Silica "sHB" chiral Pc-L* copper complexes for halogen-free solvent cyclopropanation reactions

Article abstract of DOI:10.1039/c3ra44806a

The grafting of the preformed Pc-L* (pyridine containing macrocyclic ligands) copper(i) complexes on different ordered and non-ordered silicas, and their use, under heterogeneous batch conditions, as catalysts for the olefin cyclopropanation are reported.

Full text of DOI:10.1039/c3ra44806a

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Silica “SHB” chiral Pc-L* copper complexes for halogen-
free solvent cyclopropanation reactions†
Cite this: RSC Adv., 2013, 3, 22199
a
a
ab
c
Brunilde Castano, Paolo Zardi, Yvonne C. H o¨ nemann, Anne Galarneau,
a
b
a
b
Emma Gallo, Rinaldo Psaro, Alessandro Caselli* and Vladimiro Dal Santo*
The grafting of the preformed Pc-L* (pyridine containing macrocyclic ligands) copper(I) complexes on
different ordered and non-ordered silicas, and their use, under heterogeneous batch conditions, as
catalysts for the olefin cyclopropanation are reported. High yields (up to 99%) and good recyclability in
halogen-free solvent reactions were obtained, together with negligible copper leaching (0.1% of total
copper).
Received 24th June 2013
Accepted 18th September 2013
DOI: 10.1039/c3ra44806a
www.rsc.org/advances
stability, which is highly recommended if separation and
recovery of the catalyst are to be pursued.
The advantages of highly active and selective homogeneous
catalysts coupled with the ease of separation typical of hetero-
Introduction
Immobilized heterogenous catalysts present several inherent
advantages, such as stability and recyclability, compared to
their homogeneous counterparts. However, when chiral catalyst geneous systems is also highly recommended to implement
are immobilized onto solid surfaces, the interaction between such catalysis. Moreover, by using nanostructured supports,
the active metal complex and the surface may reduce the like mesoporous silicas or clays, it is also possible to exploit the
1
13–16
obtained stereoselectivities. Anchoring chiral complexes positive connement effects
on activity, diastereo-, and
through covalent bonds has been one of the most used methods enantio-selectivity.
to immobilize chiral ligands on silica materials, however the
structural modication needed to gra the ligand to the
support has oen a detrimental effect on the observed enan-
Here we report on the use of SHB copper catalysts in cyclo-
propanations by using bare alkanes as reaction solvents in place
of the less desirable halogeno–alkanes.
2
tioselectivities. Conversely, supported hydrogen-bonded (SHB)
catalysts (a class of non-covalent bound homogeneous catalysts)
show some peculiar advantages, with respect to the covalent
bound ones, such as remarkably mild graing protocols and the
possibility to easily recover the bound complex for further
studies by standard liquid-phase techniques. Moreover, the
SHB methodology can be protably applied to cationic metal
Results and discussion
Synthesis and characterisation of the supported catalyst
The copper(I) complex 1 (Fig. 1) was chosen as a model complex
to be supported on different silicas to be studied as a catalyst in
cyclopropanation reactions. The presence of a naphthyl
substituent confers a good stability to this copper(I) complex
and, although the synthesis is better performed under protect-
ing atmosphere, compound 1 could be manipulated in air for
limited period of time without any appreciable decomposition.
The reason of this major stability is due to the presence of the
naphthyl group that can act as a further coordination site for
the metal (Fig. 1). Four different silicas were investigated as a
support: two commercial ones, Davisil LC150 (Grace) and Aer-
osil 380 (Evonik), and two popular and oen applied meso-
3,4
complexes, with minimal or no ligand modications of the
ꢀ
parent homogeneous complex (only the presence of CF
counter-anion is necessary).
3
SO
3
5
Copper complexes SHB catalysts found interesting applica-
6
tions in several reactions, like cyclopropanations, Diels–Alder
7,8
9
additions, and epoxidations.
On the other hand, Cu(I) complexes based on functionalised
1
0
pyridine-containing macrocyclic ligands
show excellent
11,12
performances in asymmetric cyclopropanation reactions.
The macrocyclic ligands confer to the complex additional
17,18
19
porous silica support materials, MCM-41
ESI† for details). The adopted SHB methodology allowed the
easy and very mild preparation of the supported 1/SiO catalyst
and SBA-15 (see
a
Dipartimento di Chimica, Universit a` di Milano, Via C. Golgi 19, 20133 Italy. E-mail:
2
alessandro.caselli@unimi.it; Fax: +39-02-50314405; Tel: +39-02-50314372
showing a copper loading in a range of 0.3–1.8 Cu wt%,
depending on the used silica (see Table 1 in ESI†). The inter-
action, via H-bonding, was conrmed by a perusal of DRIFT
spectra: upon the Cu complex graing the strong and sharp
b
CNR – Istituto di Scienze e Tecnologie Molecolari, Via C. Golgi 19, 20133 Italy. E-mail:
v.dalsanto@istm.cnr.it; Fax: +39-02-50314405; Tel: +39-02-50314428
c
0
CNRS ICGM ENSCM, 8 Rue de l Ecole Normale, 34296 Montpellier, France
Electronic supplementary information (ESI) available: Materials preparation,
†
ꢀ
1
DRIFTS, NMR, characterization and catalytic tests. See DOI: 10.1039/c3ra44806a
band located at 3740 cm , ascribable to isolated silanols,
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 22199–22205 | 22199

View Article Online
RSC Advances
Paper
Table 1 Cu(I)/Pc-L* (S configuration) supported complexes for asymmetric
a
cyclopropanation of a-methyl styrene
d
ee (%)
b
c
c
Entry Run Catalyst
Yield (%) cis : trans
cis
trans
1
2
3
1
1
1
1
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
6
7
1
1
1
2
3
1
2
3
1
2
3
1
2
3
1
81
72
<1
74
65
68
56
45
43
40
91
78
52
57
73
58
53
77
75
97
88
99
67
71
78
65
37
62 : 38
72 : 28
—
55
35
—
35
40
39
41
35
42
38
35
39
40
36
38
46
29
34
23
39
37
33
36
37
37
37
34
46
58
58
53
62
26
—
27
31
32
30
25
33
29
24
30
30
29
29
38
22
23
20
34
31
33
26
27
26
26
25
58
54
54
47
1/Davisil_1
1/Davisil_1
e
76 : 24
72 : 28
72 : 28
71 : 29
71 : 29
72 : 28
70 : 30
67 : 33
72 : 28
71 : 29
70 : 30
73 : 27
72 : 28
68 : 32
71 : 29
72 : 28
63 : 37
62 : 38
63 : 37
69 : 31
70 : 30
71 : 29
70 : 30
70 : 30
59 : 41
64 : 36
64 : 36
63 : 37
63 : 37
67 : 33
69 : 31
68 : 32
63 : 37
63 : 37
62 : 38
68 : 32
65 : 35
63 : 37
4
5
6
7
8
9
1
1/Davisil_1
Fig. 1 Supported copper(I) complex 1.
1/SBA-15_A_2
1/SBA-15_B_2
1/SBA-15_B_1
1/MCM-41_B_1
1/MCM-41_B_1
1/MCM-41_A_1
disappeared almost completely, and a new broad band, located
ꢀ1
between 3500 and 3400 cm (originated by O–H stretching
vibration of silanols H-bonded with triate counteranion)
appeared in the spectra (Fig. 2). Moreover, DRIFT spectra
showed also that the Cu complex is graed without any modi-

cation of the ligand structure, since the IR absorption bands
did not show any appreciable modication with respect to solid
, nor in location, nor in intensity (see ESI, Fig. S1†).
f
1
Spectroscopic properties
0
The nature of the graed Cu complex was further investigated
by CO–DRIFT spectroscopy, since CO is a very good probe
molecule for Cu(I) sites. In fact, stretching vibrations of a copper
bound CO, n(C^O), is very sensitive to the coordination and
g
e, f
1
1
1/MCM-41_A_1 <10
20,21
electronic density on Cu site.
64
f
Cu complex in the solid state (diluted in KBr, trace E in 12
1/MCM-41_A_1
1/Aerosil_2
1/Aerosil_2
3/Davisil_1
69
99
96
73
99
91
99
95
98
63
71
65
Fig. 3) did not adsorb CO under mild pressure (up to 2 bar),
while, when CuL was dissolved in dichloroethane, an intense
n.d. n.d.
1
3
39
39
39
33
50
46
31
30
30
33
51
48
ꢀ
1
band located at 2111 cm
(
readily appeared in solution
1
1
Scheme 1, trace D in Fig. 3). The observed frequency of the
f
coordinated CO was slightly higher than those reported for the 14
CO adsorbed on tetra coordinated Cu(I) of [bis(2-pyridylmethyl)
amine] copper complexes (2097 cm
lower efficiency of aliphatic amines in transferring electron
density to the metal centre compared to pyridine.
Notably also the graed complex exposed to CO ow showed
the presence of a strong CO band, slightly red-shied and
ꢀ
1
22
)
probably due to the
h
1
5
ꢀ59 ꢀ60
ꢀ65 ꢀ56
ꢀ67 ꢀ60
a
ꢀ2
Reactions were performed with [Cu(I)] (3.0 ꢁ 10 mmol) in n-hexane
(5 mL). Slow addition of EDA (1 mmol) in n-hexane (1 mL) over 100 min
ꢀ
1
ꢂ
b
broadened, located in the range 2115–2119 cm , (traces A–C in at 0 C. Support materials: letter A or B refer to different materials,
number 1 or 2 refers to different graing procedures, see Table S1.†
Fig. 3) that suggests a “solution like” complete accessibility to the
Cu sites of the supported complex to small molecules (like CO).
c
1
Determined by GC and H NMR, with 2,4-dinitrotoluene as internal
d
standard. Determined by chiral HPLC; absolute congurations: cis-
cyclopropanes were (1R,2S), trans-cyclopropanes were (1R,2R); The
opposite enantiomers were obtained in the same ee when employing
Pc-L* with a R conguration. Sheldon test (upper row). Reaction
performed in 1,2-dichloroethane. In the 7 run reaction was slower
and further 45 minutes aer the addition were needed to reach
completion. Opposite enantiomers were obtained.
Catalytic cyclopropanation of olens
e
f
g
th
The catalytic activity of copper complexes supported on silica in
cyclopropanation reactions has been investigated. As model
reaction we chose the cyclopropanation of a-methyl styrene by
EDA (EDA ¼ ethyl diazoacetate). Catalytic reactions were run by
h
2
2200 | RSC Adv., 2013, 3, 22199–22205
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
volatile organic solvent, its use should be intended only as a
proof of concept. In fact, for practical applications it may be
easily substituted with other hydrocarbons with higher boiling
ꢂ
point. All reactions were carried out at 0 C, since this was the
optimal temperature for cyclopropanation reactions catalysed
by complex 1 in the homogeneous phase. We rst tested the
reaction by changing the employed silica support. The results
are summarised in Table 1, compared with a typical result for
the cyclopropanation reaction in the homogeneous phase by
using the same copper(I) complex 1 as catalyst. All reported
yields have been determined by GC and conrmed by quanti-
1
tative H NMR and based on EDA. In all tested cases we
observed a complete conversion of the starting EDA aer the
end of the addition (100 min) and cyclopropanes were obtained
as major products; fumarate and maleate, the homo-coupling
products of EDA, were the only other compounds detected and
accounted for the reaction mass balance.
Fig. 2 DRIFT spectra showing the interaction via H-bonding; (A) bare Davisil; (B)
1
1
/Davisil_1; (C) bare MCM-41_A; (D) 1/MCM-41_A_1; (E) bare SBA-15_B; (F)
/SBA-15_B_2.
The catalyst activity was almost unchanged in the heteroge-
neous phase with respect to the homogeneous one, although in
some cases (entry 5, Table 1) the yield in cyclopropanation
products is somewhat lower, due both to a slight increase in the
coupling product fumarate and maleate and to a partial
absorption of the cyclopropanes on silica (see next section).
Actually, in selected cases, especially when employing n-hexane
as the solvent, reaction yields in cyclopropane products increase
in the second run. This can be explained with a better release of
the formed product from the silica (see entries 6, 9, 10 and 13,
Table 1). Experimental data show only a very weak dependence
of the reaction efficiency from the surface area characteristics of
the employed silica. In particular, we have obtained almost
indistinguishable diastereo- and enantio-selectivities using
either commercial Davisil or Aerosil 380, either ordered meso-
porous silicas MCM-41 or SBA-15. The possibility to use
commercially available silica as a support in the present system
is very interesting and pave the way to the employment of this
immobilization technique also in laboratories not equipped for
the synthesis of mesoporous materials. Other authors, using
covalently bonded bis(oxazoline)–copper complexes on silicas,
recently reported that in those conditions commercially avail-
able silica were not suitable materials, probably due to its
Fig. 3 CO–DRIFT spectra of Cu(I) Pc-L* supported complex. (A) 1/Davisil_1; (B)
/MCM-41_A_1; (C) 1/SBA-15_B_2; (D) 1 in dichloroethane solution; (E) 1 solid
1
state (KBr pellet).
2
acidity.
23
The Sheldon test showed that the catalyst is strongly bound
to the support and that the ltered solution has no catalytic
activity, while the ltrate keeps the same catalytic activity as the
original material (compare entries 2 and 3, Table 1). Aer the
catalytic run, the ltrate solution was analyzed by ICP-OES to
determine the copper content and the results (0.1% of total Cu
originally present in the catalyst) conrmed that leaching is
Scheme 1 Reaction of complex 1 with CO.
syringe pump slow addition of a EDA solution to a suspension negligible when employing n-hexane as the reaction solvent. A
containing the olen and the catalyst (Cu–EDA–olen ratio blank test (MCM-41 not loaded with copper complex) showed
1
: 35 : 170); the disappearance of the band due to the stretch- that the silica has no catalytic activity in the present reaction.
ꢀ1
ing of the N
2
moiety (n ¼ 2114 cm ) was followed by IR spec-
As oen observed for heterogeneous catalysts, the ees are in
troscopy. These conditions are very close to those used in the any case slightly lower if compared to the homogeneous system,
12
2
homogeneous phase and allow for a direct comparison of the especially for the trans isomer. Compared to the homogeneous
obtained results. The major difference is that the solvent used reaction (entry 1, Table 1), better diastereomeric excesses in
in the homogeneous phase, 1,2-dichloroethane, was replaced favor of the cis isomer were obtained in n-hexane, a fact that is
with the more eco-friendly n-hexane. Although the latter is a in agreement with those reported for clay-immobilized copper
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 22199–22205 | 22201

View Article Online
RSC Advances
Paper
24
complexes. In a low polarity solvent, in fact, tight ion pairs are runs) we observed a decrease in the enantioselectivities (ee cis
to be expected and it is reasonable to assume that in this case, 19%, trans 20%) and a complete drop of the diastereoselectivity
the steric hindrance of the support must favor the preferential (cis : trans ¼ 50 : 50). Similar negative effects related to the
15
formation of the cis isomer. It should be pointed out that diazoacetate steric hindrance have been observed also in the
12
complex 1 is not soluble in n-hexane and thus the reaction homogeneously catalysed reaction and have been explained in
28
cannot be performed in this solvent under homogeneous terms of overcrowding of the transition state.
conditions. On the other hand, reactions in 1,2-dichloroethane Under optimal conditions, other alkenes were employed to
yielded to better enantioselectivities, especially for the trans determine the substrate scope of the copper catalysed cyclo-
isomer (entries 9, 11, 12 and 14, Table 1). This may well be due propanation reactions. Since the results in terms of diastereo-
to the enhancement of the polarity of the medium that will and enantioselectivity were comparable for all tested materials,
25
affect the strength of ion pairs and ions–silica interactions,
nally leading to the observed differences in enantioselectivity support. At a Cu(I)–EDA–alkene ratio of 1 : 35 : 170, the formed
it should be pointed out how cyclopropanation reactions can complex catalysed the reaction of all the tested substrates
the less expensive commercial Davisil silica was chosen as

(
14
be affected by several parameters in a complex way).
yielding the cyclopropanes in acceptable yields and enantiose-
The improvement in the yield of the rst run in these last lectivities. It should be pointed out that for these substrates,
1
cases can be explained with the better solubility of the cyclo- reported yields have been determined by quantitative H NMR
propanes in chlorinated solvents. However, in this last solvent, and based on EDA. However, the high volatility of some cyclo-
some leached catalyst could beactive in solution, asshown by the propane products requires a careful removal of the reaction
Sheldon test performed in 1,2-dichloroethane (entry 11, Table 1). solvent. Fumarate and maleate were again the only detected side
This may explain the higher enantiomeric excesses obtained in products. The absence of a-substituents on the styrene affected
the second and third run of the reaction when employing Aerosil the diastereoselectivity of the reaction, although also in this case
3
80 as a support (entry 14, Table 1), while ee obtained in n-hexane a slight cis preference was observed compared to the results
12
remains constant during all the consecutive runs (entry 13, Table obtained in the homogeneous phase. We have obtained almost
). To further asses eventual leaching of the catalyst employing indistinguishable diastereo- and enantioselectivities using
1
dichloroethane as reaction solvent, also in this case the ltered styrene (entry 1, Table 2), 4-methyl styrene (entry 2) and 4-chloro
reaction mixture aer catalysis was analyzed by ICP-OES, styrene (entry 3). Enantioselectivities obtained for the cis cyclo-
showing Cu concentrations below 12.4 ppm, corresponding to a propane products with those substrates were higher than those
12
maximum leaching of 4.3% of copper present in the catalysts. It observed in the homogeneous reactions. Low yields and
should be pointed out that in such low concentration in solution, enantioselectivities were obtained with diphenylethylene (entry
ꢂ
complex 1 is a poor cyclopropanation catalyst, especially at 0 C. 4, Table 2). Interestingly, benzophenone (less than 10% with
Copper leaching in chlorinated solvents should be promoted by respect to starting diphenylethylene) was found in this case
2,26
the presence of EDA.
Given the efficacy of the supported catalysts in the cyclo-
amongst the reaction products.
To explore the scope of the reaction, we next studied the
propanation reaction, we tested their reusability both using cyclopropanation of two different dienes. Excellent diaster-
n-hexane and 1,2-dichloroethane as single solvents. As eoselectivity to yield the desired trans isomer (attack only at the
mentioned previously, n-hexane appears as a valuable greener non-substituted double bond), was obtained with methyl-2-
29
solvent with respect to chlorinated ones. It is noticeable that furoate (entry 5, Table 2). The cyclopropane obtained from this
quantitative conversions were still observed and cyclopropanes reaction is an important building block in the synthesis of
30
were obtained in almost unchanged yields, diastereo- and bioactive compounds.
enantio-selectivities in all consecutive runs. All catalysts were With 2,5-dimethyl-2,4-hexadiene, an important precursor to
31
recycled at least three times, without any noticeable deactiva- the chrysanthemic acid synthesis, the catalytic reaction yiel-
tion. In particular, in the case of catalyst 1/MCM-41_A_1 the ded the desired cyclopropanes (cyclopropanation of only one
32
reaction was repeated for several consecutive runs and only double bond was observed ) although in modest yields (37%,
aer the seventh recycle (two days) a decrease in activity was isolated yield), if a large excess of the olen (Cu–EDA–olen ¼
observed. It is well known that silica can act as ligand for copper 1 : 35 : 500) is employed (entry 6, Table 2). Worth to note,
27
leading to non-enantioselective catalytic processes. We guess slightly higher ees with respect to the homogeneously catalysed
that this must not be the case in the present system, since if that reaction were observed. Even aliphatic alkenes that are gener-
33
were truth we should expect a progressive decrease in the ally less reactive in cyclopropanation reactions, gave very good
enantioselectivity aer consecutive runs. To further exclude this results. For instance, the cyclopropanation occurs also with
hypothesis, in order to remove eventual free copper species reasonable yields with the non-activated double bond of
from the reaction mixture, the catalyst was washed with 1,2- n-octene, although in this case the olen has been used as the
dichloroethane aer the rst run in n-hexane. Then, the reac- solvent (entry 7, Table 2).
tion was repeated using the hydrocarbon solvent and no change
in yield, diastereo- and enantioselectivity was observed.
Finally, since the Cu(I) complex, 3 (Fig. 4), of the more
sterically hindered ligand, possessing two further stereocenters
When employing the bulkier tert-butyl diazoacetate, tBuDA, on the ring skeleton (6-[(R)-1-naphthylethyl]-3,9-ditosyl-3,8-
conditions as in the caption to Table 1, 1/Davisil_1 as sup- [(S,S)-iso-propyl]-3,6,9,15-tetraazabicyclo[9,3,1]pentadeca-1(15),
(
ported catalyst) together with lower yields (35% average of three 11,13-triene), gave the best results in term of enantioselectivity
2
2202 | RSC Adv., 2013, 3, 22199–22205
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
a
Table 2 Asymmetric cyclopropanation of alkenes by 1/Davisil_1
c
ee (%)
b
b
Entry
1
Alkene
Yield (%)
cis : trans
cis
trans
36
61 : 39
61 : 39
61 : 39
58 : 42
60 : 40
57 : 43
45
37
38
42
39
39
35
29
30
35
35
35
33
3
5
6
6
5
9
5
8
2
3
Fig. 4 Copper(I) complex 3.
4
6
5
8
1
0
56 : 44
56 : 44
56 : 44
42
42
42
38
38
38
general, CO bands undergo a small redshi to lower wave-
ꢀ
1
33
—
21
numbers at around 2113 cm (see ESI, Fig. S3†), possibly due
4
5
36
—
—
23
25
n.d.
n.d.
n.d.
38
to the reversible coordination of reaction products on copper
2
4
9
6
34
site. Coordination of by-products (diethyl maleate and fuma-
>1 : 99
>1 : 99
>1 : 99
60 : 40
60 : 40
49
49
44
33
33
35
rate) and formation of diazoacetate polymers have been
43
proposed in the literature as possible deactivation processes for
bis(oxazoline) copper complexes. Also in the present system
slightly longer reaction times, aer the third run, were
observed. However, high yields and constant stereoselectivities
were obtained in all consecutive runs.
4
3
3
4
7
7
38
d
6
e
7
n-octene
50
56 : 44
40
35
a
ꢀ2
Reactions were performed with [Cu(I)] (3.0 ꢁ 10 mmol) in n-hexane
5 mL). Slow addition of EDA (1 mmol) in n-hexane (1 mL) over 100 min
Conclusions
(
ꢂ
b
1
at 0 C. Determined by H NMR, with 2,4-dinitrotoluene as internal
c
standard. Determined by chiral HPLC; absolute congurations: for In summary, we have developed new supported hydrogen-
entries 1–3 cis-cyclopropanes were (1R,2S), trans-cyclopropanes were
bonded (SHB) chiral copper(I) complexes showing good
performances in cyclopropanation reactions allowing the use of
(
1R,2R); entry 4: the absolute conguration was (1R), entry 5: trans-
cyclopropane was (1S, 2S, 6S). For entries and were not
1 : 35 : 500; isolated yield. more environmentally friendly n-hexane as solvent in place of
n-Octene as the solvent: Cu(I)–EDA–olen ¼ 1 : 35 : 1060; isolated yield.
6
7
d
determined.
Cu(I)–EDA–olen
¼
e
1,2-dichloroetane. Although a vefold excess of the olen has to
be used to keep low the side products derived from EDA self-
condensation, it can be easily recovered at the end of the reac-
in the homogeneous phase, we next studied its reactivity when tion by simple distillation and re-used for further reactions. The
supported on silica. Again commercial Davisil LC150 was heterogeneous systems showed higher or comparable activities
11
chosen as support.
than the homogeneous counterpart and a good recyclability. In
A complete conversion of the starting EDA was observed all cases, the diastereoselective outcome of the reaction is
also in this case (entry 15, Table 1), although we should point strongly affected by the catalyst graing on the silica support
out that the reaction in the third run was slightly slower and and we observed an inversion of the selectivity in favour of the
further 15 minutes aer the addition were needed to reach less sterically hindered cis cyclopropane derivative. This is an
completion. Remarkably, 1/Davisil_1 and 3/Davisil_1 gave interesting result, since cis cyclopropanes are present in several
cyclopropane products in very similar yields (compare entry 4 compounds biologically active. The catalyst can be easily recy-
with entry 15, Table 1). It should be pointed out that complex 3 cled (up to 6 times) without any change in the obtained ster-
under homogeneous conditions gave equal amounts of both eoselectivity. We did not observe any signicant Cu leaching
isomers whilst, when supported on Davisil, a slight preference when employing n-hexane as a reaction medium and the cata-
for the cis isomer is again obtained. As expected, observed lytic system is of truly heterogeneous nature, since the ltered
enantiomeric excesses were higher (67% for the trans, 60% for solution is not catalytically active. The observed effects on
the cis isomer).
stereoselectivity are more dependent on the employed solvent
Graed complexes were investigated at the end of the (non-polar vs. halogenated) than to the kind of support (ordered
catalysis. The intact Cu complex structure was conrmed by or non-ordered). Worth to note is the fact that even commercial
means of IR spectroscopy, showing the presence of all the bands silica can be used as a support, without any need of structural
in the skeletal range of the spectrum typical of 1. These results modication of the ligand in order to strongly gra the
conrmed that the graed complex is stable under the reaction complex.
ꢀ
1
conditions. The presence of bands at 2986 and 1746 cm
Cyclopropanes were obtained in good to excellent yields and
suggests the presence of adsorbed –COOR compounds (IR enantiomeric excesses up to 67%. The high stability showed by
spectra of reaction products pure cyclopropanes show similar the copper(I) Pc-L* complexes will be the basis for more work in
ꢀ1
bands located at 2980 and 1730 cm ) (see ESI, Fig. S2†). The the direction of their use as supported catalysts and we are
state of Cu(I) site was further investigated by CO–DRIFT and, in currently moving towards the use of these catalysts in reactors.
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 22199–22205 | 22203

View Article Online
RSC Advances
Paper
Activation of all silicas was performed in a Schlenk ask at
Experimental procedures
Materials
ꢂ
3
1
00 C for 2–3 h in air, subsequently in high vacuum (at least
ꢀ5
0
mbar) overnight.
Solvents were dried prior use by standard procedures and stored
2
under dinitrogen. a-Methyl styrene was distilled over CaH and
I
Graing of [Cu (Pc-L*)]CF SO complex 1, on silica. Typical
3
3
˚
stored under dinitrogen. Davisil (Grace Davison, LC 150 A, 35–70
micron) and Aerosil 380 (Evonik) are commercially available.
All other starting materials were commercial products and
were used as received.
procedure
Method 1. Complex 1 (0.0461 g, 0.0629 mmol) was dissolved
in CH Cl (10 mL). The resulting colourless solution was added
to activated Davisil LC150 (0.400 g), the mixture was stirred at
RT for 4 h under inert atmosphere, ltered, washed with CH Cl
2
2
2
2
Instrumentation and measurements
(
3 ꢁ 5 mL) and dryed overnight to yield the immobilized cop-
NMR spectra were recorded on Bruker Avance 300-DRX or
per(I) complex.
Avance 400-DRX spectrometers. Chemical shis (ppm) are
2 6 6
Method 2. [Cu(OTf)] $(C H ) (0.140 g, 0.277 mmol) was
added to a C H Cl (28 mL) solution of Pc-L* ligand (0.371 g,
.555 mmol). The resulting colorless solution was stirred for
h, than 5.5 mL of solution was added to activated SBA-15_A
0.340 g), the mixture was stirred at RT for 4 h under inert
1
reported relative to TMS. The H NMR signals of the
2
4
2
compounds described in the following have been attributed by
COSY and NOESY techniques. Assignments of the resonance in
0
1
(
1
3
C NMR were made using the APT pulse sequence and HSQC
and HMBC techniques. Infrared spectra were recorded on a
BIO-RAD FTS-7 spectrophotometer. Elemental analyses and
mass spectra were recorded in the analytical laboratories of
Milan University. GC-MS analysis were performed on a Shi-
madzu GCMS-QP5050A instrument. Optical rotation were
atmosphere, ltered, washed with C Cl
dryed overnight to yield the immobilized copper(I) complex.
2
H
4
2
(3 ꢁ 5 mL) and
General procedure for the catalytic cyclopropanation
reactions
measured on a Perkin Elmer instruments model 343 plus; [a]
D
ꢀ
1
2
ꢀ1
I
values are given in 10
deg cm
g
. The water and air
In a typical experiment, [Cu (Pc-L*)]$(CF
ꢁ
3
SO
3
)/SiO
2
([Cu] ¼ 3.0
ꢀ
2
sensitive compounds were handled in a dry-box, model “MB-
0-Compact”. Metal loadings are determined by ICP-OES using
10 mmol) and a-methyl styrene (0.650 mL, 5.0 mmol) were
suspended in n-hexane (5 mL) and the reaction mixture was
cooled at 0 C. Then a n-hexane solution (1 mL) of EDA (0.114 g,
1
a Thermo X Series II apparatus. 15 mg of each sample are
mineralized by adding 3 mL of 37% HCl, 1 mL of concentrated
ꢂ
0
.105 mL, 1 mmol) was slowly added by a syringe pump during
100 minutes. The reaction was monitored by IR, following the
disappearance of the band due to the stretching of N moiety at
114 cm . The reaction was considered to be nished when the
HNO , 1 mL of 98% H SO . CO-DRIFT spectra of the samples
3
2
4
were recorded using a FTS-60A spectrophotometer consisting
of a homemade reaction chamber. Aer purging the apparatus
with ultra-pure He, spectra of the samples were recorded at RT
in He and CO ow, before and aer catalysis. HPLC analyses
2
ꢀ1
2
absorbance of the EDA was below 0.03 (by using a 0.1 mm thick
cell). The solution was ltered by cannula, the solid catalyst was
washed with n-hexane (3 ꢁ 5 mL), then 2,4-dinitrotoluene was
added as internal standard and the solution was analysed by
GC. The solution was then evaporated to dryness in vacuo and
were performed on
a Hewlett-Packard 1050 instrument
equipped with DAI-CEL CHIRALCEL, IB, OJ and AD chiral
columns.
1
analysed by H NMR. The residue puried by silica gel chro-
Preparations
matography (eluant ethyl acetate–n-hexane ¼ 0.3 : 10). Enan-
tiomeric excess were determined by HPLC.
Unless otherwise specied, all the reactions were carried out in
a argon or dinitrogen atmosphere employing standard Schlenk
techniques and magnetic stirring.
The solid catalyst could be recycled for further reactions
employing the same reagent amounts. In all cases (with few
exceptions reported in the tables captions) a quantitative
conversion of the starting EDA was observed.
The copper(I) complex 1 and 2 were synthesized as already
11
reported.
MCM-41 (ref. 36) and SBA-15 (ref. 37 and 38) were synthe-
sized as already reported. The characteristic (pore diameter,
pore volume, surface area) are listed below:
Acknowledgements
Financial support of the European Community's 7th Frame-
work Programme through the Marie Curie Initial Training
Network NANO-HOST, (Grant Agreement no. 215193), and of
MIUR (project ItalNanoNet) is gratefully acknowledged.
MCM-41_A 6124: pore diameter 3.6 nm; pore volume
ꢀ
1
2
ꢀ1
0
0
.61 mL g ; surface area 827 m g
.
MCM-41_B 6170: pore diameter 3.6 nm; pore volume
ꢀ
1
2
ꢀ1
.73 mL g ; surface area 967 m g
.
ꢂ
SBA-15_A (prepared at 60 C): pore diameter 6.7 nm; pore
volume 0.69 mL g ; surface area 786 m g
SBA-15_B (prepared at 130 C): pore diameter 9.6 nm; pore
volume 1.02 mL g ; surface area 525 m g
Before use, MCM.41 and SBA-15 were calcinated at 550 C for
h in air.
ꢀ1
2
ꢀ1
.
Notes and references
ꢂ
ꢀ1
2
ꢀ1
.
1 D. Rechavi and M. Lemaire, Chem. Rev., 2002, 102, 3467–3494.
2 F. Fakhfakh, L. Baraket, A. Ghorbel, J. M. Fraile, C. I. Herrer ´ı as
and J. A. Mayoral, J. Mol. Catal. A: Chem., 2010, 329, 21–26.
ꢂ
8
2
2204 | RSC Adv., 2013, 3, 22199–22205
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
RSC Advances
3
4
C. Bianchini, D. G. Burnaby, J. Evans, P. Frediani, A. Meli, 19 D. Zhao, J. Feng, Q. Huo, N. Melosh, G. H. Frederickson,
W. Oberhauser, R. Psaro, L. Sordelli and F. Vizza, J. Am.
Chem. Soc., 1999, 121, 5961–5971.
B. F. Chmelka and G. D. Stucky, Science, 1998, 279, 548–552.
20 M. Kujime, T. Kurahashi, M. Tomura and H. Fujii, Inorg.
Chem., 2007, 46, 541–551.
V. Dal Santo, M. Guidotti, R. Psaro, L. Marchese, F. Carniato
and C. Bisio, Proc. R. Soc. London, Ser. A, 2012, 468, 1904– 21 S. H. Strauss, Dalton, 2000, 1–6.
1
926.
22 H. R. Lucas, G. J. Meyer and K. D. Karlin, J. Am. Chem. Soc.,
2010, 132, 12927–12940.
5
6
7
8
C. Bianchini, V. Dal Santo, A. Meli, W. Oberhauser, R. Psaro
and F. Vizza, Organometallics, 2000, 19, 2433–2444.
M. M. Diaz-Requejo, T. R. Belderrain, M. C. Nicasio and
P. J. Perez, Organometallics, 2000, 19, 285–289.
H. Wang, X. Liu, H. Xia, P. Liu, J. Gao, P. Ying, J. Xiao and
C. Li, Tetrahedron, 2006, 62, 1025–1032.
23 R. A. Sheldon, M. Wallau, I. W. C. E. Arends and
U. Schuchardt, Acc. Chem. Res., 1998, 31, 485–493.
24 A. I. Fern ´a ndez, J. M. Fraile, J. I. Garc ´ı a, C. I. Herrer ´ı as,
J. A. Mayoral and L. Salvatella, Catal. Commun., 2001, 2, 165–170.
25 J. M. Fraile, J. I. Garc ´ı a, J. A. Mayoral and T. Tarnai, J. Mol.
Catal. A: Chem., 1999, 144, 85–89.
P. O'Leary, N. P. Krosveld, K. P. De Jong, G. Van Koten and
R. J. M. Klein Gebbink, Tetrahedron Lett., 2004, 45, 3177– 26 T. M. Rosario, M. N. Blanco, C. V. Caceres, J. M. Fraile and
180. J. A. Mayoral, J. Catal., 2010, 275, 70–77.
M. M. Diaz-Requejo, T. R. Belderrain and P. J. Perez, Chem. 27 J. K. Park, S. W. Kim, T. Hyeon and B. M. Kim, Tetrahedron
Commun., 2000, 1853–1854. Asymmetry, 2001, 12, 2931–2935.
0 B. Castano, T. Pedrazzini, M. Sisti, E. Gallo, F. Ragaini, 28 A. M. Harm, J. G. Knight and G. Stemp, Tetrahedron Lett.,
3
9
1
N. Casati and A. Caselli, Appl. Organomet. Chem., 2011, 25,
24–829.
1 B. Castano, S. Guidone, E. Gallo, F. Ragaini, N. Casati,
1996, 37, 6189–6192.
8
29 C. B ¨o hm, M. Schinnerl, C. Bubert, M. Zabel, T. Labahn,
E. Parisini and O. Reiser, Eur. J. Org. Chem., 2000, 2000,
2955–2965.
1
P. Macchi, M. Sisti and A. Caselli, Dalton Trans., 2013, 42,
2451–2462.
30 C. Bohm and O. Reiser, Org. Lett., 2001, 3, 1315–1318.
1
1
1
1
1
1
2 A. Caselli, F. Cesana, E. Gallo, N. Casati, P. Macchi, M. Sisti, 31 M. Itagaki and K. Suenobu, Org. Process Res. Dev., 2007, 11,
G. Celentano and S. Cenini, Dalton Trans., 2008, 4202–4205. 509–518.
3 J. I. Garc ´ı a, G. Jim ´e nez-Os ´e s, B. L ´o pez-S ´a nchez, J. A. Mayoral 32 A. G. M. Barrett, D. C. Braddock, I. Lenoir and H. Tone,
and A. V ´e lez, Dalton Trans., 2010, 39, 2098–2107.
J. Org. Chem., 2001, 66, 8260–8263.
4 J. M. Fraile, J. I. Garc ´ı a, C. I. Herrer ´ı as, J. A. Mayoral and 33 N. S. Youssef, A. M. A. El-Seidy, M. Schiavoni, B. Castano,
E. Pires, Chem. Soc. Rev., 2009, 38, 695–706.
F. Ragaini, E. Gallo and A. Caselli, J. Organomet. Chem.,
2012, 714, 94–103.
34 J. M. Fraile, J. I. Garc ´ı a, J. A. Mayoral, T. Tarnai and
M. A. Harmer, J. Catal., 1999, 186, 214–221.
5 J. M. Fraile, J. I. Garc ´ı a, G. Jim ´e nez-Os ´e s, J. A. Mayoral and
M. Rold ´a n, Organometallics, 2008, 27, 2246–2251.
6 J. I. Garc ´ı a, B. L ´o pez-S ´a nchez, J. A. Mayoral, E. Pires and
I. Villalba, J. Catal., 2008, 258, 378–385.
35 L. J. Liu, Y. Song and H. Li, Polym. Int., 2002, 51, 1047–1049.
7 J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, 36 T. Martin, A. Galarneau, F. Di Renzo, F. Fajula and D. Plee,
C. T. Kresge, K. D. Schmitt, C. T. W. Chu, D. H. Olson, Angew. Chem., Int. Ed., 2002, 41, 2590–2592.
E. W. Sheppard, et al., J. Am. Chem. Soc., 1992, 114, 10834– 37 A. Galarneau, N. Cambon, F. Di Renzo, R. Ryoo, M. Choi and
0843. F. Fajula, New J. Chem., 2003, 27, 73–79.
8 C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli and 38 A. Galarneau, H. Cambon, F. Di Renzo and F. Fajula,
J. S. Beck, Nature, 1992, 359, 710–712. Langmuir, 2001, 17, 8328–8335.
1
1
This journal is ª The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 22199–22205 | 22205