Chiral Copper(II) Bis(oxazoline) Complexes Directly Coordinated to Amine-Functionalized Phenylene/Biphenylene Periodic Mesoporous Organosilicas as Heterogeneous Catalysts

Article abstract of DOI:10.1002/ejic.201501196

Copper(II) complexes with commercial chiral bis(oxazoline)s were for the first time directly coordinated onto amine-modified mesoporous phenylene and biphenylene silicas. All final materials maintained the 2D ordered mesoporous structure. The copper metal coordinates directly to the amine groups within the walls of the materials. The materials were tested as asymmetric heterogeneous catalysts in the kinetic resolution of hydrobenzoin. They were active, selective, and enantioselective in this reaction. The material containing the (S)-(-)-2,2′-isopropylidene-bis(4-phenyl-2-oxazoline) (Me₂PhBox) ligand, the best ligand in the homogeneous phase, presented the highest enantioselectivity of all the materials in the first cycle (73 %). The 2,2′-methylenebis[(4S)-4-phenyl-2-oxazoline] (PhBox) ligand rendered the catalyst more stable, independent of the anion or organic moiety in the periodic mesoporous organosilica material; this catalyst was reused over five cycles without any significant loss of catalytic activity or enantioselectivity. This type of ligand plays an important role in the stability of the corresponding copper(II) complex upon immobilization and in the robustness of the heterogeneous catalyst upon reuse.

Full text of DOI:10.1002/ejic.201501196

DOI: 10.1002/ejic.201501196
Full Paper
Supported Catalysts
Chiral Copper(II) Bis(oxazoline) Complexes Directly Coordinated
to Amine-Functionalized Phenylene/Biphenylene Periodic
Mesoporous Organosilicas as Heterogeneous Catalysts
Liliana Carneiro,[a] Ana Rosa Silva*[a] Mirtha A. O. Lourenço,[b] Alvaro Mayoral,[c]
Isabel Diaz,[d] and Paula Ferreira*[b]
Abstract: Copper(II) complexes with commercial chiral bis(oxaz- ligand in the homogeneous phase, presented the highest enan-
oline)s were for the first time directly coordinated onto amine- tioselectivity of all the materials in the first cycle (73 %). The
modified mesoporous phenylene and biphenylene silicas. All 2,2′-methylenebis[(4S)-4-phenyl-2-oxazoline] (PhBox) ligand
final materials maintained the 2D ordered mesoporous struc- rendered the catalyst more stable, independent of the anion or
ture. The copper metal coordinates directly to the amine groups organic moiety in the periodic mesoporous organosilica mate-
within the walls of the materials. The materials were tested as rial; this catalyst was reused over five cycles without any signifi-
asymmetric heterogeneous catalysts in the kinetic resolution of cant loss of catalytic activity or enantioselectivity. This type of
hydrobenzoin. They were active, selective, and enantioselective ligand plays an important role in the stability of the corre-
in this reaction. The material containing the (S)-(–)-2,2′-isoprop- sponding copper(II) complex upon immobilization and in the
ylidene-bis(4-phenyl-2-oxazoline) (Me2PhBox) ligand, the best robustness of the heterogeneous catalyst upon reuse.
Introduction
Asymmetric catalysis is unique in the sense that a minute quan-
tity of chiral catalyst is able to generate large amounts of enan-
tiomerically pure compounds that are vital for many applica-
tions. There is a wide range of synthetic catalysts based on a
diverse class of chiral organic ligands that can achieve excellent
levels of enantioselectivity for many different types of reac-
tions.[1] Bis(oxazoline)s are among this class of privileged li-
gands.[2,3] They are bidentate organic ligands with a N2 coordi-
nation sphere and are able to coordinate several types of transi-
tion metals (Figure 1). The copper(II) complexes are the most
used catalysts for several reactions such as the asymmetric cy-
Figure 1. Copper(II) complexes with bis(oxazoline) ligands; OTf = trifluoro-
methylsulfonyl.
clopropanation of alkenes, Diels–Alder cycloaddition, kinetic
resolution of 1,2-diols, and so on.[3–5]
However, these catalysts work in the homogeneous phase
and are difficult to separate at the end of a reaction, which
hinders their recycling. This constitutes a major drawback in
[a] CICECO – Aveiro Institute of Materials, Department of Chemistry, University
of Aveiro,
3810-193 Aveiro, Portugal
homogeneous asymmetric catalysis, especially given that most
of the chiral ligands used are costly. A number of strategies to
make this type of catalyst recyclable have been designed.[4,6]
The most popular is immobilization of the homogeneous cata-
lyst onto the surface of a porous material (e.g., zeolites, clays,
ordered mesoporous silicas, carbon materials).[4,6] Different
methodologies have been adopted as well, such as encapsula-
tion, incipient wet impregnation, electrostatic interactions, and
covalent attachment.[4,6–8] The use of simple and leaching-ef-
fective immobilization strategies is highly desirable to maintain
the reusability and economical feasibility of the heterogeneous
E-mail: ana.rosa.silva@ua.pt
www.ua.pt
[b] CICECO – Aveiro Institute of Materials, Department of Materials & Ceramic
Engineering, University of Aveiro,
3810-193 Aveiro, Portugal
E-mail: pcferreira@ua.pt
[c] Laboratorio de Microscopias Avanzadas, Instituto de Nanociencia de
Aragon, Universidad de Zaragoza,
50018 Zaragoza, Spain
[d] Instituto de Catálisis y Petroleoquímica, CSIC,
28049 Madrid, Spain
Supporting information and ORCID(s) from the author(s) for this article
are available on the WWW under http://dx.doi.org/10.1002/ejic.201501196.
Eur. J. Inorg. Chem. 2016, 413–421
413
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
catalyst. Because the majority of homogeneous catalysts are ICP-AES also showed that the copper(II) complexes were intro-
transition-metal complexes, they can potentially be directly co- duced at the surface of the amino-modified mesoporous orga-
ordinated onto the surface of a conveniently functionalized po- nosilicas. However, the amount of copper found was larger than
rous material.
the amount of bis(oxazoline) ligand found, as can be confirmed
in Table 1 by the Cu/Box ratio. Copper(II) complexes with bis(ox-
azoline) ligands are usually generated in situ by combining suit-
able amounts of the ligand and the copper(II) salt.[1,7,8] Herein,
the complexes were allowed to form in a ligand/copper ratio of
1:1 until the solution stabilized, as determined by UV/Vis spec-
troscopy. The amine-modified mesoporous organosilicas were
then added to the complexes. Thus, the lower amount of ligand
present at the surface of the amine-modified mesoporous orga-
nosilicas relative to that of copper may be due to decoordina-
tion of part of the bis(oxazoline) ligands during coordination of
the copper(II) complex onto the surface amine groups of the
materials. This is probably due to a lower formation constant,
as bis(oxazoline) ligands are bidentate and rigid.[16] It is also
curious to note that elemental analysis and ICP-AES revealed
that the amount of Me2PhBox ligand present on the NH2PhPMO
material is lower than the amount of PhBox ligand. This sug-
gests that the latter ligand has a higher formation constant with
copper(II) than the former ligand. Nevertheless, steric constrains
associated with the size of the ligand and the effect of curva-
ture of the support material may not be ruled out.
Periodic mesoporous organosilicas (PMOs),[9–11] prepared for
the first time in 1999, are interesting materials, because they
possess high specific surface areas and pore volumes, a narrow
pore-size distribution, and an alternating hydrophobic–hydro-
philic character. Moreover, PMOs are interesting materials, as
they may act as “channel ligands” by binding transition metals.
In 2002, the preparation of PMOs with aromatic bridges with
phenylene[12] (Ph) and biphenylene[13] (Bph) moieties was re-
ported, and the synthesis of these materials was remarkable
owing to the presence of crystal-like walls. Additionally, the aro-
maticity of these bridges allowed easy incorporation of amine
functionalities into the framework of these PMOs to give rise to
NH2PhPMO and NH2BphPMO through the use of simple electro-
philic substitution organic reactions.[14,15] The coordination of
copper complexes to the primer functional amine group in the
framework of these materials is a step to be explored. Herein,
we describe the direct coordination of copper(II) complexes
with bis(oxazoline) ligands onto amine-modified phenylene-
and biphenylene-bridged PMOs (i.e., NH2PhPMO and
NH2BphPMO). The materials were tested as heterogeneous cat-
alysts in the kinetic resolution of hydrobenzoin.
The X band electron paramagnetic resonance (EPR) spectra
of the copper(II) complexes immobilized on the amine-modified
organosilicas at 120 K are shown in Figure 2. The spectra are
typical of magnetically diluted copper(II) species. The hyperfine
structure from the coupling of the free electron with the
CuII nucleus (I = 3/2) can be clearly seen in the case of the
CuPhBox(OTf)2@NH2BphPMO material. The CuMe2PhBox-
(OTf)2@NH2PhPMO and CuPhCl2@NH2PhPMO materials possess
higher copper contents (Table 1) and broadening of the hyper-
fine structure is observed, which is not completely resolved in
the latter case. Nevertheless, the results of the simulation of the
spectra by using the WinEPR software are compiled in Table 2,
together with parameters from the free copper(II) salts and
complexes with related bis(oxazoline) ligands found in the liter-
ature.[1] The spectra are axial or almost axial with Az higher than
the corresponding free copper(II) salts, which is indicative of
the presence of nitrogen in the coordination sphere. Further-
more, the Az values are high in the case of the CuMe2PhBox-
(OTf)2@NH2PhPMO and CuPhBox(OTf)2@NH2BphPMO materials
and are typical of square-planar structures. The Az value for the
Results and Discussion
Characterization
Figure 1 depicts the copper(II) complexes with bis(oxazoline)
ligands immobilized in the periodic amino-functionalized
mesoporous phenylene- and biphenylene-bridged PMOs, con-
taining 1.82 and 2.61 mmol g–1 of amine groups, respectively
(Table 1),[14,15] to form CuMe2PhBox(OTf)2@NH2PhPMO,
CuPhCl2@NH2PhPMO, and CuPhBox(OTf)2@NH2BphPMO. The
materials were characterized by elemental analysis (C, N, and
H) and copper inductively coupled plasma atomic emission
spectroscopy (Cu ICP-AES); the results are compiled in Table 1.
From the increase in the nitrogen content, it can be concluded
that the copper(II) complexes (Figure 1) were introduced at the
surface of the amino-modified mesoporous organosilicas. Given
that each PhBox molecule possesses two nitrogen atoms, this
increase in nitrogen content corresponds to 57, 368, and
94 μmol g–1 of bis(oxazoline) ligand in CuMe2PhBox- CuPhCl2@NH2PhPMO material is lower, which is indicative of
(OTf)2@NH2PhPMO, CuPhCl2@NH2PhPMO, and CuPhBox- tetrahedral distortions. Nevertheless, the value is higher than
(OTf)2@NH2BphPMO, respectively. Analysis of these materials by that for CuCl2, which indicates the presence of nitrogen in the
Table 1. Elemental analysis and copper ICP-AES results.
Sample
C
[%]
H
[%]
N
[%]
Cu
[%]
N
Box
Cu
Cu/Box
[mmol g–1
]
[μmol g–1
]
[μmol g–1
]
NH2phPMO
34.22
34.40
36.85
48.39
44.16
2.66
2.85
3.31
3.40
3.74
2.55
2.71
3.58
3.66
3.92
1.82
1.94
2.56
2.61
2.80
CuMe2PhBox(OTf)2@NH2PhPMO
CuPhBoxCl2@NH2PhPMO
NH2BphPMO
CuPhBox(OTf)2@NH2BphPMO
Cu(OTf)2@NH2BphPMO
2.20
3.87
57
368
347
610
6.1
1.7
1.17
0.92
94
185
145
2.0
Eur. J. Inorg. Chem. 2016, 413–421
www.eurjic.org
414
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 2. EPR parameters obtained by simulation for the copper(II) complexes coordinated onto amine-modified mesoporous organosilicas at 120 K.
[b]
[b]
[b]
Sample
gx
gy
gz
Ax
Ay
Az
Ref.
[a]
[1]
[1]
[1]
[1]
Cu(OTf)2
CuCl2
2.083
2.061
2.064
2.057
2.083
2.090
2.078
2.080
2.083
2.061
2.073
2.057
2.083
2.090
2.078
2.080
2.412
2.316
2.313
2.280
2.300
2.316
2.270
2.300
4.4
18.3
5.0
4.4
18.3
4.8
134.6
152.7
169.0
132.0
182
162
185
170
[a]
[a]
CuMe2PhBox(OTf)2(H2O)2
CuMe2PhBoxCl2
CuMe2PhBox(OTf)2@NH2PhPMO
CuPhBoxCl2@NH2PhPMO
CuPhBox(OTf)2@NH2BphPMO
Cu(OTf)2@NH2BphPMO
[a]
11.0
11.0
[c]
[c]
this work
this work
this work
this work
[c]
[c]
[c]
[c]
[c]
[c]
[a] In frozen THF/CH2Cl2 solution at 140 K. [b] Values × 10–4 cm–1. [c] Not simulated owing to larger linewidths in the solid-state EPR spectra.
coordination sphere. It is also noteworthy that the Az value for shows higher values of gz and lower Az, and this indicates that
the CuMe2PhBox(OTf)2@NH2PhPMO material is higher than that the former has more nitrogen incorporated in its coordination
for the free complex in THF/CH2Cl2 solution (1:1). This may be sphere.
due to a change in the geometry of the coordination sphere
from distorted octahedral to square planar[1] and/or coordina-
tion to the nitrogen atom of the amine-modified mesoporous
NH2PhPMO. The gz values of the copper(II) complexes immobi-
lized on the amine-modified mesoporous organosilicas are simi-
lar and lower than the value for the corresponding copper(II)
salts, which is consistent with the incorporation of nitrogen into
their coordination spheres.
The FTIR attenuated total reflectance (ATR) spectra of the
parent amine-modified mesoporous organosilicas and the coor-
dinated copper(II) complexes are depicted in Figure 3. The
amine-modified NH2PhPMO and NH2BphPMO materials possess
a typical N–H stretching vibration at ν ≈ 3380 cm–1 and a N–H
˜
bending vibration at ν = 1625 and 1608 cm–1. The materials
˜
with the coordinated copper(II) complexes present spectra that
are more complex with additional bands: C–H stretching vibra-
tions at ν = 2915 and 2853 cm–1, C=N stretching vibrations
˜
˜
between ν =1654 and 1660 cm–1, a shoulder for the parent
material resulting from N–H bending vibration, and a S=O
stretching vibration at ν = 638 cm–1 for materials in which cop-
˜
per(II) triflate was used. Thus, the presence of these additional
bands from both the copper(II) salt and the bis(oxazoline) li-
gand further proves the presence of copper(II) complexes at the
surface of the amine-modified mesoporous organosilicas.
Figure 4 presents the powder X-ray diffraction (PXRD) data
of NH2PhPMO (Figure 4, a) and NH2BphPMO (Figure 4, b) before
and after copper coordination to evaluate the structural order
of the prepared materials. The PXRD patterns of NH2PhPMO and
NH2BphPMO present characteristic low-angle diffraction peaks
at d = 4.65 and 4.75 nm, respectively. Additionally, molecular-
scale periodicity along the wall is observed through the pres-
ence of medium-range reflections at
d = 0.76 nm for
NH2PhPMO and d = 1.20 nm for NH2BphPMO. Thus, both pris-
tine materials possess meso- and molecular-scale periodicities.
Upon coordination of copper to the amine functionalities of
NH2PhPMO, the 100 reflection is maintained at d = 4.65 nm,
whereas the coordination of CuPhBox(OTf)2 to NH2BphPMO
leads to a slight increase from d = 4.75 to 4.80 nm. Both
CuMe2PhBox(OTf)2@NH2PhPMO and CuPhboxCl2@NH2PhPMO
presents peaks in the medium-scattering angle range at d =
0.76 nm, whereas CuPhBox(OTf)2@NH2BphPMO processes
peaks related to the molecular scale at d = 1.18 nm.
Figure 2. X-band EPR spectra of the copper(II) complexes immobilized on the
amine-modified organosilicas at 120 K.
The amine-modified materials are also able to coordinate
copper(II) salts directly. As an example, the NH2BphPMO mate-
rial was able to coordinate 145 μmol g–1 of copper(II) triflate
(Table 1) from a methanol solution containing 1.15 mmol of
copper(II) salt per gram of material. The EPR spectrum of the
resulting Cu(OTf)2@NH2BphPMO material is depicted in Fig-
ure S1 (Supporting Information) and the corresponding param-
The mesoporosity of the prepared materials is also evi-
eters are shown in Table 2. The spectrum is also typical of a denced by nitrogen adsorption–desorption isotherms acquired
copper(II) magnetically diluted species with lower gz and higher at –196 °C (Figure S2) through the observation of type IV iso-
Az values than the free copper(II) triflate (Table 2). This is con- therms (IUPAC classification) characteristic of mesoporous mate-
sistent with the incorporation of nitrogen into the copper(II) rials. A reduction in the values of SBET, Vp, and dp upon coordina-
coordination sphere. Relative to that shown by the tion of the copper ligand in the amine functionalities of
CuPhBox(OTf)2@NH2BphPMO material, Cu(OTf)2@NH2BphPMO NH2PhPMO and NH2BphPMO is observed (Figure S2 and
Eur. J. Inorg. Chem. 2016, 413–421
www.eurjic.org
415
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 3. (a) FTIR ATR spectra of the parent amine-modified mesoporous organosilicas and with the coordinated copper(II) complexes; (b) expansions of the
600–1000, 1250–1850, and 2500–4000 cm–1 regions.
To corroborate the presence of copper in the materials and
to assure the mesopore nature after Cu incorporation, trans-
mission electron microscopy (TEM) observation and chemical
analyses by energy-dispersive X-ray spectroscopy (EDS)
were performed. Figure
5 (a) depicts the structure of
CuMe2PhBox(OTf)2@NH2PhPMO along the [001] orientation,
which presents the hexagonal pore array, and the correspond-
ing fast-Fourier transform (FFT) in the inset, which was indexed
by assuming a p6mm symmetry. A magnified TEM image per-
pendicular to the pores is also presented in the inset, which
reveals the molecular-scale order. The corresponding EDS analy-
sis is shown and confirms the Si, O, and C composition of the
mesoporous material and, in addition, the presence of Cu an-
chored in the walls and uniformly distributed (inset in the EDS
graphic). Figure 5 (b) exhibits a similar analysis acquired for
CuPhBoxCl2@NH2PhPMO; first, a low-magnification image of the
material is presented with the FFT (inset), which corresponds to
the particle marked by a white dashed rectangle orientated
along [001]. A micrograph perpendicular to the pores clearly
reveals the good crystallinity of the sample with an interatomic
distance of 0.77 nm, which corresponds to the molecular perio-
dicity of the walls. The FFT, shown in the inset, presents two
Figure 4. PXRD patterns of pristine and metal-functionalized NH2PhPMO and
NH2BphPMO.
Table 3). The maxima in the pore-size distribution (PSD) curves
(Figure S2 and Table 3) shift from 3.5 to 3.1 nm for NH2PhPMO
and from 3.0 to 2.5 nm for NH2BphPMO upon coordination of
CuPhBoxCl2 and CuPhBox(OTf)2 to the amine functionalities of types of spots named “a” that can be attributed to the meso-
the PMOs.
scale and “b” at a larger distance to the central spot that corre-
Eur. J. Inorg. Chem. 2016, 413–421
www.eurjic.org
416
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 3. Physical properties of the prepared materials.
[a]
[c]
[d]
[e]
Sample
d100
a[b]
[nm]
SBET
VP
dP
b[f]
[nm]
[nm]
[m2 g–1
]
[cm3 g–1
]
[nm]
NH2PhPMO
4.65
4.65
4.65
4.75
4.80
5.36
5.36
5.36
5.48
5.54
631
436
313
533
522
0.84
0.55
0.45
0.54
0.44
3.5
3.2
3.1
3.0
2.5
1.82
2.19
2.25
2.46
3.00
CuMe2PhBox(OTf)2@NH2PhPMO
CuPhBoxCl2@NH2PhPMO
NH2BphPMO
CuPhBox(OTf)2@NH2BphPMO
[a] Spacing of the 100 reflection. [b] Unit cell parameter calculated as (2d100/√3). [c] BET surface area. [d] Pore volume. [e] Pore width obtained from the BJH
method with the corrected Kelvin equation, that is, KJS–BJH method at the maximum of pore-size distribution calculated on the basis of adsorption data.
[f] Pore wall thickness calculated as (2d100/√3 – dP), in which the first term is the unit cell parameter.
Figure 5. TEM images of (a) CuMe2PhBox(OTf)2@NH2PhPMO, (b) CuPhBoxCl2@NH2PhPMO, and (c) CuPhBox(OTf)2@NH2BphPMO. The EDS and inset figures
show the presence of Cu and the uniform distribution along the particles.
sponds to the molecular scale. The chemical analysis also cor-
Scanning transmission electron microscopy (STEM) coupled
roborates the presence of Cu incorporated in the material and with a high angle annular dark field detector (HAADF) was em-
the homogeneous distribution.
ployed to characterize CuPhBox(OTf)2@NH2BphPMO (Figure 5,
Eur. J. Inorg. Chem. 2016, 413–421
www.eurjic.org
417
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
c). Although clear order is not observed in this sample, both a of free silanols during the functionalization reactions, which re-
mesostructure and a molecular structure can be identified. In duces the intensities of the T1 and T2 signals.
this case, the interatomic distances are 1.17 nm. EDS analysis
corroborates the presence of copper in the material.
Thermogravimetric analysis (TGA) of NH2PhPMO and
NH2BphPMO before and after copper coordination (Figure S5)
revealed a first weight loss below 100 °C owing to desorption
of physisorbed water. The NH2PhPMO material has thermal sta-
bility up to 360 °C. Above this value, decomposition of the
amine groups takes place, and above 500 °C, decomposition of
the phenylene organic moieties occurs. The thermal stability
decreases upon introduction of the CuMe2PhBox(OTf)2 and
CuPhBoxCl2 copper ligands to NH2PhPMO. For both materials,
CuMe2PhBox(OTf)2@NH2PhPMO and CuPhBoxCl2@NH2PhPMO,
decomposition of the bis(oxazoline) ligands starts at 135 °C,
which is followed by degradation of the triflate anion, in the
case of CuMe2PhBox(OTf)2@NH2PhPMO. Finally, the decomposi-
tion of the NH2PhPMO material starts above 300 °C for
CuMe2PhBox(OTf)2@NH2PhPMO and above 400 °C for CuPh-
BoxCl2@NH2PhPMO. TGA of NH2BphPMO revealed a weight loss
above 225 °C, which corresponds to cleavage of the amine
bond. Decomposition of this PMO materials starts above 450 °C.
Coordination of CuPhBox(OTf)2 to NH2BphPMO leads to a
weight loss up to 200 °C, which corresponds to decomposition
of the bis(oxazoline) ligands followed by degradation of the
triflate anion and decomposition of the PMO material above
300 °C. Note that this temperature degradation is inferior to
that achieved for the Cu(OTf)2@NH2BphPMO material (above
350 °C), and this is due to the presence of the organic bis(ox-
azoline) ligand.
The solid-state 13C cross polarization magic angle spinning
(CP MAS) NMR spectra of the NH2PhPMO and NH2BphPMO ma-
terials before and after copper coordination are shown in Fig-
ure S3. NH2PhPMO shows resonances at δ= 120.6, 123.5, 133.5,
and 150.3 ppm assigned to the sp2 carbon atoms of the amine-
functionalized phenylene bridge of the PMO. Note that the res-
onance at δ= 150.3 ppm corresponds to the carbon atom of
the phenylene group linked to the amine functionality.[14,17] Co-
ordination of the copper ligands to the amine group of
NH2PhPMO leads to enormous differences in the 13C NMR spec-
tra. As CuII has a paramagnetic nucleus, the carbon nuclei close
to it are not visible in the 13C CP MAS NMR spectra. In this way,
the carbon signals related to both the bis(oxazoline) ligands
and the amine-functionalized phenylene group are absent. The
13C CP MAS NMR spectra of CuMe2PhBox(OTf)2@NH2PhPMO
and CuPhBoxCl2@NH2PhPMO only display a small amount of
noncoordinated carbon atoms. For example, the signal at δ=
150.2 ppm almost disappears in both materials, which corro-
borates the coordination of the copper ligands to the amine
phenylene bridge of the PMO material. The intense resonance
at δ= 133.5 ppm also corresponds to the sp2 carbon atoms of
the phenylene moieties from the 1,4-bis(triethoxylsilyl)benzene
precursor that are not aminated during the amination proc-
ess.[14,17] Other small resonances in the spectrum between δ=
10 and 75 ppm may correspond to residual carbon atoms of
the alkyl chain of the surfactant. The 13C CP MAS NMR spectrum
of NH2BphPMO presents five resonances at δ= 126.1, 130.4,
135.2, 140.0, and 147.7 ppm that are assigned to biphenylene
moieties functionalized with amine groups. The signal at δ=
147.7 ppm is assigned to the sp2 carbon atom directly linked
to the NH2 group.[15] Once more, this signal disappears upon
coordination of CuPhBox(OTf)2 to the amine group of
NH2BphPMO, as expected. The presence of resonances at δ=
130.0, 130.2, 135.2, and 140.5 ppm are assigned to the biphen-
ylene moieties that are not aminated during the amination re-
action.
Figure S4 shows the 29Si MAS and CP MAS NMR spectra of
NH2PhPMO and CuMe2PhBox(OTf)2@NH2PhPMO. The parent
NH2PhPMO exhibits signals at ca. δ= –59, –71, and –80 ppm
assigned to T1, T2, and T3 [Tm = RSi(OSi)m(OH)3–m] organosili-
ceous species, respectively, as described in the literature.[14,17]
Coordination of CuMe2PhBox(OTf)2 to the NH2PhPMO material
leads to differences in the 29Si MAS and CP MAS NMR spectra.
The T1 signal almost disappears and the intensity of the T2 reso-
nance suffers a decrease with respect to the intensity of the T3
sites. As mentioned before, CuII is a paramagnetic nucleus and
its presence may affect the signals of the diamagnetic nuclei
just by being in the immediate vicinity. In this way, the intensi-
ties of all the signals in the spectra are reduced, as only the
silicon atoms away from the CuII center give rise to a signal.
This is supported by a higher level of noise specific to the CP/
Catalytic Experiments
The amine-modified mesoporous organosilicas with the coordi-
nated copper(II) complexes were tested as heterogeneous cata-
lysts in the kinetic resolution of hydrobenzoin. The results of
the catalytic experiments and the recycling experiments of the
catalysts are compiled in Table 4. Taking into consideration that
in kinetic resolution the maximum product yield that can be
obtained is 50 %, it can be seen that all the materials are active,
selective, and enantioselective in the monobenzoylation of
hydrobenzoin (Figure 6). The material containing the Me2PhBox
ligand, the best ligand in the homogeneous phase (Table 4),
presented the highest enantioselectivity of all the materials in
the first cycle (73 %). However, upon reuse, the enantioselectiv-
ity dropped drastically, which indicates that there was consecu-
tive leaching of the ligand into the solution. This material con-
tains the lowest amount of ligand, and thus the constant for
the formation of the complex between copper(II) and the
bis(oxazoline) ligand must play an important role in the stability
of the complex,[16] especially if coordinated to the amine
groups of the material. On the other hand, the materials con-
taining the PhBox ligand presented lower enantioselectivities,
but the catalysts were robust upon recycling over five cycles.
This indicates that the CuPhBox complex with either chloride
or triflate counteranions is more stable even if further coordi-
nated to amine groups. The material containing the highest
MAS spectra after functionalization of the amine groups of the amount of ligand and copper, that is, CuPhBoxCl2@NH2PhPMO,
materials. Moreover, it is possible to enhance the condensation is more active and enantioselective than CuPhBox-
Eur. J. Inorg. Chem. 2016, 413–421
www.eurjic.org
418
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 4. Kinetic resolution of hydrobenzoin by the copper(II) complexes with bis(oxazoline) ligands coordinated to the amine-modified mesoporous organosili-
cas.[a]
Sample
Time [h] Run
Cu[b] [mol-%]
5.5
Ligand[b] [mol-%]
Yield[c] [%]
ee[d] [%]
S[e]
TON[f]
7
CuCl2·2H2O
without catalyst
CuCl2 + (S,S)-PhBox
2
24
2
1
1
38
21
42
46
47
42
40
43
41
39
42
40
35
32
30
31
29
27
0.1
0
1
5.4
1.0
1.0
5.2
4.6
4.4
3.9
3.5
2.9
2.8
2.6
1.5
1.5
1.4
1.3
1.3
4.8
1.0
1.0
2.1
1.8
1.7
1.5
1.4
0.5
0.5
0.4
0.8
0.7
0.7
0.7
0.7
79
85
99
57
57
53
44
47
73
12
2
43
47
43
42
44
15
27
581
5
5
5
3
4
11
1
1
3.0
3.4
3.0
2.9
3.0
8
[Cu(OTf)2] + (S,S)-PhBox
[Cu(OTf)2] + (S,S)-Me2PhBox
CuPhBoxCl2@NH2PhPMO
CuPhBoxCl2@NH2PhPMO
CuPhBoxCl2@NH2PhPMO
CuPhBoxCl2@NH2PhPMO
CuPhBoxCl2@NH2PhPMO
CuMe2PhBox(OTf)2@NH2PhPMO
CuMe2PhBox(OTf)2@NH2PhPMO
CuMe2PhBox(OTf)2@NH2PhPMO
CuPhBox(OTf)2@NH2BphPMO
CuPhBox(OTf)2@NH2BphPMO
CuPhBox(OTf)2@NH2BphPMO
CuPhBox(OTf)2@NH2BphPMO
CuPhBox(OTf)2@NH2BphPMO
24
24
24
24
24
24
24
46
47
8
1
2
3
4
5
1
2
3
1
2
3
4
5
9
10
11
11
15
15
14
21
20
22
22
21
24
24
24
24
24
[a] Reactions were performed for 24 h at 0 °C by using (R,R)-hydrobenzoin (0.48 mmol), (S,S)-hydrobenzoin (0.48 mmol), DIPEA (1.00 mmol), Cu (1.0 mol-%),
and benzoyl chloride (0.50 mmol) in of CH2Cl2 (5.0 mL). [b] Percentage of copper and ligand in the catalyst in relation to hydrobenzoin (see Table 1); for the
recycling experiments, corrected for the weight of heterogeneous catalyst loss. [c] Yield of the isolated monobenzoylated product (Figure 6). [d] Enantiomeric
excess of the monobenzoylated product, determined by HPLC. [e] Selectivity (S) = ln[1 – yield (1 + ee)]/ln[1 – yield (1 – ee)]. [f] Turnover number (TON) = mol
of isolated monobenzoylated product/mol of Cu.
(OTf)2@NH2BphPMO. However, after the third cycle, the enanti- Conclusions
oselectivity of the former catalyst dropped, whereas the latter
Three copper(II) complexes with commercial bis(oxazoline) li-
one remained the same. Relative to the enantioselectivities ob-
gands were immobilized onto amine-modified mesoporous or-
tained for the homogeneous phase reaction with the free com-
ganosilicas for the first time by direct coordination onto the
plexes, the enantioselectivities obtained for the heterogeneous
amino groups. Partial decoordination of the bis(oxazoline) li-
phase were always lower. As the amine-modified mesoporous
gands occurred upon immobilization onto the periodic meso-
organosilicas with the coordinated copper(II) complexes pos-
porous organosilica surface amino groups, which resulted in
sess higher amounts of copper than the bis(oxazoline) ligand,
lower amounts of ligand than copper on the materials. The ma-
the decrease in the enantioselectivity can be partially explained
terials were active, selective, and enantioselective heterogene-
by racemization through non-enantioselective monobenzoyla-
ous catalysts for the kinetic resolution of hydrobenzoin. In two
tion by the copper(II) centers not coordinated to the bis(ox-
cases, the materials were reused over five cycles without any
azoline) ligand. Nevertheless, in comparison to control experi-
significant loss of catalytic activity or enantioselectivity. The en-
ments run with the copper(II) salt and without catalyst, the re-
antioselectivities were, however, lower than those in the homo-
actions present enantioselectivity supporting the presence of
geneous phase owing to partial racemization of the product by
bis(oxazoline) ligands in the coordination sphere of the coordi-
non-enantioselective pathways. The bis(oxazoline) ligand and
nated copper(II) complexes.
the corresponding constant of formation for the copper(II) com-
plex seem to play important roles in the immobilization yields
of the copper(II) bis(oxazoline) complexes (Figure 1) and in the
robustness of the heterogeneous catalysts upon reuse. Steric
constrains related to the size of the ligands (Me2PhBox larger
than PhBox) and the curvature typical for mesoporous materials
Figure 6. Kinetic resolution of hydrobenzoin; DIPEA = N,N-diisopropylethylam-
ine.
cannot be ruled out.
With the use of a more complex immobilization strategy by
Experimental Section
functionalization of the PhBox bis(oxazoline) bridge and by
Materials: Copper(II) trifluoromethanesulfonate {copper(II) triflate,
grafting onto the surface ordered mesoporous silicas and their
[Cu(OTf)2], 98 %}, 2,2′-methylenebis[(4S)-4-phenyl-2-oxazoline]
carbon replicas, more enantioselective heterogeneous catalysts
(PhBox, 97 %), (S)-(–)-2,2′-isopropylidene-bis(4-phenyl-2-oxazoline)
(Me2PhBox, 97 %), dry tetrahydrofuran (THF, ≥99.9 %), (R,R)-1,2-di-
phenyl-1,2-ethanediol (99 %), (S,S)-1,2-diphenyl-1,2-ethanediol
(99 %), N,N′-diisopropylethylamine (DIPEA, ≥99 %), and potassium
bromide (FTIR grade, ≥99 %) were purchased from Aldrich and were
could be obtained in the kinetic resolution of hydrobenzoin.[18]
The amount of copper(II) coordinated to the material was con-
siderably lower than the amount of bis(oxazoline), which must
be essential to attain high enantioselectivities.
Eur. J. Inorg. Chem. 2016, 413–421
www.eurjic.org
419
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
used as received. Ethanol (p.a.) and methanol (p.a.) were from
Riedel-de Haën. Dichloromethane was purchased from Romil and
was HPLC grade.
ethanol. A few drops of the suspension were then placed onto
holey carbon nickel microgrids. Transmission electron microscopy
(TEM) was performed with a Tecnai F30 operated at 300 kV. The
microscope, which allowed a spatial resolution of 0.19 nm, was
equipped with an EDS (EDAX) detector for chemical analysis, a Ga-
tan bottom-entry CCD 2K × 2K camera, STEM module, and a Gatan
Tridiem energy filter for spectroscopic measurements.
Immobilization of the Copper(II) Complexes onto the Amine-
Modified Mesoporous Organosilicas
CuMe2PhBox(OTf)2@NH2PhPMO: A methanol solution (40 mL)
containing Me2PhBox (0.162 mmol) and copper(II) trifluorometh-
anesulfonate (0.143 mmol) was stirred for 2 h (Figure 1). Then,
NH2PhPMO (0.1730 g, synthetized according to the literature[14]),
previously dried at 120 °C, was added, and the mixture was heated
at reflux for 48 h. The material was filtered, washed, and heated at
reflux for 3 h in fresh methanol (40 mL) to remove physisorbed
copper(II) complexes. Finally, the material was filtered and dried at
120 °C for 6 h. The material acquired a green color.
Catalysis Experiments: All the catalytic reactions of the prepared
materials were performed in batch reactors at atmospheric pressure
and with constant stirring. Kinetic resolution of 1,2-diphenylethane-
1,2-diol (Figure 6) was performed at 0 °C by using (R,R)-1,2-diphen-
ylethane-1,2-diol (0.48 mmol), (S,S)-1,2-diphenylethane-1,2-diol
(0.48 mmol), DIPEA (170 μL, 1.00 mmol), heterogeneous catalyst
(1.0 mol-% Cu), and benzoyl chloride (58 μL, 0.50 mmol) in CH2Cl2
(5.00 mL).[5,18] The mixture was stirred for 24 h, and after filtration
of the heterogeneous catalyst, the solvent was evaporated from the
filtrate and the monobenzoylated product (Figure 6) was isolated
by column chromatography (silica gel, n-hexane/ethyl acetate, 3:1).
The yields were calculated on the basis of the 1,2-diol.[5,18] The en-
antiomeric excess of the monobenzoylated product was deter-
mined by HPLC at λ= 254 nm by using a Chiralcel OD column
(250 mm × 4.6 ID, 5 μm) and n-hexane/2-propanol (9:1) as the elu-
ent. The retention times of the enantiomers of the (R)-mono-
benzoylated and (S)-monobenzoylated products were identified by
comparison with those of a racemic monobenzoylated product.[5,18]
The reaction selectivity (S) was calculated on the basis of the yield
of the isolated monobenzoylated product and the respective enan-
tiomeric excess by using the formula: ln[1 – yield (1 + ee)]/ln[1 –
yield (1 – ee)]. The isolated materials at the end of the reactions
were washed extensively with the appropriate solvent, dried under
vacuum, and reused in another cycle under the same experimental
conditions. Control experiments were also performed by using the
same experimental procedure in the homogeneous phase with
equimolar quantities of [Cu(OTf)2] and either PhBox or Me2PhBox
to compare with the heterogeneous ones.
CuPhBoxCl2@NH2PhPMO: A methanol solution (40 mL) containing
PhBox (0.210 mmol) and copper(II) chloride (0.215 mmol) was
stirred for 18 h (Figure 1). Then, NH2PhPMO (0.2322 g, prepared
according to the literature[14]), previously dried at 120 °C, was
added, and the mixture was heated at reflux for 72 h. The material
was filtered, washed, and heated at reflux in fresh methanol (40 mL)
for 3 h to remove physisorbed copper(II) complexes. Finally, the
material was filtered and dried at 120 °C for 6 h. The material ac-
quired a green color.
CuPhBox(OTf)2@NH2BphPMO: A methanol solution (40 mL) con-
taining PhBox (0.203 mmol) and copper(II) trifluoromethane-
sulfonate (0.209 mmol) was stirred for 5 h (Figure 1). Then,
NH2BphPMO (0.1730 g, prepared according to the literature[15]) was
added, and the mixture was heated at reflux for 48 h. The material
was filtered, washed, and heated at reflux in fresh methanol (40 mL)
for 6 h to remove physisorbed copper(II) complexes. Finally, the
material was filtered and dried at 120 °C for 6 h. The material ac-
quired a green color.
Physicochemical Characterization: Elemental analysis (C, H, and
N) was done by using a TruSpec 630–200–200 CNHS Analyzer at
University of Aveiro. ICP-AES analysis was performed at the Univer-
sity of Vigo (Spain), and EPR analysis at 120 K was performed at
University of Santiago de Compostela (Spain). FTIR spectroscopy
was performed with a FTIR Bruker Tensor 27 instrument with a
Golden Gate ATR (Attenuated Total Reflectance). PMO powders
were dehydrated at 110 °C overnight before FTIR analysis. Powder
X-ray diffraction (PXRD) data were recorder with a Rigaku Geigerflex
D Max-C Series diffractometer by using Cu-Kα radiation. Samples
were step scanned in 0.02° 2θ steps with a counting time of 20 s
per step. Nitrogen adsorption–desorption isotherms were collected
at –196 °C by using a Gemini V 2.00 instrument model 2380. The
PMO materials were dehydrated overnight at 120 °C to an ultimate
pressure of 1024 mbar and then cooled to room temperature prece-
ding the adsorption. 13C NMR, 29Si NMR, and 15N NMR spectra were
acquired by using a double resonance MAS probe on a Bruker Av-
ance III 400 spectrometer operating at 9.4 T. 13C CP MAS NMR spec-
tra were acquired by employing a 4 μs 1H 90° pulse; a contact time
of 3 ms; a spinning rate of 8, 9, 10, or 12 kHz; and a recycle delay
(RD) of 5 s. 29Si MAS NMR spectra were gathered by using 40° flip
Supporting Information (see footnote on the first page of this
article): HPLC chromatograms of racemic and enantioenriched 2-
hydroxy-1,2-diphenylethyl benzoate, 29Si MAS and 29Si CP MAS NMR
spectra, nitrogen adsorption–desorption isotherms collected at
–196 °C, STEM-HAADF images with EDS analysis, XRD, and thermo-
gravimetry.
Acknowledgments
This work was developed within the scope of the project CI-
CECO-Aveiro Institute of Materials, POCI-01-0145-FEDER-007679
(FCT Ref. UID/CTM/50011/2013), financed by national funds
through the Fundação para a Ciência e a Tecnologia (FCT)/Min-
istério da Educação e Ciência (MEC) and, when appropriate, co-
financed by FEDER under the PT2020 Partnership Agreement.
A. R. S. and P. F. acknowledge support by the FCT (IF/01300/
2012 and IF/00327/2013, respectively). M. A. O. L. is thankful to
the FCT for a PhD grant (SFRH/BD/80883/2011). Research lead-
angle pulses, a spinning rate of 5.0 kHz, and a RD of 60 s. 29Si CP ing to these results received funding from the European Union
1
MAS NMR spectra were performed with 4 μs H 90° pulses, a CT of
8 ms, a spinning rate of 5 kHz, and a RD of 5 s. Tetramethylsilane
was used as reference for the chemical shifts in the 13C NMR and
29Si NMR spectra. Thermogravimetric analysis (TGA) was performed
with a Shimadzu TGA-50 instrument with a program rate of
5 °C min–1 in air. For electron microscopy observations, the samples
were crushed by using a mortar and pestle and were dispersed in
Seventh Framework Programme under grant agreement num-
ber 312483 - ESTEEM2 (Integrated Infrastructure Initiative-I3).
Keywords: Heterogeneous catalysis · Asymmetric catalysis ·
Periodic mesoporous organosilicas · Immobilization
Eur. J. Inorg. Chem. 2016, 413–421
www.eurjic.org
420
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[1] M. E. Owen, E. Carter, G. J. Hutchings, B. D. Ward, D. M. Murphy, Dalton
Trans. 2012, 41, 11085–11092.
[11] B. Melde, B. Holland, C. Blanford, A. Stein, Chem. Mater. 1999, 11, 3302–
3308.
[2] T. P. Yoon, E. N. Jacobsen, Science 2002, 298, 1904–1905.
[3] G. Desimoni, G. Faita, K. A. Jørgensen, Chem. Rev. 2011, 111, PR284–
PR437.
[12] S. Inagaki, S. Guan, T. Ohsuna, O. Terasaki, Nature 2002, 416, 304–307.
[13] M. P. Kapoor, Q. Yang, S. Inagaki, J. Am. Chem. Soc. 2002, 124, 15176–
15177.
[14] M. Ohashi, M. P. Kapoor, S. Inagaki, Chem. Commun. 2008, 7, 841–843.
[15] M. A. O. Lourenço, A. Mayoral, I. Díaz, A. R. Silva, P. Ferreira, Microporous
Mesoporous Mater. 2015, 217, 167–172.
[4] D. Rechavi, M. Lemaire, Chem. Rev. 2002, 102, 3467–3493.
[5] Y. Matsumura, T. Maki, S. Murakami, O. Onomura, J. Am. Chem. Soc. 2003,
125, 2052–2053.
[16] J. M. Fraile, J. I. Garcia, C. I. Herrerias, J. A. Mayoral, O. Reiser, A. Socuella-
mos, H. Werner, Chem. Eur. J. 2004, 10, 2997–3005.
[17] M. Lourenço, R. Siegel, L. Mafra, P. Ferreira, Dalton Trans. 2013, 42, 5631–
5634.
[18] A. R. Silva, L. Carneiro, A. P. Carvalho, J. Pires, Catal. Sci. Technol. 2013, 3,
2415–2424.
[6] A. F. Trindade, P. M. P. Gois, C. A. M. Afonso, Chem. Rev. 2009, 109, 418–
514.
[7] J. M. Fraile, J. I. Garcia, J. A. Mayoral, Chem. Rev. 2009, 109, 360–417.
[8] J. M. Fraile, J. I. Garcia, C. I. Herrerias, J. A. Mayoral, E. Pires, Chem. Soc.
Rev. 2009, 38, 695–706.
[9] T. Asefa, M. J. MacLachlan, N. Coombs, G. A. Ozin, Nature 1999, 402, 867–
871.
[10] S. Inagaki, S. Guan, Y. Fukushima, T. Ohsuna, O. Terasaki, J. Am. Chem.
Soc. 1999, 121, 9611–9614.
Received: October 17, 2015
Published Online: December 23, 2015
Eur. J. Inorg. Chem. 2016, 413–421
www.eurjic.org
421
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim