Highly enantioselective olefin epoxidation controlled by helical confined environments

Article abstract of DOI:10.1016/j.jcat.2013.08.032

Helical mesoporous materials of the MCM-41 type are important materials that can be prepared by onepot synthesis procedures with a co-surfactant. A control of the characteristics at a local level is of the most important in the view of the applications of such materials. However, there are not many studies relating such features with synthetic approaches. In this work, we prepared both helical and regular channel materials from Si-based MCM-41 type. Afterward, a bpy derivative was used as ligand to coordinate Mo^II/VI. The complexes and the new materials were tested as the catalytic precursors in the epoxidation of cis-cyclooctene, styrene, 1-octene, R-(+)-limonene and trans-hex-2-en-1-ol, using tert-butylhydroperoxide (TBHP) as oxidant. Although almost all the catalysts were 100% selective toward the epoxide, the conversions were in general good. The major achievement of these catalysts is an outstanding stereocontrol of the reaction products. In addition, these catalysts were found to be very effective under several circumstances. This is certainly an important contribution for such concept and may render such materials further applications where chiral recognition is important.

Full text of DOI:10.1016/j.jcat.2013.08.032

Journal of Catalysis 309 (2014) 21–32
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Highly enantioselective olefin epoxidation controlled by helical confined
environments
a,
Cristina I. Fernandes a, Marta S. Saraiva a, Teresa G. Nunes b, Pedro D. Vaz a, , Carla D. Nunes
⇑
⇑
a CQB, Departamento de Química e Bioquímica, Faculdade de Ciências da, Universidade de Lisboa, 1749-016 Lisboa, Portugal
b Centro de Química Estrutural, Universidade de Lisboa, Instituto Superior Técnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
a r t i c l e i n f o
a b s t r a c t
Article history:
Helical mesoporous materials of the MCM-41 type are important materials that can be prepared by one-
pot synthesis procedures with a co-surfactant. A control of the characteristics at a local level is of the
most important in the view of the applications of such materials. However, there are not many studies
relating such features with synthetic approaches. In this work, we prepared both helical and regular
channel materials from Si-based MCM-41 type. Afterward, a bpy derivative was used as ligand to
coordinate MoII/VI. The complexes and the new materials were tested as the catalytic precursors in the
epoxidation of cis-cyclooctene, styrene, 1-octene, R-(+)-limonene and trans-hex-2-en-1-ol, using tert-
butylhydroperoxide (TBHP) as oxidant. Although almost all the catalysts were 100% selective toward
the epoxide, the conversions were in general good. The major achievement of these catalysts is an
outstanding stereocontrol of the reaction products. In addition, these catalysts were found to be very
effective under several circumstances. This is certainly an important contribution for such concept and
may render such materials further applications where chiral recognition is important.
Received 18 July 2013
Revised 23 August 2013
Accepted 27 August 2013
Keywords:
Asymmetric catalysis
Chiral recognition
Helical materials
Molybdenum
Olefin epoxidation
Stereoselectivity
Mesoporous materials
Oxidation catalysis
Ó 2013 Elsevier Inc. All rights reserved.
1. Introduction
a widely employed method based on the use of poly-amino acids,
which was reported some time ago [14,15]. Although it was devel-
Olefin epoxidation is very important in the preparation of rele-
vant building blocks for organic synthesis or polymer science [1–
3]. Among several processes developed for asymmetric oxidation
of olefins, one of the earliest was that by Sharpless using a TiIV cat-
alyst in the presence of tartrate, especially effective with allylic
alcohols [4,5]. The Jacobsen–Katsuki epoxidation was later re-
ported relying on the use of a Mn-salen catalyst [6,7]. Both meth-
ods were important milestones in olefin epoxidation although they
present some drawbacks such as being less reactive and selective
toward Z-olefins (Sharpless) or require low (ca. 273 K) reaction
temperatures (Jacobsen–Katsuki) [8]. Shibasaki reported asymmet-
ric epoxidation reactions based on a La(BINOL) catalyst [9]. This
reaction shows an improvement over the previous ones as it can
proceed under mild conditions (r.t.), and it also works reasonably
well with Z-olefins [10,11]. A number of metal-free alkene oxida-
tion methods have also been reported. The Shi epoxidation reac-
tion relies on the use of a fructose-based chiral dioxirane to
conduct the transformation [12,13]. While it shows to be active
on a wide scope of substrates, as a shortcoming, it requires high
catalyst loadings to achieve good yields. The Juliá–Colonna is also
oped over a relatively narrow range of substrates initially, it has
suffered major improvements in recent times [10,11].
Almost all effective ligands reported for homogeneous catalytic
reactions share one prominent common feature – a rigid and bulky
structure that is vital to create effective asymmetric microenviron-
ments around the catalytically active site. By the same token of
confinement effects, solid surfaces have been used with success
to promote heterogeneous asymmetric catalysis, thus preserving
or improving the high enantioselectivity of product ee values com-
pared to their homogeneous analogues [16–23]. More recently,
chiral helical polymers, reported by Suginome and coworkers,
were used in asymmetric reactions with stereocontrol induced
by the helical backbone [24–27].
To this respect, the preparation of nanostructured materials
with helical morphology opened a new field of research due to
their potential applications such as chiral selective separation
and recognition and enantioselective catalysis [28–32]. Still, a
drawback is the lack of active sites of mesoporous materials, which
does not exclude the helical silica-based materials as well. To cir-
cumvent this, flaw different strategies are available to introduce
such active sites, ranging from co-condensation or post-synthesis
procedures to matrix doping for introduction of such active species
[33,34].
⇑
Corresponding authors. Fax: +351 217 500088.
E-mail addresses: pmvaz@fc.ul.pt (P.D. Vaz), cmnunes@fc.ul.pt (C.D. Nunes).
0021-9517/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcat.2013.08.032

22
C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
As part of our ongoing research toward more efficient and selec-
Prior to the grafting experiments, physisorbed water was re-
moved from the materials by heating at 453 K in vacuum
(10ꢀ2 Pa) for 2 h.
tive systems based on single-site active species which allow a fast
inflow of reactants and outflow of products [35], we report for the
first time – to the best of our knowledge – the incorporation of
MoII/VI active species inside pores of left-handed helical mesostruc-
tured silica materials [32]. The Mo-derivatized helical materials led
to highly active heterogeneous catalysts for asymmetric oxidation
of olefins and were benchmarked with similar Mo homogeneous
complexes.
Molybdenum catalysts are known for being very effective cata-
lysts in olefin epoxidation reactions [36–43]. Although many efforts
have been addressed concerning development of asymmetric reac-
tions, only a few were successful [39–43]. Some authors have inclu-
sively mentioned some reasons for this [40,42].
Materials MSh-bpy and MS-bpy resulted from the addition of a
suspension of the (ClCO)2bpy ligand (0.281 g; 1 mmol) in acetoni-
trile (CH3CN) to either a MSh or MS suspension (1 g) in acetonitrile
(CH3CN), and the mixture was heated at 358 K for 14 h. The
resulting solid was filtered off and washed twice with dichloro-
methane (CH2Cl2) and then dried in vacuum at 323 K for 2 h.
MSh-bpy-MoII and MSh-bpy-MoVI materials were obtained by adding
a solution of [MoI2(CH3CN)2(CO)3] (0.300 g, 0.56 mmol) or [MoO2-
Cl2(thf)2] (0.300 g, 0.87 mmol) in dry dichloromethane (CH2Cl2)
to a suspension of 1 g of MSh-bpy material in dry dichloromethane
(CH2Cl2). The reaction mixture was stirred under
a
N2
In addition, the helical host materials reported here have Lewis
and Brønsted acid properties due to the template extraction proce-
dure (based on an extraction using methanol acidified with HCl in-
stead of calcination) as reported by us recently for related
materials [35,44].
atmosphere at room temperature for 14 h. The resulting material
was then filtered off, washed twice with dichloromethane (CH2Cl2),
and dried under vacuum for 3 h. The corresponding MS-bpy-MoII
and MS-bpy-MoVI materials were also obtained by similar
procedure.
For benchmarking purposes and to understand the role of the
helical matrix in the catalytic performance, regular MCM-41 mate-
rials were also prepared.
2.2. Catalysts characterization
FTIR spectra were obtained either as KBr pellets (complexes) or
Diffuse Reflectance (DRIFT) measurements (materials) on a Nicolet
6700 in the 400–4000 cmꢀ1 range using 2 cmꢀ1 resolution. Powder
XRD measurements were taken on a Philips Analytical PW 3050/60
X’Pert PRO (theta/2 theta) equipped with X’Celerator detector and
with automatic data acquisition (X’Pert Data Collector (v2.0b) soft-
2. Experimental
2.1. Catalysts preparation
All reagents were obtained from Aldrich and used as received.
Commercial grade solvents were dried and deoxygenated by
standard procedures distilled under nitrogen and kept over 4 Å
molecular sieves (3 Å for acetonitrile). The complexes [MoI2(CO)3(-
CH3CN)2], [MoO2Cl2(thf)2] and Me2bpy-MoVI were prepared
according to the literature methods [45–47]. The (ClCO)2bpy ligand
(4,40-dicarbonyl-2,20-bipyridine chloride) was obtained from the
corresponding carboxylic acid (4,40-dicarboxylic-2,20-bipyridine),
which had previously been obtained by oxidation of the methyl
derivative (4,40-dimethyl-2,20-bipyridine) [48].
ware), using a monochromatized Cu K
a radiation as incident beam,
using 40 kV and 30 mA. 1H and 13C solution NMR spectra were ob-
tained with
temperature.
a Bruker Avance 400 spectrometer at room
Solid-state NMR measurements were performed at room tem-
perature on a Bruker MSL 300P spectrometer operating at 59.60
and 75.47 MHz for the observation of 29Si and 13C resonances,
respectively. The standard magic angle spinning (MAS) cross-
polarization–dipolar decoupling RF pulse sequence (CP–DD) was
carried out at ca. 4 kHz spinning rate. For the acquisition of 29Si
spectra, 5 ms contact time, 6 s recycling delay, and a number of
scans always higher than 3000 were selected; the Hartmann–Hahn
condition was optimized using tetrakis(trimethylsilyl)silane, and
tetramethylsilane (tms) was the external reference to set the
chemical shift scale (d = 0). 13C spectra were recorded with 2 ms
contact time, 4 s recycling delay, and a number of scans higher
than 900. The Hartmann–Hahn condition was optimized using gly-
cine, also the external reference to set the chemical shift scale
The complex Me2bpy-MoII was prepared by adding a methanol
solution (10 mL) of 4,40-dimethyl-2,20-bipyridine (Me2bpy)
(0.092 g; 0.5 mmol) to a methanol solution (10 mL) of the precur-
sor complex [MoI2(CO)3(CH3CN)2] (0.258 g; 0,5 mmol). The result-
ing solution was stirred for 14 h at r.t. under inert atmosphere.
After, the solution was filtered off and the solid washed with
dichloromethane (CH2Cl2) and dried under vacuum for 4 h at
323 K.
The helical mesostructured silica materials (MSh), two batches,
were prepared using an entropy-driven procedure with achiral
cationic surfactant templates. In this case, we used cetyltrimethyl-
ammonium bromide (C16TAB) and ammonia as surfactant and co-
surfactant, respectively, according to a literature procedure [32]. In
a typical catalyst synthesis procedure, 0.4 g of C16TAB (Fluka,
96.0%) was dissolved in 50 mL of aqueous ammonia solution (Sig-
ma–Aldrich, 25 wt%) at 313 K. Then, 2.0 mL of tetraethoxysilane
(TEOS) (Aldrich, 98%) was added to the solution. The sol was al-
lowed to react for 3 h at that temperature under stirring
(800 rpm). After this period, the mixture was transferred to a Tef-
lon-lined stainless steel autoclave and aged at 373 K for 24 h.
Two batches of MCM-41 (hereafter denoted as MS) were syn-
thesized as well, one for each Mo species. Both MCM-41 were syn-
thesized by adopting a methodology previously described, using
[(C14H33)N(CH3)3]Br as structure directing agent [35].
(
13CO at 176.1 ppm).
The N2 sorption measurements were obtained in an automatic
apparatus (ASAP 2010; Micromeritics). BET specific surface areas
(SBET, p/p0 from 0.03 to 0.30) and specific total pore volume Vp were
estimated from N2 adsorption isotherms measured at 77 K. The
pore size distributions (PSD) were calculated by the BJH method
using the modified Kelvin equation with correction for the statisti-
cal film thickness on the pore walls [49,50]. The statistical film
thickness was calculated using Harkins–Jura equation in the p/p0
range from 0.1 to 0.95.
Microanalyses for CHN and Mo quantification were performed
at CACTI, University of Vigo. CHN analyses were performed on a Fi-
sons EA 1108; Mo quantification was performed on a Perkin Elmer
Optima 4300DV using In as internal standard.
SEM images were obtained on a FEG-SEM (Field Emission Gun
Scanning Electron Microscope) from JEOL, model JSM-7001F.
TEM images were obtained on a Hitachi microscope, model
H-8100 with a LaB6 filament using an acceleration tension of
200 kV.
To remove the surfactant, the MSh and MS materials (3 g) were
added to a solution of MeOH (250 mL) and HCl 36% (6.0 g). After
stirring for 6 h at 323 K, the solid was filtered off and dried in vac-
uum during 2 h.

C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
23
2.3. Catalytic tests
2,20-bipyridine (Me2bpy) in 1:1 ratio, in methanol or dichlorometh-
ane (Scheme 2).
The complexes and materials were tested in the epoxidation of
olefins and allylic alcohols, such as cis-cyclooctene, styrene, 1-oc-
tene, trans-hex-2-en-1-ol, and R-(+)-limonene, using t-butylhydro-
peroxide (tbhp) as oxidant (Aldrich, 5.5 M in n-decane). The
catalytic oxidation tests were carried out at 328 K under air in a
reaction vessel equipped with a magnetic stirrer and a condenser.
In a typical experiment, the vessel was loaded with olefin or alco-
hol (100 mol%), internal standard (dibutyl ether, dbe), catalyst
(1 mol%), oxidant (200 mol%), and 3 mL of dichloromethane as sol-
vent. The final volume of the reaction is ca. 6 mL. Addition of the
oxidant determines the initial time of the reaction. Conversion,
product yields and stereochemistry were monitored by sampling
periodically and analyzing them using a Shimadzu QP2100-Plus
GC/MS system and a capillary column (Teknokroma TRB-5MS/
TRB-1MS or Restek Rt-bDEXsm) operating in the linear velocity
mode. All reactions were conducted under normal atmosphere.
R-(+)-limonene epoxidation was carried out at different tempera-
tures, 328 K, 353 K, and 393 K, using dichloromethane, ethanol,
and toluene as solvents, respectively.
They were characterized by elemental analysis, FTIR, and 1H
NMR spectroscopy. Elemental analysis of the Me2bpy-MoII and Me2-
bpy-MoVI complexes is shown in Table 1. As can be seen, both
complexes were synthesized successfully as formulated with
experimental values matching closely expected values. For the
Me2bpy-MoVI complex, our experimental values are also in agree-
ment with the literature data [47].
The 1H NMR spectra of Me2bpy-MoII and Me2bpy-MoVI com-
plexes are presented in Table 2. The changes experienced by the
protons are in some cases very clear as shown in Table 2 which
summarizes the observed chemical shift for complexes Me2bpy-
MoII and Me2bpy-MoVI and comparison with the free ligand. As
can be seen from Table 2, all signals shift downfield in the com-
plexes compared to the free Me2bpy ligand. This is most probably
due to the presence of the Mo centers that cause a lower electron
density at the ligand. The 13C NMR spectra were not recorded due
to low solubility.
Characterization by vibrational spectroscopy was also accom-
plished to track formation of the complexes, according to Fig. 1.
In Me2bpy-MoII complex, the strong absorptions of the
mC„O
modes are clearly observed at 2026, 1963, and 1916 cmꢀ1, con-
firming the presence of the [MoI2(CO)3] core.
2.4. Leaching and recycling tests
In complex Me2bpy-MoVI, some characteristic bands are also ob-
served. The symmetric and asymmetric Mo@O stretching modes
In general, these experiments were carried out as described in
the previous section for cyclooctene and R-(+)-limonene epoxida-
tion using material MSh-bpy-MoII as catalyst; conversion and prod-
uct yields were monitored as described above in Section 2.3.
For the leaching experiments after 2 h reaction, the catalyst was
filtered off, and the reactions continued under the same conditions.
In the case of the recycling experiments, after each cycle (24 h), the
catalyst was filtered, washed with CH2Cl2, and dried prior to reuse
in a new catalytic cycle.
are observed as a couple of strong bands at 944 and 917 cmꢀ1
,
respectively. These bands confirm the presence of the [MoO2Cl2]
core and that the initial MoVI oxidation state is preserved. Another
important probe for Mo coordination is the mC@N mode of Me2bpy.
In fact, in the free form, it appears at 1691 cmꢀ1 while after coor-
dination to Mo, it redshifts to 1617 and 1614 cmꢀ1 in Me2bpy-MoII
and Me2bpy-MoVI, respectively.
Characterization of complex Me2bpy-MoII by elemental analysis
confirmed its formulation as proposed in Scheme 2.
2.5. Study on chirality of MSh materials
3.2. Synthesis and characterization of heterogeneous catalysts
This study was based on a literature protocol with slight alter-
ations (phenylalanine was used instead of valine) [51]. A solution
of 50 mg of pure L-phenylalanine (or D-phenylalanine) was
dissolved in 20 mL of MilliQ water followed by adding 20 mg of
template-free MSh material previously activated. The mixture
was magnetically stirred under ambient conditions and at a con-
stant temperature of 298 K. The concentration of L-phenylalanine
(or D-phenylalanine) was measured using UV spectroscopy
(absorption at 260 nm) by sampling at regular time intervals. The
procedure was applied to 5 different batches of MSh materials,
and each one was tested in triplicate.
The regular MCM-41 type mesostructured silica materials (MS)
were prepared using myristyltrimethylammonium bromide
(C14TAB) as surfactant, according to a literature procedure [35].
All subsequent derivatization operations were similar to those
made for the MSh helical materials. Therefore, we will only describe
the procedure for the latter.
The helical mesostructured silica materials (MSh) were prepared
using an entropy-driven procedure with achiral cationic surfactant
template and ammonia. In this particular case, we used cetyltri-
methylammonium bromide (C16TAB) and ammonia as surfactant
and co-surfactant, respectively, according to a literature procedure
[32].
3. Results and discussion
Afterward, a bipyridine (bpy) derivative (Scheme 3) was used as
ligand to coordinate MoII/VI centers after being grafted to the inner
silanol surface of the material. Two batches were produced: one for
preparing the MoII catalyst and another for the MoVI counterpart.
Grafting of bpy ligand was straightforward by reacting (ClCO)2-
bpy with a suspension of MSh in acetonitrile. Afterward, the MoII
and MoVI centers were introduced by suspending the MSh-bpy
materials in dichloromethane and then adding the precursor
complexes, either [MoI2(CO)3(CH3CN)2] or [MoO2Cl2(thf)2]. This
afforded both MSh-bpy-MoII and MSh-bpy-MoVI materials, and
according to elemental analysis, the Mo contents were found to
be 4.26% and 1.21%, corresponding to 0.45 mmol gꢀ1 and
0.13 mmol gꢀ1, respectively. CHN analyses for MSh-bpy-MoII and
MSh-bpy-MoVI revealed values of 11.93 %C, 1.78 %H, and 1.76 %N
for the former and 6.46 %C, 2.0 %H, and 1.10 %N for the latter. Based
on the N content, these results also show that the loading of bpy
3.1. Synthesis and characterization of homogeneous catalysts
4,40-Dicarbonyl-2,20-bipyridine chloride was obtained by reac-
tion of 4,40-dimethyl-2,20-bipyridine (Me2bpy) with concentrated
sulfuric acid (H2SO4) and chromium oxide (CrO3) [48]. The (CO2-
H)2bpy intermediary product was characterized by 1H and 13C
NMR in its sodium salt form using deuterated water as solvent
and was found to be similar to the literature report [48].
After that, reaction of the obtained product (CO2H)2bpy with
thionyl chloride (SOCl2) yields the desired product adopting a liter-
ature procedure [52]. The compound was used without isolation
for the derivatization of the MSh materials, since it decomposes
promptly when exposed to air (see Scheme 1).
The complexes Me2bpy-MoII and Me2bpy-MoVI were prepared
from [MoI2(CO)3(CH3CN)2] or [MoO2Cl2(thf)2] with 4,40-dimethyl-

24
C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
Scheme 1. Synthesis of ligand (ClCO)2bpy for further functionalization of MSh materials. Numbering scheme refers to NMR assignment.
Scheme 2. Synthesis of complexes Me2bpy-MoII and Me2bpy-MoIV
.
Table 1
CHN elemental analysis of Mo complexes.
Complex
Calculated
C
Experimental
H
N
C
H
N
Me2bpy-MoII
Me2bpy-MoIV
29.15
37.62
1.96
3.16
4.53
7.31
28.99
37.45
2.11
3.62
4.60
7.44
Table 2
1H chemical shifts for Me2bpy ligand and respective complexes Me2bpy-MoII and
Me2bpy-MoIV
.
d (ppm)
H3
H5
H6
H7
Me2bpy
8.23
9.16
8.51
7.11
8.10
7.67
8.52
9.18
8.71
2.42
2.63
3.47
Me2bpy-MoII
Me2bpy-MoIV
Fig. 1. FTIR spectra of Me2bpy-MoII and Me2bpy-MoIV complexes.
derivative inside the pores is 1.15 mmol gꢀ1 and 0.72 mmol gꢀ1
,
respectively. By the same token for MS-bpy-MoII and MS-bpy-MoVI
materials, CHN analyses revealed values of 7.34 %C, 1.37 %H, and
0.98 %N for the former and 12.31 %C, 1.75 %H, and 1.42 %N for
the latter. Based on the N content, these results also show that
the loading of bpy derivative inside the pores is 0.64 mmol gꢀ1
and 0.93 mmol gꢀ1, respectively. The Mo content for these materi-
als was found to be 4.36% and 4.97%, corresponding to
0.45 mmol gꢀ1 and 0.52 mmol gꢀ1, by the same order.
All materials were conveniently characterized by DRIFT, powder
XRD, SEM and TEM, and 29Si and 13C CP-MAS SS/NMR. Sorption/
desorption N2 isotherms were also carried out for textural param-
eters estimation. All spectroscopic/textural characterization fea-
tures discussed in the following lines were found to be in
agreement with related hybrid matrix mesoporous materials. The
helical matrices of the as-prepared MSh materials were confirmed
by SEM and TEM measurements Fig. 2. The SEM image shows that
MSh materials consist of rod-like particles. The length of the rods
ranges from hundreds of nanometers to 2 lm, whereas the rod
diameter varies from 80 to 150 nm. TEM images confirm the helical
morphology and illustrate the periodic appearance of lattice
fringes (Fig. 2, right) along the rods, indicating the presence of heli-
cal channels within the rods [31,32]. The periodicity of the lattice
fringes indicates that there is a regular environment of the chan-
nels where reactants and products may diffuse along without
much disturbing.
The local structure was also confirmed by powder X-ray diffrac-
tion measurements, which agree with the published data and also

C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
25
d1 0 0 value is 42.4 Å with a corresponding lattice constant of
a = 49.0 Å; for MSh-bpy-MoII, the values are, respectively,
d1 0 0 = 41.2 Å and a = 48.4 Å. Similar observations were made for
the set of materials MSh, MSh-bpy, and MSh-bpy-MoVI used for the
MoVI derivatives. In the case of the MS materials, XRD powder pat-
terns were found to provide similar results to those already re-
ported by us for related systems [35,44,52]. Despite this, all data
from all materials are collected in Table 3, summarizing the rele-
vant textural properties of all materials leading to the preparation
of both MoII and MoVI derivatives. The observed peak intensity
reduction is common to all materials; it is even more significant
in the materials with the Mo cores. This is not due to a crystallinity
loss, but rather to an X-ray scattering contrast reduction between
the silica walls and the pore-filling material. This has been ob-
served for other types of materials and is well described in the lit-
erature [53,54].
Scheme 3. Derivatization of MSh materials with MoII/VI species.
Nitrogen sorption/desorption studies at 77 K were performed
and have revealed that MSh samples exhibit a reversible type IV
isotherm (Fig. 4), typical of mesoporous solids (pore width be-
tween 2 nm and 50 nm, according to IUPAC) [55]. The calculated
textural parameters (SBET and VP) of these materials (Table 3) agree
with the literature data [56,57]. The capillary condensation/evapo-
ration step in pristine MSh samples appears in the 0.34–0.44 rela-
tive pressures range, while the sharpness of this step reflects a
uniform pore size distribution.
The functionalized material MSh-bpy isotherm revealed much
lower N2 uptake, accounting for the decreases in both SBET (18%
and 14%) and VP (14% and 19%), according to Table 3. These findings
indicate that the ligand immobilization on the internal silica sur-
face was accomplished (Fig. 4, Table 3). For the MSh-bpy-MoII mate-
rial, the SBET and VP relative decrease related to MSh is 31% and 29%,
respectively. This result is in agreement with the p/p0 coordinate
decrease in the isotherm inflection points after post-synthesis
treatments [58]. Furthermore, the maximum of the PSD curve
(Fig. 4) determined by the BJH method, dBJH, for MSh material
changes from 35.3 Å to 34.4 Å (Table 3). For the MSh-bpy-MoVI
material, the relative decrease in SBET and VP when compared to
MSh is 27% for both parameters. These are in line with those de-
scribed above for the other set of materials.
Textural parameters for MS materials (Table 3) were also found
to match values reported in the literature for related systems
[35,44,52] and comparable to those described above for MSh
materials.
Diffuse reflectance infrared spectroscopy (DRIFT) was used to
characterize the hybrid materials, shown in Fig. 5.
Fig. 2. TEM images of MSh materials. The inset shows a SEM image of the particles.
make possible to verify that materials MSh-bpy and MSh-bpy-MoII/VI
are mesostructured.
All resulting materials were of good quality according to the X-
ray diffraction (XRD) powder patterns (Fig. 3).
The XRD powder patterns of the chiral MSh, pristine materials
exhibit four reflections indexed to a hexagonal cell as (100),
(110), (200), and (210) in the 2–10° 2h range. For material MSh,
the d1 0 0 value for reflection (100) is estimated to be 42.3 Å, corre-
p
sponding to a lattice constant of a = 48.8 Å (a = 2d1 0 0
/
3). Materi-
als MSh-bpy and MSh-bpy-MoII, obtained after subsequent stepwise
functionalization with bpy and MoII, still show three reflections
although with a slight deviation of the position maxima toward
higher 2h values as compared to MSh. For MSh-bpy material, the
The DRIFT spectrum of the MSh host material is typical of a sil-
icate evidencing a broad band in the 3600–2600 cmꢀ1 range due to
Fig. 3. Powder XRD of MSh, MSh-bpy, MSh-bpy-MoII and MSh-bpy-MoVI materials.

26
C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
Table 3
Textural parameters of host and composite materials, from powder XRD data and N2 isotherms at 77 K, for all prepared mesoporous materials.
0
0
0
a
b
Material
2h (°)
SBET (m2 gꢀ1
)
D
SBET (%)
VP (cm3 gꢀ1
)
D
VP (%)
d100 (AÅ)
a (AÅ)
dBJH (ÅA)
MS
2.26
2.47
2.49
39.0
35.5
36.1
45.0
41.0
41.6
897
818
755
–
0.75
0.62
0.59
–
34.3
31.9
32.8
MS-bpy
ꢀ9
ꢀ17
MS-bpy-MoII
ꢀ16
ꢀ21
MS
2.22
2.26
2.30
39.7
39.1
38.5
45.8
45.2
44.4
982
914
682
–
ꢀ7
0.85
0.79
0.75
–
ꢀ7
34.3
32.6
31.1
MS-bpy
MS-bpy-MoVI
ꢀ31
ꢀ12
MSh
2.08
2.08
2.11
42.3
42.4
41.2
48.8
49.0
48.4
949
775
657
–
ꢀ18
ꢀ31
0.83
0.71
0.59
–
ꢀ14
ꢀ29
35.3
34.5
34.4
MSh-bpy
MSh-bpy-MoII
MSh
MSh-bpy
MSh-bpy-MoVI
2.12
2.12
2.13
41.6
41.5
41.7
48.0
47.9
48.2
994
859
728
–
ꢀ14
ꢀ27
0.85
0.69
0.62
–
ꢀ19
ꢀ27
36.1
35.4
35.3
a
Surface area variation relatively to parent MSh.
Total pore volume variation relatively to parent MSh.
b
Fig. 4. Nitrogen adsorption studies of MSh, MSh-bpy, MSh-bpy-MoII and MSh-bpy-MoVI materials at 77 K: isotherms (bottom) and pore size distribution curves (top).
hydrogen bonding silanol groups. A sharp band at 3674 cmꢀ1 is
due to silanol groups not involved in hydrogen bonding. Other
important features comprise the band at ca. 1634 cmꢀ1 due to
OH bending modes, while the intense broad band at 1239–
950 cmꢀ1 is assigned to the asymmetric stretching vibration modes
1634 cmꢀ1 band can be related with the
to the OH bending mode of the matrix. Grafting of the bpy ligand
was also monitored by probing its C@O mode. The (ClCO)2bpy li-
mC@N mode in addition
m
gand shows this mode at 1730 cmꢀ1, which agrees with the pres-
ence of the COCl group. After grafting, this mode is redshifted to
1713 cmꢀ1, which is compatible with the expected transformation
resulting in the formation of silyl esters, as documented in the lit-
erature [60]. In this way, the absence of the 1730 cmꢀ1 band in the
MSh-bpy and MS-bpy materials strongly suggest that the bpy ligand
is grafted to the inorganic matrix in a bipodal fashion.
of the mesoporous framework (mSiAOASi) [59]. After grafting of li-
gand bpy, which affords the material MSh-bpy, the DRIFT spectrum
shows an overall similar profile dominated by the absorptions of
the host mesoporous material. Additionally, new bands were de-
tected evidencing the ligand presence within the pores. The

C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
27
Fig. 6. 29Si CP-MAS NMR spectra for MSh, MSh-bpy, MSh-bpy-MoII and MSh-bpy-MoVI
materials.
Fig. 5. Infrared spectra of MSh, MSh-bpy, MSh-bpy-MoII and MSh-bpy-MoVI materials.
The top images show a zoomed image of the C„O (left) and Mo@O modes region
from MSh-bpy-MoII and MSh-bpy-MoVI materials, respectively.
m
m
According to Fig. 6, reaction of MSh with (ClCO)2bpy ligand and
the (organo)metallic fragments [MoI2(CO)3] and [MoO2Cl2] did
not significantly change the 29Si CP-MAS spectrum, as expected,
indicating that the metal fragment reacted with the immobilized
ligand and did not interact with the wall surface. Comparing the
three materials prepared by this approach, the in situ grafting of li-
gand bpy in MSh affording MSh-bpy results in the reduction of the
Q2 and Q3 resonances. Similar results were obtained for the
corresponding MS materials, which show as well the same changes
described for the MSh ones.
After binding the molybdenum complexes [MoI2(CO)3(CH3CN)2]
and [MoO2Cl2(thf)2], which afford materials MSh-bpy-MoII and MSh-
bpy-MoVI, respectively, several changes/additional bands in the cor-
responding DRIFT spectra are detected. In what concerns material
MSh-bpy-MoII, the most remarkable feature is the observation of
three bands at 2050, 2008, and 1928 cmꢀ1 assigned to the
mC„O
modes. These bands are inclusively shifted relatively to the [MoI2(-
CO)3(CH3CN)2] precursor complex. This is indicative of the metal–
ligand binding. Additionally, the bands due to the
mC„N
vibrational modes from the acetonitrile (CH3CN) ligands are not
found, indicating that such ligands have been replaced by the
immobilized ligand.
The 13C CP-MAS SS-NMR spectra (not shown) of the MSh-bpy
and MSh-bpy-MoII materials are quite similar to those of the bpy li-
gand, confirming its binding to the surface. The same was observed
for the corresponding series of MS materials.
The most relevant spectral feature for the MSh-bpy-MoVI mate-
rial is a pair of bands observed at 946 and 914 cmꢀ1 assigned to
the
The FTIR spectra of the MS-bpy-MoII and MS-bpy-MoVI materials
show the presence of above-mentioned bands. In this way, the
C„O modes in MS-bpy-MoII are observed at 2040, 1997 and
1930 cmꢀ1. In the case of MS-bpy-MoVI material, the
Mo@O modes
mMo@O modes of the [MoO2Cl2] core.
3.3. Catalytic studies
m
The catalytic activity of the MSh and MS materials and related
homogeneous counterparts was benchmarked in olefin epoxida-
tion using t-butylhydroperoxide (tbhp) as oxygen source. The
comparison with homogeneous catalysts should be made with
some caution as the Me2bpy ligand has electron donor properties
while the CObpy in the materials shows electron withdrawing
ones. Despite this, the homogeneous catalysts perform quite well
in conversion terms.
m
are observed at 946 and 914 cmꢀ1
.
Fig. 6 shows the 29Si CP-MAS NMR spectra for pristine MSh and
the derivatized MSh-bpy, MSh-bpy-MoII, and MSh-bpy-MoVI
materials.
Unmodified MSh displays two broad convoluted resonances in
the 29Si CP-MAS NMR spectrum at ꢀ101 and ꢀ110 ppm, assigned
to Q4 and Q3 species of the silica framework, respectively [Qn = -
Si(OSi)n(OH)4ꢀn]. A weak shoulder is also observed at ꢀ91 ppm
due to the Q2 species. The Q3 sites are associated with single sila-
nols SiAOH (including hydrogen-bonded silanols), whereas the
Q2 sites correspond to geminal silanols [Si(OH)2]. The 29Si CP-
MAS spectrum of MSh-bpy also displays three broad signals at
ꢀ94, ꢀ101, and ꢀ109 ppm, assigned to Q2, Q3 and Q4 organosilicon
species, respectively.
Special focus was dedicated to stereoselectivity of products due
to the specific helical features of the matrix. In this way, we have
tested the catalysts in the epoxidation of cyclooctene, styrene, 1-
octene, trans-hex-2-en-1-ol, and R-(+)-limonene (Table 4). Blank
runs (without catalyst and in the presence of oxidizing agent) gave
virtually no conversion of starting material.
In the case of cyclooctene, both MSh-bpy-MoII/VI materials
catalyzed selectively the oxidation of the substrate to the

28
C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
Table 4
Catalytic olefin oxidation using MSh and MS heterogeneous catalysts and related homogeneous counterparts.
h
i
Entry
1
Catalyst
Epoxidea
Conv. (%)c
41
Sel. (%)d
ee (%)e
ꢀ1
TOF (ꢁ103) mol mol sꢀ1
b
Mo
Me2bpy-MoII
3
100
15
–
2
4
37
<5
3
4
5
77
0.2
15
67
25
56
100
59
<5
<5
<5f
91
6
7
Me2bpy-MoVI
MS-bpy-MoII
MS-bpy-MoVI
MSh-bpy-MoII
2
53
16
100
45
–
0.3
<5
8
9
99
1
57
28
83
100
71
<5
<5
<5f
10
13
92
11
12
14
99
92
100
1
–
0.5
10
13
14
15
80
3
62
94
77
100
99
13
13
14f
24
91
16
17
130
1
100
18
100
3
–
21
18
19
20
2
56
94
27
100
97
<5
11
15f
6
11
74
21
22
11/6/1g
0.4
100/94/86g
58
100/100/100g
14
–
32
23
24
25
0.6
69
94
100
98
82
2
100
20/4/5g
99/75/57g
100/100/100g
100/76/54f,g
26
27
7
37h
63i
90
96
83f
85f
12
28
MSh-bpy-MoVI
3
94
100
–

C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
29
Table 4 (continued)
h
i
Entry
Catalyst
Epoxidea
Conv. (%)c
26
Sel. (%)d
ee (%)e
ꢀ1
TOF (ꢁ103) mol mol sꢀ1
b
Mo
29
0.4
14
51
30
31
32
76
1
62
30
75
100
71
86
84
95f
11
86
33
14
99h
100
100
94f
93f
34
134
100i
a
All reactions carried out in dichloromethane in the presence of 2 eq. of oxidant (tbhp) and 1 mol% of molybdenum catalyst at 328 K.
Turnover Frequency calculated after 10 min.
Calculated after 24 h.
b
c
d
e
f
Calculated as ‘‘Yield of epoxide’’/‘‘Conversion’’ ꢁ 100%.
Determined by chiral GC.
Value refers do diastereomeric excess (de) of the trans form.
Values correspond to 1st, 2nd and 3rd catalytic runs.
Reaction carried out in ethanol at 353 K.
g
h
i
Reaction carried out in toluene in 383 K.
corresponding epoxide, without formation of any by-products (Ta-
therefore unactivated. Results confirm that all catalysts are also
able to perform successfully such transformation, with conversions
above 50%, showing complete selectivity to the required epoxide
product. From results achieved, we found that, compared to the
homogeneous complexes, the MSh materials showed an extraordi-
nary stereoselectivity. In fact, while the former did not show any ee
of relevance, the latter made possible obtention of octane-1,2-
epoxide as the sole product with quite a remarkable improvement
in ee (an average of 84% for MSh materials against less than 5% for
complexes). On the other hand, the MS materials also show ee
higher than the complexes, but not reaching the performance
achieved by the MSh counterparts (Table 4, entries 13, 18, 23, and
30).
Impelled by the promising results, we extended the test-set to
more sensitive substrates – trans-hex-2-en-1-ol and R-(+)-limo-
nene – to evaluate both chemo- and stereoselectivity. According
to Table 4, epoxidation of the former olefin led to the formation
of hexan-1-ol-2,3-epoxide as major product in all catalytic systems
tested. Once more, both MSh heterogeneous catalysts exceeded the
homogeneous counterparts for this substrate with MSh-bpy-MoII
catalyst reaching almost reaction completeness with virtual abso-
lute selectivity for the desired epoxide product (Table 4, entry 24).
The same results were observed for the MS catalysts which show
also very high substrate conversion yielding the epoxide as the ma-
jor product. However, comparing the measured ee’s, it can be ob-
served that once more, the MSh catalysts outperform the MS
ones. In fact, the MSh yield ee’s of 100% and 84% for the MoII and
MoVI derivatives, while the MS counterparts reach only 13% and
11%, by the same order.
ble 4, entries 21 and 28). The homogeneous counterpart complexes
– Me2bpy-MoII and Me2bpy-MoVI – showed also conversion with
complete selectivity in this transformation. For this set of catalysts,
the results are similar to those reported previously by Kühn and
Thiel for MoVI-related complexes and materials [47,61,62].
However, it is possible to observe that the yield obtained with
the heterogeneous materials is much higher than for remaining
catalysts – Me2bpy-MoII and Me2bpy-MoVI complexes. In the case
of the heterogeneous catalysts MSh-bpy-MoII/VI and MS-bpy-MoII/VI
materials, cyclooctene epoxidation occurs generally in a complete
fashion with the lowest conversion reaching 94% for MSh-bpy-MoVI.
Kinetic profiling of cyclooctene epoxidation shows that both MSh-
bpy-MoII/VI materials present rapid and very high conversion pro-
files (Fig. 7). All other catalysts converted cyclooctene to the de-
sired epoxide in yields with matching kinetics. In addition, it is
not possible to rule out that Mo loading in MSh materials (4.26%
in MSh-bpy-MoII and 1.21% in MSh-bpy-MoVI) has a direct influence
in cyclooctene epoxide yield.
Epoxidation of 1-octene was also benchmarked given that it is
harder to epoxidize as this is an unbranched terminal olefin being
The same trend was observed for R-(+)-limonene. In this case,
both MSh heterogeneous catalysts display outstanding perfor-
mance (Table 4, entries 25 and 32) where limonene epoxide was
obtained not only in excellent yield but also with perfect stereose-
lectivity. The kinetic profiles are similar for all MSh catalysts
(Fig. 8). Although they do not show a rapid beginning, they convert
the substrate to the desired product with good to excellent yield.
However, and as already mentioned, the major advantage of the
heterogeneous catalysts was shown by the resulting product stere-
oselectivity. Performance of the MS materials seems to follow the
same trend in terms of substrate conversion and epoxide yield
for limonene, as described above for MSh counterparts. The excep-
tion seems to be once more stereoselectivity which is again very
low when compared to the helical heterogeneous catalysts.
Fig. 7. Kinetics of cyclooctene epoxide yield in the presence of complexes Me2bpy-
MoII, Me2bpy-MoVI, and materials MSh-bpy-MoII, MSh-bpy-MoVI, MS-bpy-MoII and MS-
bpy-MoVI
.

30
C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
In an attempt to modulate conversion and selectivity for R-(+)-
limonene epoxidation, the effects of solvent and temperature were
tested as well using materials MSh-bpy-MoII and MSh-bpy-MoVI as
catalysts. From the results in Fig. 9 (bottom), it is obvious that reac-
tions in dichloromethane at 328 K lead to high substrate conver-
sion in both cases. But in order to investigate whether the
increase in the temperature with different solvents affected the re-
sults, we used ethanol and toluene. From the obtained results with
different solvents, it is observed that material MSh-bpy-MoII is
much more sensitive to different solvents than material MSh-bpy-
MoVI. According to Fig. 9 for the former, it is not ruled out that a
different solvent with higher reaction temperature leads to better
results in conversion and selectivity, like the MSh-bpy-MoVI coun-
terpart or as observed for other catalysts described previously by
our group [35,52,63].
When assessing catalytic performance of heterogeneous sys-
tems, it is important to evaluate catalyst stability in both terms
of active center leaching and in recycling experiments. In this
way, considering the first vector, active site leaching, we decided
to run a catalytic cycle and remove the catalyst after 2 h reaction
and keep following the reaction in the absence of the catalyst to
evaluate possible leaching of Mo species to the slurry taking mate-
rial MSh-bpy-MoII as catalyst. We found that after 2 h reaction,
cyclooctene conversion is ca. 66%, but after removing the catalyst,
the final conversion after 24 h reaches only 72% while that in the
presence of the catalyst reaches 100% (Table 4, entry 21 and
Fig. 7). It shows clearly that the reaction almost stops indicating
that there is little or no leaching to the homogeneous phase. Such
result makes possible to conclude that the catalysts are truly het-
erogeneous in nature.
The second vector was to evaluate catalyst performance in
recycling experiments. Using once more material MSh-bpy-MoII as
catalyst, we performed a test to understand how recyclable the cat-
alyst is. Recycling was accomplished three times using cyclooctene
or R-(+)-limonene as substrates (Table 4, entries 21 and 25). For
both catalytic systems, loss of activity was found after each
catalytic cycle by ca. 7% and 24% for each of the substrates, respec-
tively. In more detail, cyclooctene conversion reached 100%, 94%,
and 86%, whereas R-(+)-limonene conversion achieved 99%, 75%,
and 57% over three consecutive cycles, although selectivity to the
epoxide was not affected as shown in Fig. 9-top. Despite the epox-
ide selectivity was not affected, in the case of R-(+)-limonene, a
deterioration of the cis/trans ratio over recycle (Table 4, entry 25)
is observed. It should be mentioned that the catalyst was filtered,
washed with dichloromethane, and dried prior to a new recycling
experiment.
Fig. 9. Bottom – Conversion and selectivity for R-(+)-limonene in the presence of
materials MSh-bpy-MoII and MSh-bpy-MoVI under variable temperature and solvent
conditions after 24 h. Top – Conversion and selectivity for cyclooctene and R-(+)-
limonene in the presence of material MSh-bpy-MoII as catalyst in recycling
experiments after 24 h.
Concerning the initial catalytic activity calculated, turnover fre-
quencies (TOF) after 10 min reaction time were found to fit in a
wide range of values. This means that, apart from some exceptions,
the reactions are initially slow, but after the 24 h, they achieve
quite good results as discussed above. Despite this, for instance,
for cyclooctene epoxidation, we report values that range between
2 ꢁ 10–3 and 3 ꢁ 10ꢀ3 mol molMꢀ1o sꢀ1 forꢀt1he complexes and be-
tween 1 ꢁ 10–3 and 130 ꢁ 10ꢀ3 mol molMo sꢀ1 for the materials.
In the case of the Me2bpy-MoII complex, its TOF (2 ꢁ 10–3) is lower
than those found in the literature for related MoII complexes,
which range between 68 ꢁ 10–3 and 91 ꢁ 10ꢀ3 mol molMꢀ1o sꢀ1 with
acetonitrile, nicotinic, and picolinic acids as ligands [ꢀ512,64]. For the
Me2bpy-MoVI complex, its value of 3 ꢁ 10ꢀ3 mol molMo sꢀ1 is found
comparable to those reported by Kühn, using ionic liquids as sol-
vent [47]. The heterogeneous catalysts show also TOF values that
fit with values found in the literature [52,63,64], although there
is one system in this discussion (MS-bpy-MoVI) that outperforms
the literature systems. Such comparisons should be made with
care, since our TOF values were measured after 10 min reaction
time, whereas those found in the literature report TOF values after
only 5 or 15 min reaction time.
In addition, the higher catalytic performance of the heteroge-
neous materials face to the homogeneous ones may be due to
the higher surface acidity which may assist in substrate activation,
a situation that cannot occur with the complexes. This effect has
been described previously in the literature [35,65].
Fig. 8. Kinetics of R-(+)-limonene yield in the presence of complexes Me2bpy-MoII,
Me2bpy-MoVI, and materials MSh-bpy-MoII and MSh-bpy-MoVI
.

C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
31
3.4. Origin of stereoselectivity
matter of fact, in Ref. [32], authors mention channels as being chi-
ral (confirmed in the present work), although the helical rods are
not.
As evidenced by catalytic results for trans-hex-2-en-1-ol stereo-
control was achieved using the heterogeneous catalysts where the
desired hexan-1-ol-2,3-epoxide product was obtained in an abso-
lutely stereoselective way (Table 4, entry 24). The same is observed
for 1-octene and R-(+)-limonene epoxidation being a remarkable
achievement using a simple catalytic system based on a symmet-
ric-ligand molecular catalyst. Compared to the homogeneous
counterparts, the ee obtained with the heterogeneous helical cata-
lysts may arise from the rigid confined environment leading to
substrate hindrance on approaching the catalytic active center in
the confined space of the mesoporous helical host material. This
arises from the described channel left-handedness in the original
literature report and the fact that the grafted bpy ligand may lie
at saddle shaped sites (cf. Fig. 3 in Ref. [32]) in the walls giving rise
to a non-planar position of both N-donor atoms chelating Mo. This
yields a confined environment imposed by the rigid inorganic ma-
trix, which leads to enhanced stereocontrol over the catalytic reac-
tion products.
Face to such positive results, we investigated whether the inor-
ganic matrix has really chiral channels, adopting a literature proce-
dure [51]. Based on that literature procedure to evaluate chirality
of a given inorganic mesoporous material matrix, we have ad-
sorbed enantiopure D-phenylalanine and L-phenylalanine into
the porous system of MSh materials [51]. According to the authors
who reported the original synthesis of MSh materials [32], the heli-
cal materials have left-handed channels. However, that conclusion
was made based on TEM observations, and we think further evi-
dence at the molecular level is required to make a proof of concept
on chiral recognition.
Our results on the adsorption of enantiopure phenylalanine (D
and L; with similar dimensions to the substrates used) show that
MSh materials selectively adsorb twice as much D-phenylalanine
after 24 h, as evidenced in Fig. 10, showing a clear ‘‘preference’’
for D-phenylalanine over L-phenylalanine. This was found to be
reproducible over 5 batches (each one run in triplicate) and
strongly supports that the matrix is chiral. Compared to data from
Ref. [51], arising from adsorption of D- and L-valine, our results
show a greater discrimination between both enantiomers (phenyl-
alanine in the present case). This is most probably due to the high-
er bulkiness of this molecule compared to valine. Nevertheless, it
confirms the observed left-handedness chirality of the channels
in the original synthesis of MSh materials [32] and explains why
reactions achieve the presented results. Attention should be drawn
to the fact that in this work, we are probing channel chirality. As a
Statistical analysis on the results, evidenced by the small error
bars, is also a good tool to demonstrate that the synthesis proce-
dure of such materials is quite reproducible based on a simple en-
tropy-driven mechanism [32], Therefore, it strongly suggests that
the inorganic matrix of MSh materials is really chiral with focus
on one predominant form. This study also confirms previous obser-
vations made by Han et al. in the original synthesis description
where TEM data showed that MSh materials are left-handed [32].
However, further investigation on the nature of the chirality of
the inorganic matrix lie outside the scope of the present research
and should be directed to the original reference [32].
Such confined space chiral concept imposed by the rigid
inorganic backbone has been previously raised and discussed to
explain similar ee enhancements [16–27], where reactants are al-
lowed to approach in a single orientation. Data in Table 4 supports
this assumption by the observation of higher stereoselectivity
achieved for bulkier substrates. In this way, for MSh catalysts, the
trend is more or less, styrene < 1-octene < trans-hex-2-en-1-
ol ꢂ R-(+)-limonene. In addition, present data show that MSh mate-
rials can really be used in processes where chiral recognition is
valuable, such as catalysis or separation technology. This fact is
corroborated by catalytic results arising from experiments using
the MS set of materials which yielded low stereoselectivity results
(see Table 4).
Surprisingly, there are some results of relevance achieved with
MS materials. By the same token explained above for the MSh coun-
terparts, this may be due to the template extracting process. The
HCl used to remove the template can etch the inner walls creating
defects which may, to some extent, create local confined space
which will hinder reactants approach. In this way, since the com-
plexes are not chiral or asymmetric either, the most plausible rea-
son for the reported results may be directly related to how the
complexes graft onto the MSh walls. In fact, if they graft in such a
way that they can behave as pro-chiral catalysts which is imposed
by the walls of the chiral channels then it would result in a high
degree of steric hindrance which render several restrictions on
how the organic reactants approach the metal active center. To fur-
ther support this thesis, results from the recycling experiments
show that stereoselectivity decreases concomitantly. This may be
related to the well-known issue that mesoporous materials suffer
wall degradation by an etching process when exposed to aggres-
sive conditions [66]. In this way, the channels will start to lose
their chirality and the steric restrictions once existed may no long-
er exist, thus leading to the stereoselectivity loss.
4. Conclusion
In conclusion, we describe the first use of chiral helical meso-
porous materials as heterogeneous catalysts for asymmetric olefin
epoxidation based on molybdenum active sites. Although previous
use of such MSh materials was reported in catalytic applications,
no mention to stereocontrol has been made [67,68]. As a major
achievement, these catalysts offer an outstanding stereocontrol
of the reaction products arising from confined space imposed by
the rigid inorganic chiral backbone. In addition, these catalysts
were found to be very effective under several circumstances. For
instance, they can oxidize unactivated unbranched terminal olefins
(1-octene) with good performance. Although leaching does not
occur to a great extent, recyclability seems to be a drawback,
although after three cycles, they are still more efficient than the
homogeneous counterparts. The excellent stereocontrol of the
epoxidation reaction seems to be intrinsically related with
Fig. 10. Adsorption kinetic profiles of pure L-phenylalanine and D-phenylalanine on
MSh materials. The plot represents the average of 5 different batches each one
tested in triplicate (error bars are also included).

32
C.I. Fernandes et al. / Journal of Catalysis 309 (2014) 21–32
[24] T. Yamamoto, M. Suginome, Angew. Chem. Int. Ed. Engl. 48 (2009) 539.
[25] T. Yamamoto, T. Yamada, Y. Nagata, M. Suginome, J. Am. Chem. Soc. 132 (2010)
7899.
[26] T. Yamamoto, Y. Akai, Y. Nagata, M. Suginome, Angew. Chem. Int. Ed. Engl. 50
(2011) 8844.
[27] Y. Nagata, S. Ohashi, M. Suginome, J. Polym. Sci. Part A: Polym. Chem. 50
(2012) 1564.
[28] T. Gier, X. Bu, P. Feng, G.D. Stucky, Nature 395 (1998) 154.
[29] D. Bradshaw, T.J. Prior, E.J. Cussen, J.B. Claridge, M.J. Rosseinsky, J. Am. Chem.
Soc. 126 (2004) 6106.
[30] A.E. Rowan, R.J.M. Nolte, Angew. Chem. Int. Ed. Engl. 37 (1998) 63.
[31] S. Che, Z. Liu, T. Ohsuna, K. Sakamoto, O. Terasaki, T. Tatsumi, Nature 429
(2004) 281.
[32] Y. Han, L. Zhao, J.Y. Ying, Adv. Mater. 19 (2007) 2454.
[33] D. Trong-On, D. Desplantier-Giscard, C. Danumah, S. Kaliaguine, Appl. Catal. A
222 (2001) 299.
confinement effects imposed by the chiral inorganic backbone
which modulates the way reactants approach active sites. This
arises from experimental data confirming the chirality of the
inorganic matrix which selectively adsorbs a given enantiomer
over another. Despite this, recycling experiments also evidence a
deterioration of stereoselectivity which may be related to the deg-
radation of the chiral channel system of the host materials, which
is known to occur under aggressive conditions.
This is certainly an important contribution for such concept and
may render such materials further applications where chiral recog-
nition is important.
Acknowledgments
[34] A.M. Balu, J.M. Hidalgo, J.M. Campelo, D. Luna, R. Luque, J.M. Marinas, A.A.
Romero, J. Mol. Catal. A: Chem. 293 (2008) 17.
[35] N.U. Silva, C.I. Fernandes, T.G. Nunes, M.S. Saraiva, C.D. Nunes, P.D. Vaz, Appl.
Catal. A 408 (2011) 105.
[36] N. Grover, F.E. Kühn, Curr. Org. Chem. 16 (2012) 16.
[37] M.R. Maurya, Curr. Org. Chem. 16 (2012) 73.
[38] P.D. Vaz, C.D. Nunes, Curr. Org. Chem. 16 (2012) 89.
[39] C.D. Nunes, P.D. Vaz, in: A.C. Poehler (Ed.), Homogeneous Catalysts: Types,
Reactions and Applications, Nova Science Publishers Inc., New York, 2011, p.
43.
[40] A.J. Burke, Coord. Chem. Rev. 252 (2008) 170.
[41] F.E. Kühn, A.M. Santos, M. Abrantes, Chem. Rev. 106 (2006) 2455.
[42] A.U. Barlan, A. Basak, H. Yamamoto, Angew. Chem. Int. Ed. Engl. 45 (2006)
5849.
[43] A.U. Barlan, W. Zhang, H. Yamamoto, Tetrahedron 63 (2007) 6075.
[44] N.U. Silva, T.G. Nunes, M.S. Saraiva, M.S. Shalamzari, P.D. Vaz, O.C. Monteiro,
C.D. Nunes, Appl. Catal. B 113–114 (2012) 180.
[45] P.K. Baker, Chem. Soc. Rev. 27 (1998) 125.
[46] W.M. Carmichael, D.A. Edwards, G.W.A. Fowles, P.R. Marshall, Inorg. Chem.
Acta 1 (1967) 93.
[47] D. Betz, W.A. Herrmann, F.E. Kühn, J. Organomet. Chem. 694 (2009) 3320.
[48] N. Garelli, P. Vierling, J. Org. Chem. 57 (1992) 3046.
[49] M. Kruk, M. Jaroniec, A. Sayari, Langmuir 13 (1997) 6267.
[50] M. Kruk, V. Antochshuk, M. Jaroniec, A. Sayari, J. Phys. Chem. B 103 (1999)
10670.
[51] Z. Guo, Y. Du, Y. Chen, S.-C. Ng, Y. Yang, J. Phys. Chem. C 114 (2010) 14353.
[52] A.C. Ventura, C.I. Fernandes, M.S. Saraiva, T.G. Nunes, P.D. Vaz, C.D. Nunes,
Curr. Inorg. Chem. 1 (2011) 156.
[53] B. Marler, U. Oberhagemann, S. Voltmann, H. Gies, Micropor. Mater. 6 (1996)
375.
[54] W. Hammond, E. Prouzet, S.D. Mahanti, T.J. Pinnavaia, Micropor. Mesopor.
Mater. 27 (1999) 19.
[55] S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and Porosity, second ed.,
Academic Press, London, 1982.
[56] M.D. Alba, A. Becerro, J. Klinowski, J. Chem. Soc., Faraday Trans. 92 (1996) 849.
[57] A.A. Romero, M.D. Alba, W. Zhou, J. Klinowski, J. Phys. Chem. B 101 (1997)
5294–5300.
[58] M. Kruk, M. Jaroniec, Langmuir 15 (1999) 5410.
[59] P.D. Vaz, C.D. Nunes, M. Vasconcellos-Dias, M.M. Nolasco, P.J.A. Ribeiro-Claro,
M.J. Calhorda, Chem. Eur. J. 13 (2007) 7874.
[60] W. Bett, S. Cradoc, Monatsh. Chem. 111 (1980) 193.
[61] J. Tang, L. Wang, G. Liu, Y. Liu, Y. Hou, W. Zhang, M. Jia, W.R. Thiel, J. Mol. Catal.
A 313 (2009) 31.
Authors thank FCT, POCI and FEDER (Projects PTDC/QUI/71576/
767 2006, PEst-OE/QUI/UI0612/2013, RECI/QEQ-QIN/0189/2012
and PEst-OE/QUI/UI0100/2013). M.S.S. thanks FCT for a research
Grant (SFRH/BD/48640/2008).
We would like to thank the Editor of Journal of Catalysis, Prof.
S.T. Oyama and all the referees, for many pertinent and timely sug-
gestions that contributed for a better discussion of data.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2013.08.032.
References
[1] A. Taguchi, F. Schüth, Micropor. Mesopor. Mater. 77 (2005) 1.
[2] M.H. Valkenberg, W.F. Hölderich, Catal. Rev. – Sci. Eng. 44 (2002) 321.
[3] F. Hoffmann, M. Cornelius, J. Morell, M. Fröba, Angew. Chem. Int. Ed. Engl. 45
(2006) 3216.
[4] T. Katsuki, K.B. Sharpless, J. Am. Chem. Soc. 102 (1980) 5974.
[5] Y. Gao, R.M. Hanson, J.M. Klunder, S.Y. Ko, H. Masamune, K.B. Sharpless, J. Am.
Chem. Soc. 109 (1987) 5765.
[6] R. Irie, K. Noda, Y. Ito, N. Matsumoto, T. Katsuki, Tetrahedron Lett. 31 (1990)
7345.
[7] W. Zhang, J.L. Loebach, S.R. Wilson, E.N. Jacobsen, J. Am. Chem. Soc. 112 (1990)
2801.
[8] Q.-H. Xia, H.-Q. Ge, C.-P. Ye, Z.-M. Liu, K.-X. Su, Chem. Rev. 105 (2005) 1603.
[9] M. Shibasaki, H. Sasai, T. Arai, Angew. Chem. Int. Ed. Engl. 36 (1997) 1236.
[10] M.J. Porter, J. Skidmore, Chem. Commun. (2000) 1215.
[11] D. Díez, M.G. Núñez, A.B. Antón, P. García, R.F. Moro, N.M. Garrido, I.S. Marcos,
P. Basabe, J.G. Urones, Curr. Org. Synth. 5 (2008) 186.
[12] Y. Shi, Acc. Chem. Res. 37 (2004) 488.
[13] M. Hickey, D. Goeddel, Z. Crane, Y. Shi, Proc. Natl. Acad. Sci. USA 101 (2004)
5794.
[14] S. Juliá, J. Masana, J.C. Vega, Angew. Chem. Int. Ed. Engl. 19 (1980) 929.
[15] S. Juliá, J. Guixer, J. Masana, J. Rocas, S. Colonna, R. Annuziata, H. Molinari, J.
Chem. Soc. Perkin Trans. 1 (1982) 1317.
[62] C.D. Nunes, A.A. Valente, M. Pillinger, A.C. Fernandes, C.C. Romão, J. Rocha, I.S.
Gonçalves, J. Mater. Chem. 12 (2002) 1735.
[63] M. Vasconcellos-Dias, C.D. Nunes, P.D. Vaz, P. Ferreira, P. Brandão, V. Félix, M.J.
Calhorda, J. Catal. 256 (2008) 301.
[16] J. Wang, L. Zhao, H. Shi, J. He, Angew. Chem. Int. Ed. Engl. 50 (2011) 9171.
[17] M.D. Jones, R. Raja, J.M. Thomas, B.F.G. Johnson, D.W. Lewis, J. Rouzaud, K.D.M.
Harris, Angew. Chem. Int. Ed. Engl. 42 (2003) 4326.
[64] M. Vasconcellos-Dias, S.R.M.M. de Aguiar, C.D. Nunes, P.D. Vaz, T.G. Nunes, M.J.
Calhorda, Curr. Inorg. Chem. 1 (2011) 146.
[18] R. Raja, J.M. Thomas, M.D. Jones, B.F.G. Johnson, D.E.W. Vaughan, J. Am. Chem.
Soc. 125 (2003) 14982.
[65] C.D. Nunes, A.A. Valente, M. Pillinger, J. Rocha, I.S. Gonçalves, Chem. Eur. J. 9
(2003) 4380.
[19] M.R. Castillo, L. Fousse, J.M. Fraile, J.I. García, J.A. Mayoral, Chem. Eur. J. 13
(2007) 287.
[66] A. Galarneau, M. Nader, F. Guenneau, F. Di Renzo, A. Gedeon, J. Phys. Chem. C
111 (2007) 8268.
[20] J.M. Fraile, J.I. García, C.I. Herrerías, J.A. Mayoral, M.A. Harmer, J. Catal. 221
(2004) 532.
[67] A.I. Carrillo, E. Serrano, J.C. Serrano-Ruiz, R. Luque, J. Garcia-Martinez, Appl.
Catal. A: Gen. 435–436 (2012) 1.
[21] J.M. Fraile, J.I. García, C.I. Herrerías, J.A. Mayoral, O. Reiser, A. Socullamos, H.
Werner, Chem. Eur. J. 10 (2004) 2997.
[68] A.I. Carrillo, E. Serrano, R. Luque, J. Garcia-Martinez, Chem. Commun. 46 (2010)
5163.
[22] R.A. García, V. Morales, T. Garcés, J. Mater. Chem. 22 (2012) 2607.
[23] R.A. García, R. van Grieken, J. Iglesias, V. Morales, N. Villajos, J. Catal. 274
(2010) 221.