
COOEt Ph
COOEt
+
N2CHCOOEt
R
S
S
R
Ph
COOEt
COOEt
Ph
Fig. 7. Cyclopropanation reaction between styrene and ethyl diazoacetate.
with those obtained with other homogeneous and immobilized
catalysts, are gathered in Table 2.
In a first set of experiments, the immobilized catalysts were
tested in dichloromethane, recovered by filtration and washing
with the same solvent (method A), and reused under the same con-
ditions. The results are presented in Table 2. All the reactions, with
the freshly prepared catalysts, showed a total conversion of ethyl
diazoacetate to either cyclopropanes or dimerization products
(diethyl fumarate and maleate). Hence, the yield reported is in fact
the chemoselectivity to cyclopropanation reaction. The heteroge-
neous character of the catalysts was tested by adding ethyl diazo-
acetate to the filtrate. In all cases, the yield increase was lower than
2%, showing that the contribution of the homogeneous reaction is
negligible. The only exception was the catalyst prepared by direct
adsorption on silica (AzatBu–Cu–SR, entry 11), in which the contri-
bution of the homogeneous reaction was significant.
A first conclusion is the suitable nature of the PW anion to act as
counterion for azabox–Cu complexes. The results in ‘‘quasi-homo-
geneous” phase show that enantioselectivity similar to that ob-
tained with triflate can be reached with AzaiPr (entry 3) or
AzatBu ligands (entry 10), whereas the final yield increases with
higher catalyst amount, due to a higher chemoselectivity. The
immobilization of PW is highly positive in the case of AzaiPr–Cu
(entry 4), leading to better results than the complex immobilized
on laponite (entry 2) and even better enantioselectivity than that
originally reported in the literature (entry 1). This is also the case
of AzatBu–Cu (entry 13), when compared with the catalysts immo-
bilized on laponite (entry 6) or a nafion–silica support (entry 8).
Several parameters have been studied with AzatBu–Cu. The
morphology of the silica support has some effect on the catalytic
activity, although the selectivities are always at the same level.
Ralt-Chemie silica (entry 13) performs better than Grace–Davison
(entry 18) as support. The larger pore size, and then a better acces-
sibility to the catalytic sites, may account for this behavior. An in-
crease in particle size, with a consequent reduction in the external
surface, is detrimental for the catalytic activity (entry 16 vs. 13),
probably due to intraparticle diffusion limitations. Finally, the per-
formance of alumina (entry 24) is the lowest among the tested
supports, probably due to the strong interaction of PWn with alu-
mina and also because the tungstophosphate anion is transformed
into another species [40].
Recovery and reuse were tested in most of the catalysts. In all
cases, a significant drop in activity and enantioselectivity is ob-
served from the second run, or the third at most, when a larger
amount of catalyst is used (entry 20). As a result of this drop in
activity, conversion of ethyl diazoacetate is not complete, in con-
trast with the behavior of the fresh catalyst. This effect is not due
to leaching of active copper, as mentioned above, but probably
due to deactivation of the complex as a result of adsorption of
by-products obtained from the competitive dimerization and/or
polymerization of ethyl diazoacetate to diethyl fumarate, diethyl
maleate, and poly(ethyl 2-ylidene acetate) [44]. These products
are easily detectable in the used catalysts by IR due to the presence
of the prominent C@O band, and they contribute to the increase in
C content of the used catalysts. This increase makes it difficult to
directly compare the copper content in the used and the fresh cat-

76
M. Rosario Torviso et al. / Journal of Catalysis 275 (2010) 70–77
Table 4
Effect of the recycling method of immobilized Azabox–Cu(II) complexes.a
Catalyst
% Cat
0.22
Reaction solvent
CH2Cl2
Recycling methodb
B1
Run
% Yieldc
trans/cisc
% ee transc
% ee cisc
AzaiPr–Cu–PWn-SR
1
2
47
21
68:32
68:32
70
49
54
29
AzaiPr–Cu–PWn-SR
0.2
0.7
Styrene
Styrene
B2
B2
1
2
3
35
34
32
65:35
66:34
65:35
74
75
75
50
51
52
AzatBu–Cu0.2–PWn-SR
1
2
3
4
5
6
57
55
54
54
53
49
62:38
70:30
62:38
62:38
61:39
62:38
95
89
85
83
82
83
83
80
72
66
66
65
AzaiPr–Cu–PWn
AzaiPr–Cu–PWn
AzaiPr–Cu–PWn-SR
0.22
0.7
CH2Cl2
CH2Cl2
CH2Cl2
C
C
C
1
2
17
12
73:27
65:35
62
62
52
52
1
2
50
40
66:34
66:34
75
72
55
56
0.22
1
2
3
4
5
47
32
32
30
2
68:32
70:30
68:32
70:30
69:31
70
73
72
73
39
54
55
52
53
16
B2
a
b
c
Total conversion of ethyl diazoacetate is observed in all cases.
See Section 2 for details.
Determined by gas chromatography; 1R cyclopropanes are the major products.
[5] J.M. Fraile, J.I. García, C.I. Herrerías, J.A. Mayoral, M. Harmer, J. Catal. 221 (2004)
532.
[6] J.F. Keggin, Proc. Roy. Soc. London, A 144 (1934) 75.
[7] T. Okuhara, Chem. Rev. 102 (2002) 3641.
[8] N. Mizuno, M. Misono, Chem. Rev. 98 (1998) 199.
[9] I.V. Kozhevnikov, Chem. Rev. 98 (1998) 171.
[10] M.N. Timofeeva, Appl. Catal. A 256 (2003) 19.
[11] R. Augustine, S. Tanielyan, S. Anderson, H. Yang, Chem. Commun. (1999)
1257.
ing) to obtain good results in catalytic activity, enantioselectivity,
and recoverability in the cyclopropanation of ethyl diazoacetate
and styrene. These results open the way to the development of
new chiral heterogeneous catalysts of wide applicability in enan-
tioselective synthesis.
4. Conclusions
[12] R.L. Augustine, S.K. Tanielyan, N. Mahata, Y. Gao, A. Zsigmond, H. Yang, Appl.
Catal. A 256 (2003) 69.
3À
[13] R.L. Augustine, P. Goel, N. Mahata, C. Reyes, S.K. Tanielyan, J. Mol. Catal. A 216
(2004) 189.
Tungstophosphate (PW
O
40 ) is a suitable counterion for aza-
12
bis(oxazoline)–copper complexes used as catalysts of cyclopropa-
nation reactions. The performance of this counterion increases
when it is immobilized on supports with large surface area, such
as silica and alumina. The nature and morphology of the support
have an influence on the catalytic results, with silica Ralt-Chemie
being the most efficient. The method of separation and recovery
of the catalyst is also crucial to obtain good recycling behavior.
Centrifugation and exclusion of dichloromethane in the separation
process are key points allowing up to six consecutive runs with sta-
ble activity, reaching productivities higher than 400 mol/mol Cu
and enantioselectivity in the range of 82–96% ee for the trans
cyclopropane.
[14] Á. Zsigmond, I. Balatoni, F. Notheisz, C. Hegedüs, J. Bakos, Catal. Lett. 101
(2005) 195.
[15] Á. Zsigmond, S. Undrala, F. Notheisz, Á. Szöllösy, J. Bakos, Appl. Catal. A 303
(2006) 29.
[16] J.A.M. Brandts, P.H. Berben, Org. Proc. Res. Develop. 7 (2003) 393.
[17] Á. Zsigmond, K. Bogár, F. Notheisz, J. Catal. 213 (2003) 103.
[18] C. Simons, U. Hanefeld, I.W.C.E. Arends, T. Maschmeyer, R.A. Sheldon, J. Catal.
239 (2006) 212.
[19] S. Telalovic, U. Hanefeld, Appl. Catal. A 372 (2010) 217.
[20] G. Desimoni, G. Faita, K.A. Jørgensen, Chem. Rev. 106 (2006) 3561.
[21] J.M. Fraile, J.I. García, J.A. Mayoral, Coord. Chem. Rev. 252 (2008) 624.
[22] J.M. Fraile, J.I. García, C.I. Herrerías, J.A. Mayoral, E. Pires, L. Salvatella, Catal.
Today 140 (2009) 44.
[23] M.J. Fernández, J.M. Fraile, J.I. García, J.A. Mayoral, M.I. Burguete, E. García-
Verdugo, S.V. Luis, M.A. Harmer, Top. Catal. 13 (2000) 303.
[24] H. Wang, X. Liu, H. Xia, P. Liu, J. Gao, P. Ying, J. Xiao, C. Li, Tetrahedron 62 (2006)
1025.
[25] C. Langham, P. Piaggio, D. Bethell, D.F. Lee, P. McMorn, P.C. Bulman-Page, D.J.
Willock, C. Sly, F.E. Hancock, F. King, G.J. Hutchings, Chem. Commun. (1998)
1601.
[26] J.M. Fraile, I. Pérez, J.A. Mayoral, O. Reiser, Adv. Synth. Catal. 348 (2006)
1680.
[27] J.M. Fraile, J.I. García, J.A. Mayoral, M. Roldán, Org. Lett. 9 (2007) 731.
[28] J.M. Fraile, J.I. García, C.I. Herrerías, J.A. Mayoral, E. Pires, Chem. Soc. Rev. 38
(2009) 695.
[29] J.I. García, B. López-Sánchez, J.A. Mayoral, E. Pires, I. Villalba, J. Catal. 258
(2008) 378.
Acknowledgment
This work was made possible thanks to the generous financial
support of the Spanish Ministerio de Ciencia e Innovación (projects
CTQ2008-05138 and Consolider Ingenio 2010 CSD2006-0003), the
Diputación General de Aragón (E11 Group co-financed by the Euro-
pean Regional Development Funds), and the Agencia Española de
Cooperación Internacional (project A/5392/06).
[30] J.M. Fraile, J.I. García, G. Jiménez-Osés, J.A. Mayoral, M. Roldán,
Organometallics 27 (2008) 2246.
[31] M.J. Fabra, J.M. Fraile, C.I. Herrerías, F.J. Lahoz, J.A. Mayoral, I. Pérez, Chem.
Commun. (2008) 5402.
References
[32] H. Werner, R. Vicha, A. Gissibl, O. Reiser, J. Org. Chem. 68 (2003)
10166.
[33] M.R. Torviso, E.N. Alesso, G.Y. Moltrasio, P.G. Vázquez, L.R. Pizzio, C.V. Cáceres,
M.N. Blanco, Appl. Catal. A 301 (2006) 25.
[34] P. Váquez, L. Pizzio, C. Cáeres, M. Blanco, H. Thomas, E. Alesso, L. Finkielsztein,
B. Lantaño, G. Moltrasio, J. Aguirre, J. Mol. Catal. A 161 (2000) 223.
[35] A. Gissibl, M.G. Finn, O. Reiser, Org. Lett. 7 (2005) 2325.
[1] D.E. De Vos, I.F.J. Vankelecom, P.A. Jacobs (Eds.), Chiral Catalysts
Immobilization and Recycling, Wiley-VCH, Weinheim, 2000.
[2] M. Heitbaum, F. Glorius, I. Escher, Angew. Chem. Int. Ed. 45 (2006) 4732.
[3] J.M. Fraile, J.I. García, J.A. Mayoral, Chem. Rev. 109 (2009) 360.
[4] J.M. Fraile, J.I. García, M.J. Gil, V. Martínez-Merino, J.A. Mayoral, L. Salvatella,
Chem. Eur. J. 10 (2004) 758.

M. Rosario Torviso et al. / Journal of Catalysis 275 (2010) 70–77
77
[36] C. Geiger, P. Kreitmeier, O. Reiser, Adv. Synth. Catal. 347 (2005) 249.
[37] J.M. Fraile, J.I. García, C.I. Herrerías, J.A. Mayoral, O. Reiser, A. Socuellamos, H.
Werner, Chem. Eur. J. 10 (2004) 2997.
[40] L.R. Pizzio, C.V. Cáceres, M.N. Blanco, Appl. Catal. A 167 (1998) 283.
[41] J.M. Fraile, J.I. García, J.A. Mayoral, T. Tarnai, Tetrahedron: Asymmetry 9 (1998)
3997.
[38] A.V. Nelmark, L.I. Kheifez, V.B. Fenelonov, Ind. Eng. Chem. Prod. Res. Dev. 20
(1981) 439.
[42] C. Rocchiccioli-Deltcheff, M. Fournier, J. Chem. Soc. Faraday Trans. 87 (1991)
3913.
[39] L. Pizzio, G. Romanelli, P. Vázquez, J. Autino, M. Blanco, C. Cáceres, Appl. Catal.
A 308 (2006) 153.
[43] M. Glos, O. Reiser, Org. Lett. 2 (2000) 2045.
[44] A.F. Noels, Angew. Chem. Int. Ed. 46 (2007) 1208.