Amine-functionalised hexagonal mesoporous silica as support for copper(II) acetylacetonate catalyst

Article abstract of DOI:10.1002/ejic.200500979

Copper(II) acetylacetonate was anchored onto a hexagonal mesoporous silica (HMS) material using a two-step procedure: (i) functionalisation of the surface hydroxy groups with (3-aminopropyl)triethoxysilane (AMPTSi) and then (ii) anchoring of the copper(II) complex through Schiff condensation with free amine groups, using two different metal complex loadings. Upon the first step, nitrogen elemental analysis, XPS and DRIFT showed the presence of amine groups on the surface of the HMS material, and porosimetry indicated that the structure of the mesoporous material remained unchanged, although a slight decrease in surface area was observed. Atomic absorption, XPS and DRIFT showed that copper(II) acetylacetonate was anchored onto the amine-functionalised HMS by Schiff condensation between the free amine groups and the carbonyl groups of the copper(II) complex; using EPR an NO₃ coordination sphere was proposed for the anchored copper(II) complex. The new [Cu(acac) ₂]-AMPTSi/HMS materials were tested in the aziridination of styrene at room temperature, using PhI=NTs as nitrogen source and acetonitrile as solvent. The styrene conversion and total TON of the heterogeneous phase reaction are higher than those of the same reaction catalysed in homogeneous phase by [Cu(acac)₂]; nevertheless, the initial activity decreases and the reaction time increases due to substrate and product diffusion limitations. The heterogeneous catalyst showed a successive slight decrease in catalytic activity when reused for two more times. Wiley-VCH Verlag GmbH & Co. KGaA, 2006.

Full text of DOI:10.1002/ejic.200500979

FULL PAPER
DOI: 10.1002/ejic.200500979
Amine-Functionalised Hexagonal Mesoporous Silica as Support for Copper(II
Acetylacetonate Catalyst
)
Ana Rosa Silva,[a] Karen Wilson,[b] Adrian C. Whitwood,[c] James H. Clark,[b] and
Cristina Freire*[a]
Keywords: Copper / Anchoring / Mesoporous silica / EPR spectroscopy / Aziridination
Copper(II) acetylacetonate was anchored onto a hexagonal
mesoporous silica (HMS) material using a two-step pro-
cedure: (i) functionalisation of the surface hydroxy groups
with (3-aminopropyl)triethoxysilane (AMPTSi) and then (ii)
anchoring of the copper(II) complex through Schiff condensa-
tion with free amine groups, using two different metal com-
plex loadings. Upon the first step, nitrogen elemental analy-
sis, XPS and DRIFT showed the presence of amine groups on
the surface of the HMS material, and porosimetry indicated
that the structure of the mesoporous material remained un-
plex; using EPR an NO3 coordination sphere was proposed
for the anchored copper(II) complex. The new [Cu(acac)2]-
AMPTSi/HMS materials were tested in the aziridination of
styrene at room temperature, using PhI=NTs as nitrogen
source and acetonitrile as solvent. The styrene conversion
and total TON of the heterogeneous phase reaction are
higher than those of the same reaction catalysed in homo-
geneous phase by [Cu(acac)2]; nevertheless, the initial ac-
tivity decreases and the reaction time increases due to sub-
strate and product diffusion limitations. The heterogeneous
changed, although a slight decrease in surface area was ob- catalyst showed a successive slight decrease in catalytic ac-
served. Atomic absorption, XPS and DRIFT showed that cop-
per(II) acetylacetonate was anchored onto the amine-func-
tionalised HMS by Schiff condensation between the free
tivity when reused for two more times.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
amine groups and the carbonyl groups of the copper(II) com- Germany, 2006)
(salen)copper() complexes have also been applied as
Introduction
homogeneous catalysts for the alkene aziridination using
Currently, transition-metal complexes with Schiff-base li-
gands of the salen type are the object of intense research
due to their high activity, chemoselectivity and enantio-
selectivity in a large range of catalytic processes.[1–3] One of
the best-known examples are (salen)manganese() com-
plexes, that catalyse the epoxidation of alkenes,[1–3] and lead
(by direct atom transfer to olefins) to epoxides, which are
of great utility in organic synthesis.[1–3,4] The equivalent ni-
trogen-atom-transfer process catalysed by transition-metal
complexes has been less exploited,[4,5] despite aziridines be-
ing valuable products as intermediates in organic synthe-
sis,[4–7] efficient chiral auxiliaries and ligands in asymmetric
catalysis.[6] Evans et al. reported that copper() acetylace-
tonate is a very efficient homogeneous catalyst for the aziri-
dination of alkenes using 4-methyl-N-(phenyl-λ3-iodanylid-
ene)benzenesulfonamide (PhI=NTs) as nitrogen source;[4]
the same nitrogen source.[7]
However, homogeneous catalysts can not be separated
from the reaction media and, subsequently, can not be re-
used. Thus, one way to improve the homogeneous chemical
processes, preventing the requirement for laborious and in-
efficient extraction processes, is to immobilise the (some-
times) expensive homogeneous catalyst on solid supports.[8]
In fact, the last decade has witnessed a growing interest in
the heterogenisation of homogeneous metal complexes
using several types of supports and several immobilisation
strategies.[9–13] Initially, the complexes were just ion-ex-
changed or adsorbed on the porous supports and, conse-
quently, could be susceptible to leaching.[14,15] More re-
cently, several grafting and tethering procedures have been
developed to covalently attach transition-metal complexes
to organic polymers,[13] silica, zeolites and other micro- and
mesoporous inorganic materials.[9–12]
Copper() acetylacetonate has already been immobilised
in a polymer[6] and activated carbon,[16,17] and the new ma-
terials have acted as recoverable and reusable heterogeneous
catalysts for the aziridination of olefins using PhI=NTs as
nitrogen source.[6,17] In the first case, the copper() complex
was microencapsulated in polystyrene, and in the latter it
was anchored onto an amine-functionalised activated car-
[a] REQUIMTE/Departamento de Química, Faculdade de Ciên-
cias, Universidade do Porto,
4169-007 Porto, Portugal
Fax: +351-226082959
E-mail: acfreire@fc.up.pt
[b] Department of Chemistry, Green Chemistry Group, University
of York,
Heslington, York YO10 5DD, UK
[c] Department of Chemistry, University of York,
Heslington, York YO10 5DD, UK
Eur. J. Inorg. Chem. 2006, 1275–1283
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1275

A. R. Silva, K. Wilson, A. C. Whitwood, J. H. Clark, C. Freire
FULL PAPER
Scheme 1. Anchoring procedure: (a) functionalisation of the HMS with (3-aminopropyl)triethoxysilane and (b) anchoring copper()
acetylacetonate onto AMPTSi-functionalised HMS.
Scheme 2. Aziridination of styrene using PhI=NTs as nitrogen source.
bon,[16] through Schiff-base condensation between the free use of a cheap and neutral template, which does not require
amine group, covalently attached to the activated carbon the need of strongly acidic conditions.[22]
surface with the carbonyl groups of the acetylacetonate li-
In the present work, a new and simpler route to obtain
gand coordinated to copper(). Both heterogeneous cata- immobilised (Schiff base)copper() complexes is presented.
lysts presented similar catalytic parameters as those ob- A hexagonal mesoporous silica (HMS), organically func-
tained in homogeneous-phase reaction using [Cu- tionalised with (3-aminopropyl)triethoxysilane, was chosen
(acac)2], under similar experimental conditions, and no sig- to anchor copper() acetylacetonate. The anchoring process
nificant metal complex leaching was observed after the was done through Schiff condensation between the free
catalytic reactions.[6,17] However, in the case where activated amine group, covalently attached to the HMS surface
carbon was used as support, inherent diffusion limitations (Scheme 1), with the carbonyl groups of the acetylacetonate
were observed due to the porous structure of the support, ligand, coordinated to copper(), as depicted in Scheme 1.
and thus a decrease in the initial activity (TOF), and an All the materials were characterised by nitrogen elemental
increase in the reaction time was observed.[17] Furthermore, analysis, FTIR-DRIFT, nitrogen isotherms at 77 K and
the spectroscopic characterisation by DRIFT and UV spec- XPS, and those with immobilised copper() complex were
troscopy of the new carbon-based heterogeneous catalysts also characterised by copper atomic absorption, UV diffuse
proved to be difficult and impossible, respectively, due to reflectance spectroscopy and EPR. The catalytic activity of
the highly absorbing carbon matrix, which limited the gath- this new materials in the aziridination of styrene at room
ering of information about the copper() complex formed temperature using PhI=NTs as nitrogen source was studied
on the surface of the amine-functionalised activated car- (Scheme 2).
bon. Immobilisation on several porous supports of (Schiff
base)copper() complexes previously synthesised[18] or pre-
pared in situ on the porous structure,[19,20] have also been
reported in the literature.
Results and Discussion
Characterisation of the HMS and Amine Functionalisation
The silica-based inorganic mesoporous materials which
offer pore sizes in the range 20–100 Å, are more suitable for
The nitrogen adsorption isotherm at 77 K of the calcined
liquid-phase reactions, because they allow easy diffusion of HMS material is of type IV (Figure 1), and therefore is typi-
reactants to the active sites.[8,21] Their surface silanol groups cal of mesoporous materials, although the presence of
can also be easily functionalised by using trialkyloxysilanes, micropores can be detected in the lower relative pressure
with amine, thiol, carboxylic acid, phenyl, cyano, sulfonic region. The material has a BET surface area of 1158 m2 g–1
acid and vinyl groups, hence allowing subsequent chemical (Table 1).
attachment of the catalytic active species, namely inorganic
The infrared spectrum of the calcined HMS shows the
coordination compounds.[8,21] Furthermore, they can be typical Si–O lattice vibrations: two broad bands between
easily characterised by several spectroscopic techniques. De- 2100 and 1750 cm–1, with medium intensity, a strong and
spite of the hexagonal mesoporous silicas (HMS) being less broad band with two peaks in the region 1450–900 cm–1,
ordered than MCM-41 and SBA-15 materials, they are and two strong bands between 900 and 500 cm–1. Further-
much easier and straightforward to synthesise due to the more, a very broad band centred at 3450 cm–1, due to the
1276
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 1275–1283

Amine-Functionalised HMS as Support for CuII Acetylacetonate Catalyst
FULL PAPER
Figure 1. Nitrogen adsorption isotherms at 77 K of (a) calcined
HMS, (b) AMPTSi/HMS, (c) [Cu(acac)2]AMPTSi/HMS_A and (d)
[Cu(acac)2]AMPTSi/HMS_B.
Table 1. Selected characteristics of the modified HMS materials.
[b]
Sample
Cu [mmolg–1
AA
]
N/Cu
ABET
XPS[a] EA/AA XPS [m2 g–1
]
HMS
1158
561
AMPTSi/HMS
[Cu(acac)2]AMPTSi/
HMS_A
[Cu(acac)2]AMPTSi/
HMS_B
0.362 0.269
0.068 0.036
3.9
23
4.6
27
436
201
[a] mol Cu/weight of HMS (from XPS data in Table 2) Ϸ atom-%
Cu/[atom-% C×Ar(C) + atom-% O×Ar(O) + atom-% N×Ar(N)
+ atom-% Si×Ar(Si) + atom-% Cu×Ar(Cu)]. [b] From the nitro-
gen adsorption isotherm at 77 K.
Figure 2. FTIR-DRIFT spectra of (a) calcined HMS, (b) AMPTSi/
HMS, (c) [Cu(acac)2]AMPTSi/HMS_A, (d) [Cu(acac)2]AMPTSi/
HMS_B and (e) [Cu(acac)2] in the regions (A) 4000–500 cm–1 and
(B) 2000–1000 cm–1.
O–H stretching vibrations, and a very sharp band at
3745 cm–1, with medium intensity, which is assigned to the
stretching vibrations of isolated surface SiO–H groups, are
observed. A band at 1622 cm–1 is also observed in the spec-
trum and is attributed to the H–O–H bending vibrations of
physisorbed water (Figure 2).
Table 2. XPS elemental analysis [atom-%] of the modified HMS
materials.
The HMS material was functionalised with (3-aminopro-
pyl)triethoxysilane (AMPTSi) and both nitrogen elemental Sample
C
O
N
Si
Cu
and XPS analyses show that nitrogen (Tables 1 and 2) was
introduced on the surface of the material: 1.34 mmolg–1
and 0.999 mmolg–1, respectively. From the former value
and taking into account the AMPTSi used in the function-
alisation reaction (3.00 mmolg–1), the anchoring efficiency
of the amine derivative onto the support was 44.7%.
AMPTSi/HMS
[Cu(acac)2]AMPTSi/HMS_A 17.9
[Cu(acac)2]AMPTSi/HMS_B 13.7
18.3
57.1 1.8 22.9
53.7 2.3 25.6
51.1 1.9 33.2
–
0.5
0.07
7.0
[Cu(acac)2]
66.9
26.0
–
–
Upon amine functionalisation, a decrease in the intensity 1300 cm–1, respectively, as weak intensity bands (Fig-
of the band at 3745 cm–1 is observed (Figure 2A), indicating ure 1B).
that the reaction between the isolated silanol groups of the
No changes in the isotherm profile were observed upon
surface of the HMS with the ethoxy groups of the silane surface functionalisation of HMS with AMPTSi, suggest-
derivative took place. The N–H and C–H stretching vi- ing that the structure of the parent material was unaffected
brations of AMPTSi are observed in the region 3400– after this step (Figure 1); however, a decrease in the HMS
3000 cm–1 and 3000 cm–1, respectively, and the correspond- surface area (Table 1), as well as some blockage of the
ing H–N–H and H–C–H bending vibrations can also be micropores and a slight decrease in the mean mesopore size
observed in the regions 1700–1500 cm–1 and 1500– (Figure 1) are observed, as expected.
Eur. J. Inorg. Chem. 2006, 1275–1283
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1277

A. R. Silva, K. Wilson, A. C. Whitwood, J. H. Clark, C. Freire
FULL PAPER
Anchoring of Copper(II) Acetylacetonate onto the Amine-
Functionalised HMS
surface of the HMS material and the oxygen atoms coordi-
nated to copper(), and that therefore the anchoring pro-
cedure occurred as depicted in Scheme 1.
Two chloroform solutions containing different quantities
The electronic spectrum of free [Cu(acac)2] presents a
of copper() acetylacetonate were refluxed for 24 h with the large and asymmetric band with a maximum at λ = 558 nm
amine-functionalised HMS. Evidence for the effective and a shoulder lying lower in energy at about λ = 642 nm
anchorage of the copper() acetylacetonate complex onto (Figure 3); this spectrum is typical of CuII complexes with
the surface of AMPTSi/HMS was first provided by noting a square-planar geometry.[16] In this context the band at
that (a) the blue colour of the complex solution disappears lower energy is assigned to the dxy ǟ dz2 transition and the
during the adsorption reaction, and (b) colourless solutions band at higher energies is attributed to the envelop of two
are obtained when the materials were purified by Soxhlet other allowed d–d transitions: dxy ǟ dx2–y2 and dxy ǟ dxz
extraction.
Ϸ dyz, which have very close energies.[24]
Atomic absorption analysis (Table 1) of the materials re-
vealed that the [Cu(acac)2]AMPTSi/HMS_A material con-
tains 362 µmol of copper/g of material and the [Cu(acac)2]-
AMPTSi/HMS_B material contains 68 µmol of copper/g.
Relatively to the copper() complex quantity initially pres-
ent in each solution, these values correspond to an anchor-
ing efficiency of 73% and 82%, respectively, suggesting the
presence of some diffusion control in the metal complex
immobilisation reactions.
The N/Cu ratios obtained by elemental analysis and XPS
have similar magnitudes (Table 1) suggesting that the an-
chored copper() species have the same chemical environ-
ment throughout the HMS matrix. Nevertheless, in both
materials the spacer AMPTSi is always in excess relative to
complex (larger excess in [Cu(acac)2]AMPTSi/HMS_B).
No change in the isotherm profile was observed upon
complex immobilisation (Figure 1), suggesting that the
functionalised HMS structure was unaffected by complex
anchoring; however, a further decrease in the material sur-
face area (Table 1 and Figure 1), and a slight decrease in
the mean mesopore size are observed, confirming indirectly
the presence of the complex within the matrices. Surpris-
ingly, the material with lower copper loading, [Cu(acac)2]-
Figure 3. UV diffuse reflectance spectra of the HMS materials: (a)
calcined HMS, (b) [Cu(acac)2], (c) [Cu(acac)2]AMPTSi/HMS_A
and (d) [Cu(acac)2]AMPTSi/HMS_B.
The electronic spectrum of the [Cu(acac)2]AMPTSi/
AMPTSi/HMS_B, shows a larger decrease in the surface HMS_A material shows a broad asymmetric band at λ =
area, specially in the mesopore region, than the material 823 nm tailoring to lower energies, which corresponds to
with higher copper loading, [Cu(acac)2]AMPTSi/HMS_A.
the envelop of the three allowed d–d transitions (dxy ǟ dz2,
The IR spectra of [Cu(acac)2AMPTSi]/HMS_A and [Cu- dxy ǟ dx2–y2 and dxy ǟ dxz Ϸ dyz) with close energies. In
(acac)2AMPTSi]/HMS_B show, with respect to the parent comparison to the spectrum of free [Cu(acac)2], the maxi-
material, some changes in the regions where vibration from mum of the band is shifted towards lower energy (Figure 3);
AMPTSi occur: 3500–3000 cm–1 region, due to N–H this type of changes in the electronic spectrum of CuII com-
stretching vibrations and 1700–1300 cm–1, weak absorp- plexes can be interpreted as suggesting a tetrahedral distor-
tions due to the H–N–H and H–C–H bending vibrations, tion in the equatorial plane or axial coordination of a new
suggesting that some molecular changes in the spacer took ligand.[24] Although the first explanation could justify the
place. On the other hand, new broad bands are observed at proposed complex anchoring model, on the basis of elec-
about 1658 and 1555 cm–1 (Figure 2), which may be attrib- tronic data alone, no choice between the two hypotheses
uted to a C=N stretching vibration and to the acetylace- can be made. The profile spectrum of [Cu(acac)2]AMPTSi/
tonate ring vibrations of the metal complex, respectively. HMS_B is similar to the previous one, but its low intensity,
Furthermore, in the region between 1700 and 1300 cm–1, due to the lower copper content in this material, prevents
where the fingerprint bands of [Cu(acac)2] are usually de- any further interpretation (Figure 3).
tected, the DRIFT spectra of both [Cu(acac)2AMPTSi]/
Further insights into the copper() coordination sphere
HMS_A and [Cu(acac)2AMPTSi]/HMS_B materials are upon complex anchoring can be gathered from EPR spec-
different from that of free [Cu(acac)2] (Figure 2), contrast- troscopy. The EPR spectrum of [Cu(acac)2]AMPTSi/
ing with a previous report for silica-supported [Cu(acac)2], HMS_B at 120 K, depicted in Figure 4 (b), is typical of
which shows a very similar DRIFT spectrum to that of the copper() species: it is of the axial type and shows hyperfine
free complex.[23] These results clearly indicate that there was couplings due to the interaction of the unpaired electron
a Schiff condensation between the free amine groups on the with the copper nucleus (65Cu/63Cu, I = 3/2) in the low
1278
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 1275–1283

Amine-Functionalised HMS as Support for CuII Acetylacetonate Catalyst
FULL PAPER
magnetic field region (Figure 4b). The EPR spectrum of as free complexes and immobilised onto other supports
[Cu(acac)2]AMPTSi/HMS_A (Figure 4a) shows band found in the literature.
broadening due to the high local copper() concentration
(2.30%, Table 1), which prevents any further analysis. In
contrast, the EPR spectrum of the [Cu(acac)2]AMPTSi/
HMS_B shows smaller band-broadening effects, and thus
could be analysed in detail.
Tetracoordinated copper() complexes with an equato-
rial O4 coordination sphere generally possess higher gʈ val-
ues (2.26–2.29)[26–28] relative to copper() complexes with
an equatorial N2O2 coordination sphere (2.19–2.24),[18,29]
because the introduction of nitrogen atoms into the equato-
rial plane increases the in-plane covalence. This can be con-
firmed by the data extracted from literature and compiled
in Table 3. An NO3 coordination sphere would give rise to
intermediate g values, with some overlap with the intervals
given above, but confirmation with experimental data is
quite difficult due not only to the failure of the preparation
the free counterpart of the anchored complex (all synthetic
attempts resulted in oils which we were unable to fully char-
acterise), but also to the almost nonexistent EPR data for
compounds with these coordinating atoms: only Murphy et
al. reported a silica-anchored [Cu(salicylaldimine)] species
with an NO3 coordinating sphere, which showed a gʈ =
2.29[20] and Talzi et al. reported CuII complexes with an
equatorial NO3 coordination sphere with gʈ = 2.27–2.28.[30]
Nevertheless, some insights into the coordination sphere of
the anchored complex can be gathered indirectly by com-
parison with g and A values of the parent copper complex
and those with mixed NO3 coordination spheres shown in
Table 3.
Figure 4. EPR spectra of the copper() acetylacetonate complexes,
free and anchored onto HMS at 120 K: (a) [Cu(acac)2]AMPTSi/
HMS_A, (b) [Cu(acac)2]AMPTSi/HMS_B and (c) [Cu(acac)2]/
HMS.
Comparing the g values of [Cu(acac)2]AMPTSi/HMS_B
with those of the free copper() acetylacetonate in a non-
coordinating solvent (CHCl3, Table 3) a decrease in gʈ and
The EPR band profile of this material suggests that there gЌ; values, with consequent decrease in the gav value, is
is one major CuII species within the HMS matrix[20] and its observed; these g factors tendencies suggest that the an-
similarity with the EPR of analogous free CuII com- chored copper() complex has a different coordination
plexes[18–20,25–28] allows the same orientation scheme for the sphere with higher equatorial covalence,[28] which is typical
tensor axes: g1 = g2 = gЌ, g3 = gʈ (or g1 = gx, g2 = gy, g3 = of the involvement of nitrogen atoms in the direct coordina-
gz), where g1 and g3 refer to the lowest and highest magnetic tion with CuII.[20,28] As gʈ = 2.275 is higher than the typical
fields g values (obviously, A1 = A2 = AЌ, A3 = Aʈ or A1 = values observed for copper() complexes with an N2O2 co-
Ax, A2 = Ay, and A3 = Az). In this context, as gx = gy Ͻ gz ordinating sphere and similar to those observed by Talzi et
the ground state of CuII in the immobilised complex has al.,[30] we propose that the major Cu species immobilised
been kept and is dxy; the values of the tensors g and A of in [Cu(acac)2]AMPTSi/HMS_B have an NO3 coordinating
the [Cu(acac)2]AMPTSi/HMS_B were obtained by simula- sphere as shown in Scheme 1 (other species, for example,
tion and are compiled in Table 3, which also summarises with an N2O2 coordination sphere and present in low quan-
the values for similar copper() acetylacetonate/salen-based tities, cannot be unambiguously excluded because of the so-
complexes with O4, NO3 and N2O2 coordination spheres, lid-state broadening effects in the EPR spectra). Although
Table 3. EPR parameters of the free and immobilised [Cu(acac)2].
|Aʈ|[b]
|AЌ|[b]
Ref.
[a]
Sample (coordination sphere)
gʈ
gЌ
gav
[d]
[d]
[d]
[d]
[d]
[Cu(acac)2]AMPTSi/HMS_A[c,d]
[Cu(acac)2]AMPTSi/HMS_B[c]
2.275
2.057
2.130
180.0
12.0
[e]
[25]
[19]
[Cu(acac)2]/CHCl3 (O4)
[Cu(acac)2]/MeHO (O4)
2.285
2.290
2.060
2.067
2.135
2.141
180
165.0
12
15.4
[20]
[20]
[Cu(salimine)]/SiO2 (NO3)
[Cu(salimine)]/SiO2 (N2O2)
2.290
2.243
2.065
2.065
2.062
2.140
2.124
155
183
13
13
16.8
[18]
[Cu(4-OHsalen)]/MeHO (N2O2)
2.213
2.111
192.2
2.055
36.5
[a] gav = 1/3(2×gЌ + gʈ). [b] The values are expressed in 10–4 cm–1. [c] At 120 K. [d] Not simulated due to band-broadening effects. [e]
This work.
Eur. J. Inorg. Chem. 2006, 1275–1283
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1279

A. R. Silva, K. Wilson, A. C. Whitwood, J. H. Clark, C. Freire
FULL PAPER
the spectrum was taken at low temperature, the bandwidths HMS is hindered in the amine-functionalised HMS mate-
of the signals are still large to detect the nitrogen superhy- rial.
perfine splittings (typically 14–25 G), due to the interaction
The EPR spectrum of this material, [Cu(acac)2]/HMS,
between the copper unpaired spin and the nitrogen nuclear was also recorded and is shown in Figure 4c. Although the
spin (I = 1), hence hindering the direct confirmation of ni- spectrum could not be simulated due to the low intensity
trogen atoms coordinated to the copper() centre in the of the signal, some qualitative information relating to the
equatorial plane. Further confirmation of the proposed anchored complex can be gathered from the spectrum pro-
NO3 coordination sphere for Cu species in [Cu(acac)2]- file. With respect to the spectrum of [Cu(acac)2]AMPTSi/
AMPTSi/HMS_B and the exclusion of complex immobilis- HMS_B (Figure 4b), that of [Cu(acac)2]/HMS shows the
ation through axial coordination to the free amines of the parallel region at higher magnetic field values as a broad
support can be gathered from the AʈCu values. Axial coordi- band, which corresponds to the envelope of the unresolved
nation to a square-planar complex usually leads to a de- four features due to the Cu hyperfine splittings. The main
crease of the metal hyperfine couplings:[26–29] in our case, consequence of this is an increase in gʈ and a decrease in
upon immobilisation of the complex onto the support, no AʈCu. By using the data from analogous Cu complexes[26–29]
change in the AʈCu value was observed, suggesting that no and those referred to in Table 3, it can be clearly shown
effective axial coordination took place with the function- that in the unfunctionalised HMS, the Cu centre lies in an
alised matrix.
equatorial O4 plane and possesses axial coordination to the
Despite the band-broadening effects in the [Cu(acac)2]- matrix hydroxy groups. On the contrary, in the NH2-func-
AMPTSi/HMS_A spectrum that prevented its simulation tionalised HMS, the Cu centre has a coordination sphere
and detailed analysis, the same coordinating sphere is pro- which involves the nitrogen atoms of the modified matrix
posed, because the N/Cu ratio is much lower than in [Cu- and no significant axial coordination takes place.
(acac)2]AMPTSi/HMS_B (Table 1), and thus the in situ for-
mation of an anchored copper() complex with an N2O2
coordinating sphere is less likely to occur.
Catalytic Experiments
It is noteworthy that in a control experiment, where a
[Cu(acac)2] solution in chloroform was refluxed with the
The results obtained for the aziridination of styrene, cat-
unfunctionalised HMS for 24 h, a very low copper content alysed by the [Cu(acac)2] in homogeneous phase, and im-
(see Figure 4) was obtained after purification. This indi- mobilised onto the amine-functionalised HMS are collected
cates that the functionalisation of the support with amine in Table 4, at room temperature, using PhI=NTs as nitrogen
groups leads to higher metal loadings, as observed by us source.
when activated carbon was used as the heterogeneous sup-
Both [Cu(acac)2]AMPTSi/HMS_A and [Cu(acac)2]-
port.[16] The anchoring process of the copper() acetylace- AMPTSi/HMS_B materials acted as heterogeneous cata-
tonate onto the calcined HMS must occur through axial lysts in the aziridination of styrene. As it was not possible
coordination of the CuII centre to the isolated hydroxy to prepare the free counterparts of the anchored complexes,
groups of the HMS surface. The confirmation was achieved the catalytic activity of the heterogeneous copper complexes
by the observation in the FTIR spectrum of a decrease in were compared with the parent free complex [Cu(acac)2].
the intensity of the band at 3745 cm–1, which is due to the In this context, the former heterogeneous catalyst showed
isolated silanol groups. Taking into account that upon improved catalytic efficiency (styrene conversion and total
amine functionalisation of HMS there is a significant de- TON, Figure 5) with respect to free [Cu(acac)2] in the
crease in the intensity of this band (vide supra, Figure 2), it homogeneous phase, under similar experimental conditions.
may be concluded that the axial coordination of the CuII This is an important result, as the modification of the coor-
centre to the isolated hydroxy groups of the surface of the dination sphere of the parent complex upon immobilisation
Table 4. Aziridination of styrene catalysed by the [Cu(acac)2] in the homogeneous phase and heterogenised using PhI=NTs as nitrogen
source.[a]
Catalyst
Run
t [h][b]
0.5
% C[c]
69
Total TON[d]
11
Initial TOF [h–1][e]
21
[Cu(acac)2]
AMPTSi/HMS
[Cu(acac)2]AMPTSi/HMS_A
24
21
1st
2nd
3rd
4
24
24
92
79
40
21
24
14
8
7
5
[Cu(acac)2]AMPTSi/HMS_B
1st
3rd
24
24
15
12
3
13
0.5
0
[a] Catalyst/styrene/chlorobenzene (internal standard)/PhI=NTs (1:100:100:20), in CH3CN, at room temperature. [b] Time needed for total
consumption of PhI=NTs. [c] Corrected for limiting reagent (PhI=NTs). [d] TON = mmol converted styrene/mmol Cu. [e] Initial TOF
calculated for 1 h of reaction.
1280
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 1275–1283

Amine-Functionalised HMS as Support for CuII Acetylacetonate Catalyst
FULL PAPER
Figure 5. (a) Styrene conversion (%) and (b) total TON of the [Cu(acac)2] in the homogeneous phase and heterogenised onto an amine-
functionalised HMS.
led to an increase in the catalytic activity, contrasting with product mixture, can be related with residual catalytic ac-
usual decrease observed in the majority of the anchorage tivity of the support.
procedures, which involve the coordination sphere of the
catalytic centres.
The materials could be reused at least in two other cata-
lytic reactions with formation of aziridine. However, the
However, a decrease in the initial activity and increase in catalyst [Cu(acac)2]AMPTSi/HMS_A showed a loss of the
reaction time was observed, which may be due to the in- catalytic activity with reuse, giving, nevertheless, better
herent diffusion limitations within the HMS porous struc- catalytic efficiency (total TON) than the reaction run in
ture, as the nitrogen source is a solid with limited solubility homogeneous phase with the [Cu(acac)2], under similar ex-
in acetonitrile, and its solubilisation is controlled by its rate perimental conditions. The catalyst [Cu(acac)2]AMPTSi/
of consumption.[4] Besides the formation of p-toluenesul- HMS_B showed improved catalytic efficiency in the 2nd re-
fanamide, due to competitive homolytic cleavage of the utilisation, and then a small drop in the 3rd cycle, which
PhI=NTs,[4] only very small amounts of benzaldehyde and may be explained by a considerable decrease in the catalyst
epoxide were detected in the reaction media by gas mass during the recycling procedure.
chromatography.
Nonetheless, the typical IR bands of the immobilised
The AMPTSi/HMS support also shows some catalytic (Schiff base)copper() complex and silica backbone of both
activity in the conversion of styrene, but much smaller com- catalysts after the catalytic cycles remained unchanged, in-
pared with the new heterogeneous catalysts. Nevertheless, dicating the stability of the catalytic active species and sup-
the amount of the benzaldehyde by-product observed in the port itself under the catalysis experimental conditions.
Eur. J. Inorg. Chem. 2006, 1275–1283
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1281

A. R. Silva, K. Wilson, A. C. Whitwood, J. H. Clark, C. Freire
FULL PAPER
(HPLC grade). The nitrene donor 4-methyl-N-(phenyl-λ3-iodanyl-
idene)benzenesulfonamide (PhI=NTs) was synthesised using re-
ported procedures.[31]
Hence, the loss of catalytic activity with reuse of the new
heterogeneous catalyst may be due to slight leaching of
active species, promoted by some degradation of the bond
between the support and the spacer.
Synthesis and Modification of Hexagonal Mesoporous Silica (HMS)
Synthesis of Hexagonal Mesoporous Silica (HMS): The HMS was
The [Cu(acac)2]AMPTSi/HMS_A heterogeneous catalyst
reported herein, which consists of copper() acetylace-
tonate anchored onto an amine-functionalised hexagonal
mesoporous silica, presents higher catalytic activity in the
aziridination of styrene, at room temperature, using
PhI=NTs as nitrogen source, than [Cu(acac)2] anchored
onto a triamine-functionalised activated carbon.[17] How-
ever, it is not as stable and reusable as the latter hetero-
geneous catalyst, because this could be reused at least for 4
times with successive increase in styrene conversion, and
thus in the total turnover number (TON) and initial activity
(TOF).
synthesised using
a
reported procedure;[22] tetraethoxysilane
(37.0 mL, 0.166 mol) was added to a stirred solution of ethanol
(88.1 mL, 1.51 mol), water (88.5 mL, 4.91 mol) and 1-dode-
cylamine (10.3 mL, 0.0448 mol). The mixture was stirred at room
temperature for 24 h. The white precipitate obtained was vacuum-
filtered, washed with deionised water (100 mL) and ethanol
(100 mL). In order to remove the template (1-dodecylamine), the
precipitate was calcined at 600 °C for 24 h.
Functionalisation of the HMS with (3-Aminopropyl)triethoxysilane
(AMPTSi/HMS): A mixture of calcined HMS (3.00 g) in dry tolu-
ene (50.0 mL) and (3-aminopropyl)triethoxysilane (9.00 mmol,
2.10 mL) was refluxed for 24 h. The material was vacuum-filtered,
washed with toluene (3×50 mL) and dried overnight in an oven at
85 °C.
Conclusion
Anchoring of Copper(II) Acetylacetonate onto Amine-Functionalised
HMS: A solution of copper() acetylacetonate (0.1304 g, 498 µmol)
in chloroform (100 mL, light blue solution) was refluxed with
AMPTSi/HMS (1.02 g) for 24 h. At the end of the process, the
colour of the solution was sligthtly blue. The solid was vacuum-
filtered, washed with chloroform (3×20 mL), and then Soxhlet-
extracted with chloroform for 6 h. Finally, the solid was dried over-
night in an oven at 85 °C. This material will be referred to as [Cu-
(acac)2]AMPTSi/HMS_A. Another material with lower complex
loading was also prepared: [Cu(acac)2]AMPTSi/HMS_B. In this
case AMPTSi/HMS (0.35 g) was refluxed with copper() acetylace-
tonate (0.0077 g, 29 µmol) in chloroform (10 mL, light blue solu-
tion) for 24 h, after which an almost colourless solution was ob-
tained after filtration of the material. The material was washed with
chloroform (3×20 mL) and then Soxhlet-extracted with chloro-
form (150 mL) for 6 h. The blue solid was dried in an oven at
100 °C, after which it was light brown.
Copper() acetylacetonate was anchored onto an HMS
material using a two-step procedure: (i) functionalisation of
the surface hydroxy groups with (3-aminopropyl)triethoxy-
silane and then (ii) anchoring of the copper() complex
through Schiff condensation with free amine groups, using
two different metal loadings.
Nitrogen elemental analysis, XPS and DRIFT showed
the presence of amine groups on the surface of the HMS
material and porosimetry indicated that the structure of the
mesoporous material remained unchanged after this step.
Atomic absorption and XPS data showed that the cop-
per() acetylacetonate could be irreversibly immobilised
onto the amine-functionalised HMS through Schiff conden-
sation between the surface amine groups and the oxygen
atoms of the copper() ligand, leading to an NO3 coordina-
tion sphere for the anchored metal complex. This was con-
firmed by the combination of data extracted from FTIR
and EPR spectra of the materials.
Physico-Chemical Measurements: Nitrogen elemental analysis and
copper atomic absorption were performed by Microanalytical Ser-
vices, University of Manchester, U. K. Surface area measurements
were performed monitoring the N2 adsorption isotherms at 77 K
with a Micrometrics ASAP 2010 apparatus. Surface areas were cal-
The new [Cu(acac)2]AMPTSi/HMS materials were tested
in the aziridination of styrene, using PhI=NTs as nitrogen
source and acetonitrile as solvent, at room temperature. The
styrene conversion and total TON of the heterogeneous
phase reaction are higher than those of the same reaction
catalysed in homogeneous phase by [Cu(acac)2]; neverthe-
culated using the BET equation over the pressure range p/p0
=
0.02–0.2. X-ray photoelectron spectroscopy was carried out with a
Kratos AXIS His instrument equipped with a charge neutraliser
and an Mg-Kα X-ray source. Spectra were recorded at normal emis-
sion with an analyser pass energy of 20 eV and an X-ray power of
less, the initial activity decreases and reaction time increases 225 W. The FTIR-DRIFT spectra of the powered materials were
obtained with a Bruker Equinox 55 spectrophotometer in the re-
gion 400–4000 cm–1. The UV diffuse reflectance spectra were ob-
tained with a Shimadzu UV-3101 PC in the range 1600–200 nm,
using barium sulfate as reference. EPR spectra were recorded with
a Bruker ESP300 EPR spectrometer equipped with an X-band mi-
crowave bridge. Temperature control was achieved with a Bruker
ER-4111 variable-temperature controller, the temperature at the
centre of the EPR cavity was calibrated using a separate thermo-
couple. The spectra were taken at 120 K with the following typical
experimental conditions: modulation frequency: 100 kHz; modula-
tion amplitude: 5 G; microwave power: 1–5 mW. The EPR param-
eters were obtained by simulation using the program Win EPR
Simfonia (Bruker) assuming axial or rhombic spin Hamiltonians.
The reported values of gЌ (or gx and gy) and AЌ (or Ax and Ay)
are less accurate because of their dependence on the used line
due to substrate and product diffusion limitations. The
heterogeneous catalyst showed a successive slight decrease
in catalytic activity when reused for two more times.
Experimental Section
Materials and Reagents: All reagents and solvents used in the prep-
aration and modification of the hexagonal mesoporous silica
(HMS) were used as received. Copper() acetylacetonate, 1-dode-
cylamine and (3-aminopropyl)triethoxysilane (AMPTSi) were from
Aldrich; tetraethoxysilane was from Lancaster; toluene and other
solvents used in the synthesis procedures were from Fischer Scien-
tific (analytical reagent grade) and those used in the catalytic ex-
periments and as eluents were from Romil or Fischer Scientific
1282
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 1275–1283

Amine-Functionalised HMS as Support for CuII Acetylacetonate Catalyst
FULL PAPER
widths (50–60 G). GC-FID chromatograms were obtained with a
[1] J. F. Larrow, E. N. Jacobsen, Top. Organomet. Chem. 2004, 6,
123–152.
[2] P. G. Cozzi, Chem. Soc. Rev. 2004, 33, 410–421.
[3] T. Katsuki, Chem. Soc. Rev. 2004, 33, 437–444.
[4] D. A. Evans, M. M. Faul, M. T. Bilodeau, J. Org. Chem. 1991,
56, 6744–6746.
[5] J. B. Sweeney, Chem. Soc. Rev. 2002, 31, 247–258.
[6] M. Lakshmi Kantam, B. Kavita, V. Neeraja, Y. Haritha, M. K.
Chaudhuri, S. K. Dehury, Tetrahedron Lett. 2003, 44, 9029–
9032.
Varian CP-3380 gas chromatograph using helium as carrier gas and
a fused silica Varian Chrompack capillary column CP-Sil 8 CB
Low Bleed/MS (30 m×0.25 mm i.d.; 0.25 µm film thickness). Con-
ditions: 60 °C (3 min), 5 °C min–1, 170 °C (2 min), 20 °C min–1,
200 °C (10 min); injector temperature: 200 °C; detector tempera-
ture: 300 °C. HPLC chromatograms were obtained with a Merck
Hitachi Lachrom apparatus consisting of a pump L-7100, UV de-
tector L-7400, interface D-7000, autosampler L-7200 and a column
oven L-7300 equipped with a Phenomenex Luna 5u C18(2),
250×4.6 mm column. The eluent used was a mixture of 85% of
acetonitrile and 15% of water with a flow rate of 1 mL min–1.
[7] W. Sun, E. Herdtweck, F. E. Kühn, New J. Chem. 2005, 29,
1577–1580.
[8] P. M. Price, J. H. Clark, D. J. Macquarrie, J. Chem. Soc., Dal-
ton Trans. 2000, 101–110.
Catalysis Experiments: The activity of the catalysts in the aziridin-
ation of styrene was studied at room temperature using a molar
ratio of metal catalyst/alkene/chlorobenzene (internal standard)/
PhI=NTs = 1:100:100:20 in acetonitrile (5.00 mL), whilst stirring.
For the experiments with the catalyst [Cu(acac)2]AMPTSi/
HMS_A, styrene (1.22 mmol), chlorobenzene (1.22 mmol, internal
standard), heterogeneous catalyst (0.0300 g) and PhI=NTs
(0.244 mmol, nitrogen source) were used; for the experiments with
catalyst [Cu(acac)2]AMPTSi/HMS_B, styrene (0.20 mmol), chloro-
benzene (0.20 mmol, internal standard), heterogeneous catalyst
(0.0300 g) and PhI=NTs (0.041 mmol, nitrogen source) were used.
During the experiment, 0.05-mL aliquots were taken from the solu-
tion with a hypodermic syringe, filtered through 0.2-µm PTFE sy-
ringe filters and directly analysed by GC-FID. When the entire
PhI=NTs had disappeared from the solution and the ratio of the
areas of iodobenzene and chlorobenzene in the chromatogram was
constant, the solution was centrifuged at 3500 rpm for 15 min, and
then to the decanted solution another portion of PhI=NTs
(0.244 mmol) was added. This procedure checks the eventual exis-
tence of metal complex leaching to the reaction medium. The pres-
ence of the aziridine at the end of the reaction was confirmed by
HPLC. The heterogeneous catalysts were then washed/centrifuged
sequentially with acetonitrile (3×10 mL) and methanol (1×10 mL)
and oven-dried at 90 °C overnight. These materials were reused for
further two times using the same experimental procedure. To pro-
vide a framework for the results obtained using the heterogenised
copper() complex, styrene aziridination was also carried out under
experimental conditions comparable to those described above (a)
in homogeneous media using the same amount of [Cu(acac)2] and
(b) using the AMPTSi/HMS material, the precursor of both hetero-
geneous catalysts.
[9] Q.-H. Xia, H.-Q. Ge, C.-P. Ye, Z.-M. Liu, K.-X. Su, Chem.
Rev. 2005, 105, 1603–1662.
[10] C. Li, Catal. Rev. 2004, 46, 419–492.
[11] Q.-H. Fan, Y.-M. Li, A. S. C. Chan, Chem. Rev. 2002, 102,
3385–3466.
[12] D. Brunel, N. Belloq, P. Sutra, A. Cauvel, M. Laspéras, P. Mo-
reau, F. Di Renzo, A. Galarneau, F. Fajula, Coord. Chem. Rev.
1998, 178–180, 1085.
[13] N. E. Leadbeater, M. Marco, Chem. Rev. 2002, 102, 3217–3274.
[14] J. S. Rafelt, J. H. Clark, Catal. Today 2000, 57, 33–44.
[15] I. W. C. E. Arends, R. A. Sheldon, Appl. Catal. A 2001, 212,
175–187.
[16] A. R. Silva, M. Martins, M. M. A. Freitas, J. L. Figueiredo, C.
Freire, B. de Castro, Eur. J. Inorg. Chem. 2004, 2027–2035.
[17] A. R. Silva, J. L. Figueiredo, C. Freire, B. de Castro, Catal. To-
day 2005, 102–103, 154–159.
[18] A. R. Silva, C. Freire, B. de Castro, M. M. A. Freitas, J. L. Fig-
ueiredo, Langmuir 2002, 18, 8017–8024.
[19] R. Ferreira, M. Silva, C. Freire, B. de Castro, J. L. Figueiredo,
Microporous Mesoporous Mater. 2000, 38, 391–401.
[20] E. F. Murphy, D. Ferri, A. Baiker, S. Van Doorslaer, A.
Schweiger, Inorg. Chem. 2003, 42, 2559–2571.
[21] K. Wilson, A. F. Lee, D. J. Macquarrie, J. H. Clark, Appl. Ca-
tal. A 2002, 228, 127–133.
[22] P. T. Tanev, T. J. Pinnavaia, Science 1995, 267, 865–867.
[23] M. B. Mitchell, V. R. Chakravarthy, M. G. White, Langmuir
1994, 10, 4523–4529.
[24] A. B. P. Lever, Inorganic Electronic Spectroscopy, 2nd ed., El-
sevier, Amesterdan, 1984, vol. II, p. 554.
[25] J. V. Zanchetta, B. Deroide, B. Asri, J. C. Giuntini, Microporous
Mater. 1995, 4, 187–193.
[26] H. R. Gersmann, J. D. Swalen, J. Chem. Phys. 1962, 36, 3221–
3233.
[27] J. I. Zink, R. S. Drago, J. Am. Chem. Soc. 1972, 94, 4550–4554.
[28] H. Yokoi, A. W. Addison, Inorg. Chem. 1977, 16, 1341–1349.
[29] B. Jezowska-Trzebiatowska, J. Jezierska, J. Mol. Struct. 1973,
19, 627–634.
Acknowledgments
[30] E. P. Talzi, V. M. Nekipelov, K. I. Zamaraev, Russ. J. Phys.
Chem. 1984, 58, 165.
[31] S. Taylor, J. Gullick, P. McMorn, D. Bethell, P. C. Bul-
man Page, F. E. Hancock, F. King, G. J. Hutchings, J. Chem.
Soc., Perkin Trans. 2 2001, 1714–1723.
A. R. S. thanks the “Fundação para a Ciência e Tecnologia” (Lis-
boa) and the European Social Fund for a Post-Doctoral fellowship.
This work was funded by the Fundação para a Ciência e Tecnolo-
gia (Portugal) and the FEDER through project ref. POCTI/CTM/
45449/2002.
Received: November 3, 2005
Published Online: January 26, 2006
Eur. J. Inorg. Chem. 2006, 1275–1283
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1283