Easily recoverable titanosilicate zeolite beads with hierarchical porosity: Preparation and application as oxidation catalysts

Article abstract of DOI:10.1016/j.jcat.2015.09.017

Titanosilicate zeolite beads with hierarchical porosity and 0.2-0.5 mm diameter (HPB-TS-1) have been synthesized from a titanosilicate solution, employing a porous anion-exchange resin as shape- and structure-directing template. The characterization results showed the existence of crystalline TS-1 nanoparticles and of a network of connected large meso/macropores in the interior of the beads. These bead materials are active and selective heterogeneous catalysts for two classes of industrially relevant oxidation reactions: the hydroxylation of phenol and the epoxidation of alkenes. In both cases, a green oxidant such as hydrogen peroxide was utilized. HPB-TS-1 beads displayed high activity in the hydroxylation of phenol, due to their crystalline TS-1 structure with isolated tetrahedral Ti species. In the epoxidation of cyclohexene, HPB-TS-1 gave conversion comparable to that over a benchmark catalyst such as Ti-MCM-41, but more than double yield and selectivity of the target product, cyclohexene oxide. HPB-TS-1 was also active and selective in the epoxidation of other bulky alkenes, i.e., cis-cyclooctene and trans, trans, cis-1,5,9-cyclododecatriene. The catalytic performance of HPB-TS-1 stems from the presence of crystalline TS-1 nanodomains, which are accessible through a network of meso/macropores within the beads. Notably, the bead format of these catalysts causes their spontaneous settling upon cessation of the agitation, thus enabling their straightforward separation from the reaction mixture. The HPB-TS-1 catalysts could be efficiently recycled in the hydroxylation of phenol with aqueous H₂O₂ via a calcination approach.

Full text of DOI:10.1016/j.jcat.2015.09.017

Journal of Catalysis 333 (2016) 139–148
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Easily recoverable titanosilicate zeolite beads with hierarchical porosity:
Preparation and application as oxidation catalysts
Wenjing Cheng a,1, Yanqiu Jiang a,1, Xianzhu Xu , Yan Wang , Kaifeng Lin , Paolo P. Pescarmona b,
a
a
a,
⇑
⇑
a
Academy of Fundamental and Interdisciplinary Sciences, Harbin Institute of Technology, 150080 Harbin, China
Chemical Engineering Department, University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Titanosilicate zeolite beads with hierarchical porosity and 0.2–0.5 mm diameter (HPB-TS-1) have been
synthesized from a titanosilicate solution, employing a porous anion-exchange resin as shape- and
structure-directing template. The characterization results showed the existence of crystalline TS-1
nanoparticles and of a network of connected large meso/macropores in the interior of the beads.
These bead materials are active and selective heterogeneous catalysts for two classes of industrially rel-
evant oxidation reactions: the hydroxylation of phenol and the epoxidation of alkenes. In both cases, a
green oxidant such as hydrogen peroxide was utilized. HPB-TS-1 beads displayed high activity in the
hydroxylation of phenol, due to their crystalline TS-1 structure with isolated tetrahedral Ti species. In
the epoxidation of cyclohexene, HPB-TS-1 gave conversion comparable to that over a benchmark catalyst
such as Ti-MCM-41, but more than double yield and selectivity of the target product, cyclohexene oxide.
HPB-TS-1 was also active and selective in the epoxidation of other bulky alkenes, i.e., cis-cyclooctene and
trans, trans, cis-1,5,9-cyclododecatriene. The catalytic performance of HPB-TS-1 stems from the presence
of crystalline TS-1 nanodomains, which are accessible through a network of meso/macropores within the
beads. Notably, the bead format of these catalysts causes their spontaneous settling upon cessation of the
agitation, thus enabling their straightforward separation from the reaction mixture. The HPB-TS-1 cata-
Received 22 July 2015
Revised 23 September 2015
Accepted 26 September 2015
Keywords:
Titanosilicates
Hierarchical porosity
Connected meso/macropores
Bead-shaped catalysts
Heterogeneous oxidation catalysts
Straightforward separation
lysts could be efficiently recycled in the hydroxylation of phenol with aqueous H
approach.
2 2
O via a calcination
Ó 2015 Elsevier Inc. All rights reserved.
1
. Introduction
size to the nanometer scale [12–15], which leads to an increase in
the external surface area. However, the preparation and catalytic
application of these zeolite nanocrystals is hampered by their awk-
ward separation from the reaction mixture, normally requiring
high-speed centrifugation or special filtration. This hinders the
recovery of the products and the recycling of the catalysts and
therefore limits broad industrial application of these zeolite
nanocrystals as heterogeneous catalysts in liquid-phase oxidation.
An alternative way to overcome the diffusion limitation of
bulky molecules in zeolites is the creation of mesopores or macro-
pores in the zeolitic structure [16]. Well-defined meso- or macro-
porous arrays could show optimal fluxes, thus minimizing issues
related to diffusion. In recent years, there has been growing inter-
est in the improved transport properties of such hierarchically por-
ous zeolites with microporous crystalline domains and meso- or
macropores [16,17]. The strategies used to generate this secondary
porosity in zeolites up to now can be classified as non-templated or
templated routes [13,18–20]. The former include approaches such
as dealumination by acid leaching or steaming at high tempera-
ture, and selective desilication of zeolites in basic media [21].
Since the groundbreaking discovery of zeolite TS-1 (MFI frame-
work type), titanium-substituted zeolites have been attracting
much attention for their remarkable catalytic performance in the
selective oxidation of alkenes, alkanes, aromatics, alcohols, and
2 2
other substrates, using aqueous H O as a clean and environmen-
tally friendly oxidant [1,2]. Besides their high surface area, tunable
pore size, and high thermal stability, the excellent catalytic activity
of such titanosilicate zeolites stems from the isolated, tetrahedrally
coordinated Ti species [3–11]. However, these titanosilicate zeo-
lites cannot effectively catalyze the conversion of bulky molecules,
which are too large to access the active titanium sites located
inside the micropores (<1 nm). The diffusion limitation imposed
by the pores of zeolites can be mitigated by decreasing the particle
⇑
Corresponding authors. Fax: +31 50 3636521 (P.P. Pescarmona).
E-mail addresses: linkaifeng@hit.edu.cn (K. Lin), p.p.pescarmona@rug.nl
(
P.P. Pescarmona).
1
These authors have contributed equally to this study.
http://dx.doi.org/10.1016/j.jcat.2015.09.017
021-9517/Ó 2015 Elsevier Inc. All rights reserved.
0

1
40
W. Cheng et al. / Journal of Catalysis 333 (2016) 139–148
Though these routes are effective in creating secondary mesopores
in the crystalline structure, they do not allow controlling the volume
and diameter of the mesopores [20,22]. Alternatively, routes
employing soft or hard templates with well-defined morphology
and structure have been widely used, as they allow tailoring the size
and geometry of the hierarchical pores. With this approach, meso-
porous zeolites can be prepared using carbonaceous template mate-
rials such as carbon nanotubes, carbon black, carbon aerogels and
ordered mesoporous carbon, but the preparation of these carbon
templates is complicated and high temperatures and inert gases
are needed during carbonization [12,18,23]. For soft templating
methods, the most commonly used templates include traditional
surfactants for mesopore creation, organosilane agents, sugars, and
cationic polymers [22,24–26]. Such hierarchically porous zeolites
display very promising catalytic performances in heterogeneous
reactions with both small and bulky substrates. However, most of
the attention has been devoted to the preparation of hierarchical
aluminosilicate zeolites (ZSM-5 and Beta), whereas the synthesis
of hierarchical titanosilicate zeolites is less explored. This might be
due to the difficulty of incorporating isolated, tetrahedral Ti species
into the zeolite framework when a template for the generation of
mesopores is applied in the process. Another challenge in the syn-
thesis of these materials is to achieve a good connection between
the created mesopores, which is considered to be vital to maximize
the benefits of hierarchical porosity in catalysis [27]. Recently, the
synthesis of a hierarchical TS-1 with a well connected network of
meso/macropores inside the zeolite crystal using caramel as tem-
plate was reported [28]. This hierarchical TS-1 showed improved
catalytic performance compared with conventional TS-1 in the
selective oxidation of cyclohexene, thiophene, benzothiophene,
to the titanosilicate solution with a solution-to-resin mass ratio of
20:1. The mixture was stirred at 40 °C for 24 h, transferred into an
autoclave, and hydrothermally treated with the following tempera-
ture programming: 60 °C for 24 h, from 60 to 100 °C at 10 °C/h,
100 °C for 19 h, from 100 to 165 °C at 16 °C/h, 165 °C for 24 h. After
cooling to room temperature, the aqueous slurry was decanted and
the beads were washed several times with distilled water and finally
dried at 60 °C. Meanwhile, a powder sample was obtained from the
slurry via centrifugation. The beads obtained in this step were then
subjected to acid treatment (2 M nitric acid at 120 °C for 24 h) with
the purpose of removing possible extraframework Ti species [5,8].
The resin beads used as templating agents and other organic resi-
dues were removed by calcination in air at 550 °C for 6 h, which
led to the formation of the hierarchically porous TS-1 zeolite beads
denoted as HPB-TS-1. The as-synthesized hierarchical porous TS-1
zeolite beads calcined without prior treatment by nitric acid were
denoted as HPB-TS-1AS. The calcined powder obtained from the
slurry was denoted as Pow-TS-1. Based on the amount of tetraethyl
orthosilicate employed in thesynthesis, thetypical yield of HPB-TS-1
was 20–25% and that of Pow-TS-1 was 50–55%. Moreover, hierar-
chically porous titanosilicate beads were also prepared with
different solution-to-resin mass ratios (6.8:1 and 10:1), while all
other conditions were as described above. These materials were
named HPB-TS-1(6.8) and HPB-TS-1(10), where the value in
parentheses indicates the solution-to-resin mass ratio. Another set
of reference titanosilicate bead samples (HPB-TiSil-60, HPB-TiSil-
100, HPB-TiSil-170) was prepared according to the general proce-
dure described above, with the difference that the hydrothermal
treatment was carried out at a given fixed temperature (60, 100, or
170 °C) for 72 h instead of using the programmed temperature
profile (vide supra). All the methods used in the preparation of the
materials presented in this work are summarized in Scheme S1
(see the Supplementary Information).
2 2
and 4,6-dimethyl dibenzothiophene using H O as the oxidant.
These results underline the relevance of developing novel hierarchi-
cally porous titanosilicate zeolites with highly active Ti species and
connected networks of mesopores.
In a previous paper of some of us, a novel method for the prepa-
ration of hierarchically porous titanosilicate beads based on the
use of anion-exchange resin beads as shape- and structure-
directing agents was reported [29]. The bead-shaped catalysts dis-
played high catalytic activity, due to their large mesopores, and
were easily separated and recycled without requiring centrifuga-
tion or filtration, owing to their bead format. However, based on
the preparation conditions applied, the microstructure of titanosil-
icate beads was mainly amorphous, whereas a crystalline TS-1
structure would be expected to lead to enhanced catalytic activity
in oxidation reactions. In the present work, hierarchical TS-1 beads
with a connected network of meso/macropores were successfully
synthesized using anion-exchange resin beads as a hard template.
The obtained beads were used to catalyze the hydroxylation of aro-
2.2. Characterization
Powder XRD patterns were measured on a Bruker D8 instrument
using Cu K
a radiation. FT-IR spectra were recorded on a Perkin
Elmer 100 spectrometer. UV–vis spectra were measured with a Per-
kin Elmer Lambda 750 spectrophotometer in the region 200–
800 nm. The samples were not treated before measurement and
the UV–vis spectra were recorded in air. Prior to analysis by XRD,
FT-IR, and UV–vis measurements, the beads were ground to a pow-
der. The isotherms of nitrogen adsorption–desorption were mea-
sured at liquid nitrogen temperature using a Micromeritics ASAP
2020 system. The micropore size distribution was calculated using
the Horvath–Kawazoe (HK) model [30] and the mesopore size dis-
tribution was obtained using the Barrett–Joyner–Halenda (BJH)
model [31] applied to the desorption branch. Scanning electron
microscopy (SEM) images were taken on a Hitachi SU8010 appara-
tus. SEM analysis of the as-synthesized bead-shaped samples pro-
vided the average bead size and the surface morphology, whereas
the inner structure was investigated by SEM analysis after the
beads were ground into powder. Transmission electron microscopy
(TEM) images were taken on an FEI-T12 electron microscope. For
the preparation of the samples for TEM analysis, the beads were
ground into powder, which was then dispersed in ethanol, followed
by sonication of the suspension for 10 min. One drop of the suspen-
sion was placed on a TEM grid and allowed to dry overnight. Si/Ti
molar ratios were determined by elemental analysis using an Agi-
lent ICP-MS 7500c inductively coupled plasma mass spectrometer.
2 2
matics and the epoxidation of various alkenes with aqueous H O
as a green oxidant, and proved to display high activity and selectiv-
ity with all substrates tested.
2
. Experimental
2.1. Synthesis of the catalytic materials
Amberlite IRA-900 resin beads in the chloride form with a bead
size of 16–50 mesh and an anion-exchange capacity of 4.2 meq/g
Alfa Aesar) was used as a templating agent. Amberlite IRA-900
(
is a macroreticular resin with high porosity and with benzyltrialky-
lammonium functionalities.
In a typical synthesis of hierarchically porous TS-1 zeolite beads
(
HPB-TS-1), a titanosilicate solution was prepared by mixing aque-
2.3. Catalytic tests
ous tetrapropylammonium hydroxide (TPAOH, 50%, 2.0 mL), H
2
O
(
(
8.0 mL), titanium isopropoxide (0.2 g), and tetraethyl orthosilicate
5 mL) under stirring at 40 °C. Next, IRA-900 resin beads were added
The liquid-phase hydroxylation of phenol was carried out under
shaking at 120 rpm, in glass vials soaked in a water bath at 80 °C.

W. Cheng et al. / Journal of Catalysis 333 (2016) 139–148
141
Shaking was chosen as the agitation method because the beads
tend to deteriorate upon stirring with a magnetic bar. Typically,
the reaction mixture consisted of 470 mg of phenol (5 mmol),
prepared in this work (vide infra). A crucial feature of the method
used to prepare HPB-TS-1 was the accurate control of the
temperature profile during the hydrothermal step of the synthesis.
The HPB-TS-1 samples obtained after synthesis and calcination
consist of white beads with diameter ranging from 0.2 to 0.5 mm
(Fig. 1D and I), which is slightly smaller than that of the starting
resin beads, as observed in other amorphous titanosilicate beads
[32]. In contrast, the centrifugate from the decanted aqueous
slurry, denoted as Pow-TS-1, is a fine powder consisting of TS-1
nanoparticles (vide infra for the XRD analysis) with a size of
300–400 nm (Fig. 1H). Pow-TS-1 has morphology and particle size
similar to those of a reference TS-1 sample that was prepared by an
analogous procedure but without the resin beads (Fig. 1N). This
indicates that a fraction of conventional TS-1 grew outside the
beads of HPB-TS-1 during the applied synthesis procedure. The
surface of the as-synthesized beads (HPB-TS-1AS) is partially cov-
ered with a layer of well-defined, discrete particles with the typical
shape of MFI zeolite crystals (Fig. 1E and F), but with different mor-
phology compared to the particles in Pow-TS-1. This layer of parti-
cles was removed from the surface of the beads by the acid
treatment used in the preparation of the final catalyst (HPB-TS-1,
Fig. 1I and J). The layer of particles can also be washed away from
the surface of HPB-TS-1AS by water treatment at 80 °C for 4 h (see
Fig. S1), suggesting that the interaction between the discrete parti-
cles and the bead body is rather weak. As anticipated, HPB-TS-1
contains a large number of irregular mesopores as well as a small
number of macropores (Fig. 1J and K). This porosity reflects the
related features of the surface of the resin beads used as template
(Fig. 1B). The meso- and macropores in HPB-TS-1 are expected to
be connected because they are created by removal of the continu-
ous resin structure upon calcination.
1
5 mL of distilled water, and 50 mg of catalyst. When the temper-
ature was increased to 80 °C, aqueous H (35 wt.%) at a phenol-
to-H molar ratio of 3 was added in one dose. At the end of the
2 2
O
2 2
O
test, the bead-shaped catalysts deposited spontaneously at the bot-
tom of the vial within a few seconds after the agitation was
stopped. Therefore, the beads were easily separated from the reac-
tion solution by removal of the liquid with a pipette. The catalysts
in powder form (Pow-TS-1, TS-1) were separated from the reaction
solution by centrifugation. After reaction for 4 h, the products were
determined using a gas chromatograph (GC) SP-6800A equipped
with an SE-54 capillary column (30 m, 0.32 mm) and an FID detec-
tor. The reusability of the bead-shaped catalysts in consecutive
runs was evaluated following two different recycling approaches:
the first involving washing with ethanol, and the second including
calcination at 550 °C for 4 h in air after the washing step.
For the epoxidation of cyclohexene with aqueous
.5 mmol of cyclohexene, 4.5 mL of acetonitrile, 2.25 mmol of
aqueous H (35 wt.%), and the catalyst (60 mg) were mixed in
2 2
H O ,
4
2 2
O
the vial under shaking and heated to 60 °C. After reaction for 5 h,
the products were analyzed using an Agilent 7890A GC system
equipped with an HP-5 column (30 m, 0.32 mm).
The assignment of the products was confirmed by gas chro-
matography–mass spectrometry analysis (GC–MS) on an Agilent
7
890A gas chromatograph (HP-5MS column, 30 m, 0.25 mm) cou-
pled to an Agilent 5975 MSD mass spectrometer.
2
The amount of H decomposed in H O and O during the cat-
2
O
2
2
alytic test was determined by means of the titration of the reaction
solution (after 4 h at 80 °C for phenol hydroxylation or after 5 h at
6
0 °C for cyclohexene epoxidation) with a 0.1 M solution of Ce
High-magnification SEM images of the interiors of HPB-TS-1AS
and HPB-TS-1 indicate that the beads are composed of particles
with the typical morphology of MFI-type crystals, with a rather
uniform size of ca. 150–200 nm (Fig. 1G and K). The formation of
these nanocrystals with uniform size is attributed to their compet-
itive growth in the confined space within the resin beads. Note that
the zeolitic particles are interconnected, so that they build a robust
bead. Therefore, the mesopores originating from the interstices
between the zeolite nanocrystals should be considered structural
(
SO
4
)
SO
2
, obtained by dissolving Ce(SO
(28 mL) and bidistilled H O (28 mL), and diluted to a total
O. The obtained colorless solution was
4
)
2
Á4H
2
O (50.0 mmol) in
H
2
4
2
volume of 500 mL with H
titrated with the 0.1 M Ce solution until it turned yellow
2
IV
4
+
3+
+
(
2 2 2
2Ce + H O ? 2Ce + 2H + O ).
3
. Results and discussion
In this work, a titanosilicate with a macroscopic bead format
(
and thus different from the interparticle mesopores that can be
observed in the case of discrete nanoparticles).
containing TS-1 nanocrystals surrounded by a network of meso-
and macropores (HPB-TS-1) was prepared and tested as a hetero-
geneous catalyst for oxidation reactions with hydrogen peroxide
as oxidant. The TS-1 nanocrystals are expected to contain highly
active tetrahedral Ti sites and to display a relatively high external
surface area, while the hierarchical porous structure will grant
accessibility to the TS-1 domains. Moreover, the bead format will
allow straightforward separation of the catalyst from the reaction
mixture. The successful synthesis of a catalytic material with these
highly desirable features was demonstrated by full characteriza-
tion of HPB-TS-1 and of selected reference materials with a combi-
nation of techniques.
Transmission electron microscopy (TEM) images of HPB-TS-1
confirm the presence of large mesopores and macropores in the
interior of the beads (Fig. 2A and B). TEM images with higher mag-
nification (Fig. 2C and D) show the well-defined crystalline
domains of the microporous MFI zeolite, located in the meso/
macroporous structure of HPB-TS-1. The combination of SEM and
TEM analyses showed that the ordered micropores of the TS-1
crystals in HPB-TS-1 are partially open onto the meso/macropores,
leading to the formation of the desired interconnected hierarchical
pore system.
Characterization of HPB-TS-1 by powder X-ray diffraction (XRD)
evidences a well-resolved MFI pattern (Fig. 3), thus confirming the
presence of the TS-1 crystals observed by SEM and TEM. Notably,
the crystallinity of the acid-treated HPB-TS-1 is superior to that of
the as-synthesized HPB-TS-1AS, as indicated by the disappearance
of the broad band at 23° assigned to amorphous (titano)silicate
(Fig. 3A and B). This implies that amorphous species could be
mostly removed via acid treatment. It is worth noting that the acid
treatment procedure (reflux at 120 °C for 24 h) is similar to the
procedure employed for evaluating the hydrothermal stability of
porous materials (treatment in boiling water) [34]. Importantly,
the crystallinity level of the TS-1 particles within the beads in
HPB-TS-1 is analogous to that of the TS-1 sample prepared in the
absence of the bead template (Fig. 3D) and to the powder sample
3.1. Synthesis and characterization of the catalysts
The HPB-TS-1 material was prepared employing Amberlite IRA-
00 resin beads as shape- and structure-directing agent. These
9
resin beads display irregular pores in the mesoporous range and,
to a lesser extent, in the macroporous range (Fig. 1A–C). The choice
of these resin beads was inspired by previous reports [29,32,33],
but the synthesis method was carefully tuned in this work to
achieve the formation of TS-1 nanocrystals within the porous
beads. This was a challenging yet rewarding task, as demonstrated
by the less attractive features of the other bead materials that were

1
42
W. Cheng et al. / Journal of Catalysis 333 (2016) 139–148
Fig. 1. SEM images of IRA-900 (A: the whole beads; B: the surface; C: the interior); HPB-TS-1AS (D: the whole beads, E: the surface and interior of a fractured bead; F: the
discrete crystals covering the surface; G: the interior); Pow-TS-1 (H); HPB-TS-1 (I: the whole beads; J: the surface; K: the interior of a fractured bead); regenerated HPB-TS-1
(
L: the whole beads; M: the interior of a fractured bead); TS-1 crystals (N).

W. Cheng et al. / Journal of Catalysis 333 (2016) 139–148
143
Fig. 2. TEM images of fragments of the HPB-TS-1 beads. Characteristic meso- and macropores are indicated by the arrows.
silanol groups. Moreover, the porous structure within the beads
can favor trapping of water within the material.
The importance of the programmed temperature profile
employed in the hydrothermal synthesis of HPB-TS-1 is demon-
strated by comparison with titanosilicate bead samples prepared
at a given, fixed temperature (HPB-TiSil-60, HPB-TiSil-100, HPB-
TiSil-170, where the number indicates the temperature in °C).
The XRD patterns of HPB-TiSil-60 and HPB-TiSil-100 only display
the broad band centered at 23° due to amorphous (titano)silicate,
indicating that these temperatures are too low to achieve crystal-
lization into a TS-1 structure (Fig. S3). On the other hand, the
XRD pattern of HPB-TiSil-170 shows the characteristic peaks for
MFI zeolites, though the signal is noisier than in HPB-TS-1AS
(Fig. S3). These results clearly reveal that temperature plays an
important role in the formation of TS-1 crystals, regardless of the
presence of resin beads, and that crystallization occurs both inside
and outside the beads at higher temperature, confirming the free
mass transport of titanosilicate species from the starting solution
into the beads via ion exchange. In the attempt to minimize the
observed formation of powder TS-1 material outside the beads,
the titanosilicate beads were prepared with a larger amount of
resin relative to the titanosilicate solution (solution-to-resin mass
ratios of 10 and 6.8). However, the crystallinity of the obtained
materials is inferior to that of HPB-TS-1AS, as shown by the higher
relative intensity of the broad band at 23° assigned to amorphous
species (Fig. S3). This implies that the chemistry of the resin itself,
such as active groups and counterions, influences the crystalliza-
tion process. These results underline the importance of careful tun-
ing of the synthesis parameters to obtain the desired bead
materials consisting of TS-1 nanocrystals.
Fig. 3. XRD patterns of HPB-TS-1AS (A), HPB-TS-1 (B), Pow-TS-1 (C), and TS-1 (D)
samples.
(
Pow-TS-1) that was obtained together with the beads (Fig. 3C). In
agreement with the XRD data, the FT-IR spectra of the beads
Fig. S2) display the characteristic peaks of TS-1 [35,36]. The spectra
(
of the bead materials also show a large and intense band at 3450–
À1
3
300 cm , attributed to silanol groups interacting with water
adsorbed in the pores of the bead [37]. The presence of adsorbed
water is confirmed by the water bending mode at about
À1
1
640 cm . These bands are significantly less intense in the FT-IR
spectrum of Pow-TS-1 (Fig. S2), indicating higher hydrophilicity
of the bead materials. This feature can be explained by the smaller
size of the TS-1 crystals in the bead materials, which implies a
relatively higher external surface and thus a higher population of
The textural properties of the prepared materials were
investigated by N physisorption (Fig. 4). The adsorption–desorption
2

1
44
W. Cheng et al. / Journal of Catalysis 333 (2016) 139–148
Fig. 4. Nitrogen adsorption–desorption isotherms of HPB-TS-1AS (A) and HPB-TS-1
3
(
B). The isotherm of A is offset by 150 cm /g along the vertical axis to facilitate its
visualization.
isotherms of the hierarchically porous titanosilicate beads with and
0
without acid treatment show a steep uptake of N
2
below p/p = 0.05,
which is characteristic of microporous materials prepared using
tetrapropylammonium hydroxide (TPAOH) as template [28]. In
addition to this uptake, the two samples display a defined hysteresis
0
loop above p/p = 0.7, showing the presence of meso- and/or macro-
pores connected through windows of smaller size, leading to the so-
called bottleneck effect [38].
Based on the Horvath–Kawazoe (HK) method, the sizes of the
micropores in HPB-TS-1AS and HPB-TS-1 are about 0.54 and
0
.55 nm (Fig. 5), which correspond well with the diameter of
micropores in TS-1 or in other MFI zeolites templated by TPAOH
28]. The BJH pore size distribution indicates that very large meso-
Fig. 5. Pore-size distributions of HPB-TS-1 (A and B) and HPB-TS-1AS (C and D).
[
pores and macropores are present in the beads (Fig. 5), mainly in
the range 20–100 nm. Although this information is only qualitative
due to the limitations of N physisorption analysis in the large
2
mesopore range, it is in agreement with the porosity observed by
SEM (vide supra) and with previous reports on titanosilicate beads
3
[
29]. The pore volumes were calculated to be 0.56 and 0.55 cm /g
for HPB-TS-1AS and HPB-TS-1, respectively. The specific surface
2
area of HPB-TS-1 (227 m /g) is lower than that of HPB-TS-1AS
(
A
B
2
387 m /g), probably due to the removal of the layer of TS-1 crys-
tals on the external surface of the beads via acid treatment (vide
supra).
The coordination of titanium species in zeolites and meso-
porous titanosilicates can be monitored based on the position of
the UV–vis absorption of the material [39]. Ideally, the titanium
species are incorporated into the framework of crystalline zeolite
TS-1 in tetrahedral coordination sites, and this is regarded as the
key factor in their catalytic activity and stability [40]. The UV–vis
spectrum of HPB-TS-1AS shows a broad band in the range
C
200
300
400
500
600
700
800
Wavelength (nm)
Fig. 6. UV–vis spectra of HPB-TS-1AS (A), HPB-TS-1 (B), and Pow-TS-1 (C).
2
15–320 nm (Fig. 6), ascribed to the coexistence of various types
of tetrahedral Ti sites, such as Ti(OSi) and Ti(OH)(OSi) , and a cer-
4
3
tain amount of higher-coordination Ti species, such as reversible
pentacoordinated Ti species generated by the interaction of Ti spe-
cies with moisture and polymerized hexacoordinated Ti species
restrictions within the resin bead template and/or to the types of
functional groups and counterions present in the resin. Although
HPB-TS-1AS contains extraframework Ti sites with higher coordina-
tion (penta, hexa), these are effectively removed by the treatment
with nitric acid [8], as shown by the narrow band at 210 nm in the
UV–vis spectrum of HPB-TS-1, with only a narrow shoulder at
higher wavelengths (Fig. 6). This result is fully in agreement with
the XRD data (vide supra), and confirms that the acid treatment
allows removing the amorphous species and generating a material
in which titanium is mainly in isolated tetrahedral sites, the
desired active sites for catalytic application.
[
39]. On the other hand, the UV–vis spectrum of the powder mate-
rial obtained together with HPB-TS-1AS, i.e., Pow-TS-1, exhibits a
narrow band centered at ca. 215–220 nm, which corresponds to
the characteristic signal of isolated tetrahedral Ti sites in TS-1 zeo-
lites [10], with a shoulder extending to higher wavelengths (Fig. 6).
These results indicate that the use of the resin beads as shape- and
structure-directing agents partially hinders the formation of the
crystalline TS-1 framework, possibly due to the pore space

W. Cheng et al. / Journal of Catalysis 333 (2016) 139–148
145
Based on the discussion of the synthesis and characterization
results, we propose a path for the formation of HPB-TS-1 by means
of templating with anion-exchange resin beads (Scheme 1). In the
alkaline synthesis medium, negatively charged titanosilicate oligo-
mers can be formed in solution and, once the beads are added, can
exchange with the anions of the resin beads. Subsequently, the
condensation and crystallization of the titanosilicate species occurs
in the resin–titanosilicate composites during the hydrothermal
treatment. These composites consist of two interconnected 3D net-
works: an organic part originating from the resin, and an inorganic
part consisting of the condensed titanosilicate matter [29]. If the
temperature of the hydrothermal step is low, the condensed
titanosilicate matter is amorphous, whereas the titanosilicate
domains gradually grow into TS-1 zeolite crystals if the hydrother-
mal treatment takes place at T P 165 °C. The crystallization
becomes more efficient if the temperature profile is carefully con-
trolled. Upon calcination, the organic resin bead is removed, thus
generating titanosilicate beads with a unique inner structure in
which aggregates of TS-1 zeolite crystals are interconnected by a
three-dimensional network of meso- and macropores, as proven
The first reaction in which HPB-TS-1 was tested as a heteroge-
neous catalyst is the hydroxylation of phenol (Table 1 and Fig. S5),
which is an important industrial reaction for the production of
dihydroxybenzenes and their derivatives [43]. This reaction is
widely employed in the evaluation of porous titanosilicate cata-
lysts because it is very sensitive to the features of the catalyst
[44,45]. For example, the well-known ordered mesoporous Ti-
MCM-41 is almost inactive for the hydroxylation of phenol. This
has been attributed to the lower oxidation ability of titanium spe-
cies in the amorphous mesoporous walls of Ti-MCM-41, compared
with the Ti sites in the crystalline framework of TS-1 [46]. In this
work, HPB-TS-1 beads displayed high catalytic turnover frequency
À1
(TOF = 49.7 h ), which is comparable to that obtained on TS-1
À1
nanocrystals (TOF = 52.5 h ), as summarized in Table 1. The minor
difference in activity can be ascribed to the relatively low surface
area of the HPB-TS-1 material. Also the selectivity of HPB-TS-1 is
very similar to that of TS-1, with catechol and hydroquinone being
the major products (Table 1). These results clearly indicate that the
Ti species in the beads have analogous catalytic properties to those
in TS-1 nanocrystals, as expected based on the presence of TS-1
crystals with tetrahedral Ti sites within HPB-TS-1, which were
evidenced by XRD and UV–vis characterization (vide supra). In
contrast, HPB-TS-1AS displayed significantly lower conversion than
HPB-TS-1 (Table 1). This is ascribed to the presence of
weakly active or inactive extraframework Ti-species with higher
coordination in HPB-TS-1AS, as proved by UV–vis spectroscopy.
Pow-TS-1 gave phenol conversion and product selectivity
analogous to those for the TS-1 sample (Table 1). This confirms
the similarity between these two materials evidenced by the
physicochemical characterization.
2
by SEM, TEM, N physisorption, and XRD. Compared with conven-
tional ordered mesoporous materials, the meso/macropore size of
the HPB-TS-1 sample is much larger and the size distribution is rel-
atively broad, which is potentially favorable for mass transport of
reactants and products during catalytic reactions.
The temperature program used in the synthesis of HPB-TS-1
proved to be vital for the formation of robust beads. This was
demonstrated by the tendency of the sample prepared under direct
hydrothermal treatment at 170 °C (HPB-TiSil-170) to disintegrate
upon calcination (Fig. S4).
These catalytic results indicate that the connected meso/macro-
pore system in HPB-TS-1 allows efficient mass transport of the
reactants to the TS-1 domains containing the titanium active sites.
This was confirmed by means of the Weisz–Prater criterion [47]
using the measured catalytic data. This analysis indicates that
the HPB-TS-1 catalyst does not suffer from diffusion limitations
in the meso/macroporous network present inside the beads (see
the Supplementary Information). It is concluded that our catalyst
combines the high activity of TS-1 nanocrystals with the
straightforward separation endowed by its bead format. Indeed,
3
.2. Catalytic activity
The titanosilicate material HPB-TS-1 was studied as a heteroge-
neous catalyst for two types of oxidation reactions: the hydroxyla-
tion of phenol and the epoxidation of alkenes. In both cases,
aqueous hydrogen peroxide was employed as the oxidant. This is
considered a very sustainable oxidant, being relatively inexpen-
sive, safer than molecular oxygen, and environmentally friendly,
as it generates H
2
O as by-product [41,42].
resin bead
TS-1 crystallizaꢀon
(hydrothermal
treatment)
ion-exchange
Cl-
removal of
the resin
(calcinaꢀon)
negaꢀvely charged
tanosilicate oligomers
ꢀ
network of meso-
and macropores
aggregates of TS-1 zeolite crystals
Scheme 1. Proposed steps leading to the formation of the hierarchically porous beads consisting of TS-1 zeolite crystals, via templating with anion-exchange resin beads.

1
46
W. Cheng et al. / Journal of Catalysis 333 (2016) 139–148
Table 1
Catalytic performance of the titanosilicate materials in the hydroxylation of phenol with aqueous H
2
O
2
:a
Si/Tib
TOF (h )c
À1
Conv. (%)d
Product selectivity (%)e
CAT
Catalyst
HQ
BQ
6.8
45.7
7.8
HPB-TS-1
HPB-TS-1AS
Pow-TS-1
TS-1
41
33
41
35
49.7
8.6
50.9
52.5
17.6
4.0
17.8
21.3
51.0
46.9
47.2
50.8
42.2
7.4
45.0
43.7
5.5
a
b
c
Reaction conditions: 15 mL of water, 0.47 g of phenol, H
Determined by ICP-MS.
Turnover frequency is defined as mol of phenol converted per mol of Ti atoms of the catalyst per hour, calculated based on the conversion after 30 min of reaction (see
2 2 2 2
O (35 wt.%) in molar ratio H O :phenol = 1:3, 50 mg of catalyst, 80 °C, 4 h.
Fig. S5 and Table S1).
d
Based on phenol; the maximum theoretical conversion achievable under the employed reaction conditions is 33%.
CAT = catechol, HQ = hydroquinone, and BQ = para-benzoquinone. The tar product is not included, and the product selectivity is calculated as CAT (or HQ or BQ)/(CAT
e
+
HQ + BQ).
2
5
0
5
0
5
0
important commercial products [48]. In our strategy, efficient mass
transport through the connected meso/macroporous network in
the beads in combination with the small size of the TS-1 domains
2
1
1
(
implying a larger fraction of active sites at the micropore mouth
and on the external surface of the crystallites) should lead to
enhanced epoxidation activity with substrates that suffer from dif-
fusion limitation in the micropores of TS-1 [16,49]. To evaluate
this, the epoxidation of cyclohexene was chosen as a test reaction.
Cyclohexene is frequently employed as a test substrate to evaluate
titanosilicate catalysts, because its bulkiness limits its diffusion in
the micropores of TS-1 [49]. Indeed, the TS-1 sample showed the
expected low conversion of cyclohexene (Table 2), with the activity
being mainly ascribed to the conversion on the external surface of
the TS-1 particles. Coherently, the catalytic behavior of Pow-TS-1 is
very similar to that of the TS-1 sample (Table 2). Notably, the con-
version of cyclohexene, the TOF value, and the epoxide yield over
HPB-TS-1 beads were much higher than those over TS-1 (Table 2
and Fig. S6), proving the anticipated accessibility of the active Ti
sites exposed on the meso/macropores in our catalyst (i.e., the Ti
sites at the micropore mouth or on the external surfaces of the
TS-1 nanocrystals).
0
1
2
3
4
5
Runs
Fig. 7. Recycling tests of HPB-TS-1 as catalyst for the hydroxylation of phenol with
aqueous H . Recycling approach: washing ( ) or calcination ( ).
2 2
O
The catalytic performance of HPB-TS-1 was also compared with
that of Ti-MCM-41 [29], which is known to display high activity in
the epoxidation of cyclohexene due to the sufficiently large size of
its ordered mesoporous channels [50]. Although HPB-TS-1 showed
relatively lower conversion than Ti-MCM-41, presumably due to its
lower specific surface area, the yield of cyclohexene oxide, i.e., the
target product of the epoxidation, was more than double with
HPB-TS-1 compared to Ti-MCM-41 (Table 2). This result stems
from the much higher selectivity toward the epoxide and lower
selectivity toward 1,2-cyclohexanediol over HPB-TS-1, which sug-
the beads deposited automatically at the bottom of the reaction
vial few seconds after the agitation was stopped. This allowed sep-
aration from the reaction solution without centrifugation or filtra-
tion, which is a very attractive feature for any practical application
of these catalysts. Importantly, SEM images of the used catalyst
show that neither the bead shape nor the TS-1 crystals within
HPB-TS-1 were affected during the catalytic test (Fig. 1L and M).
The reusability of the catalyst in consecutive runs was evaluated
following two different recycling approaches. If the protocol only
involved washing with ethanol, the phenol conversion decreased
in subsequent runs (Fig. 7), probably due to the deposition of reac-
tion residues (e.g., tar) in the micropores of the TS-1 domains. In
line with this hypothesis, when the regeneration of HPB-TS-1
was carried out via calcination at 550 °C for 4 h in air, most of
the original activity of the catalyst was recovered (Fig. 7).
2
gests that the hydrolysis of the epoxide product with H O was less
favorable on HPB-TS-1. Ti-MCM-41 is characterized by amorphous
walls that are intrinsically rich of surface silanols, which favor the
adsorption of water on the surface, thus promoting the hydrolysis
of the epoxide to the diol [29,51]. Although the FT-IR spectra
showed that the bead materials also contain silanol groups that
tend to adsorb water (vide supra), the much larger meso/macrop-
ores of HPB-TS-1 compared with the narrow mesopores character-
istic of Ti-MCM-41 are expected to allow easier diffusion of water
(formed during the reaction) from the active sites, thus decreasing
the chances of hydrolysis of the formed epoxide.
The second class of reactions that was selected for testing HPB-
TS-1 was the epoxidation of alkenes with aqueous hydrogen perox-
ide. This is an attractive reaction, because the obtained epoxides
are highly useful intermediates for the manufacture of a range of

W. Cheng et al. / Journal of Catalysis 333 (2016) 139–148
147
Table 2
Catalytic performance of the titanosilicate materials in the epoxidation of cyclohexene with aqueous H
2 2
O
:a
Catalyst
Si/Tib
TOF
Conv.
(%)d
Yield
(%)e
H
2
O
2
decomp.
Product selectivity (%)f
À1
)c
(
h
(%)
8.0
30.7
6.2
3.1
n.d.
CHO
CHD
ALLYLIC
HPB-TS-1
HPB-TS-1AS
Pow-TS-1
TS-1
41
33
41
35
59
23.3
23.4
7.7
15.9
22.5
4.4
4.5
28.8
4.7
5.4
1.3
1.7
2.6
29.3
24.1
30.3
37.9
8.9
37.5
48.6
59.2
54.8
79.7
33.2
27.3
10.5
7.3
8.0
n.d.
Ti-MCM-41g
h
h
11.4
a
b
c
Reaction conditions: 4.5 mL of MeCN, 4.5 mmol of cyclohexene, 2.25 mmol of H
Determined by ICP-MS.
Turnover frequency is defined as mol of cyclohexene converted per mol of Ti atoms of the catalyst per hour, calculated based on the conversion after 30 min of reaction
2
O
2
(35 wt.%), 60 mg of catalyst, 60 °C, 5 h.
(
see Fig. S6 and Table S1).
d
Based on cyclohexene; the maximum theoretical conversion achievable under the employed reaction conditions is 50%.
Yield of cyclohexene oxide.
CHO = cyclohexene oxide, CHD = 1,2-cyclohexanediol, ALLYLIC = products of the allylic oxidation, i.e., 2-cyclohexene-1-ol and 2-cyclohexene-1-one.
Results from Ref. [29]. Reaction conditions: 4.5 mL of MeCN, 4.5 mmol of cyclohexene, 2.25 mmol of H
Not determined.
e
f
g
2 2
O (50 wt.%), 60 mg of catalyst, 60 °C, 5 h.
h
hydrogen peroxide occurs on the former. This suggests that the
bead format of the catalysts may influence the selectivity in the
oxidation of cyclohexene.
Table 3
Catalytic performance of HPB-TS-1 in the epoxidation of various alkenes with
aqueous H
a
2 2
O .
Conv. (%)b Epoxide selectivity (%)
Finally, the potential of HPB-TS-1 in the epoxidation of bulky
substrates was further tested using other alkenes with larger
molecular size (Table 3). Although the catalyst gave lower conver-
sion than with cyclohexene, it is still active with these much bulk-
ier alkenes, indicating that a fraction of the titanium species are
accessible through the meso/macropores of HPB-TS-1. Because
the large alkenes cannot enter the micropores of TS-1, their
conversion most likely occurs on the external surface of the TS-1
nanocrystals. The selectivity toward the epoxide was very high
with all substrates (>90%), which is attributed to the higher
structural stability of these epoxides relative to the more strained
cyclohexene oxide [29].
Substrate
cis-Cyclooctene
15.0
3.3
91.0
98.5
97.0
Cyclododecenec
trans, trans, cis-1,5,9-Cyclododecatriene 15.6
a
Reaction conditions: 4.5 mL of CH
35 wt.%), 60 mg of catalyst, 60 °C, 5 h.
3 2 2
CN, 4.5 mmol of alkene, 2.25 mmol of H O
(
b
Based on the alkene: the maximum theoretical conversion achievable under the
employed reaction conditions is 50%.
c
Mixture of cis and trans.
The higher conversion of cyclohexene, though with lower epox-
ide selectivity, observed on HPB-TS-1AS compared to HPB-TS-1
(Table 2) is ascribed to the higher surface area of the former. This
result and its comparison with the opposite trend between the
two catalysts observed in the hydroxylation of phenol (vide supra)
suggest that some of the Ti species with higher coordination pre-
sent in HPB-TS-1AS contribute to the epoxidation activity [52],
whereas the activity in the hydroxylation of phenol stems virtually
only from isolated tetrahedral Ti sites in a crystalline framework.
4. Conclusions
Hierarchically porous TS-1 beads with a diameter of 0.2–
0.5 mm (HPB-TS-1), built with TS-1 nanocrystals connected
through meso/macropores, were prepared by using anion-
exchange resin beads as a hard template. These titanosilicate beads
proved to be active and versatile heterogeneous catalysts for oxi-
dation reactions of both small and bulky substrates using the green
oxidant hydrogen peroxide. The catalytic performance stems from
the combination of the advantages of zeolitic nanodomains with
those of a meso/macroporous structure within the beads. In the
hydroxylation of phenol, HPB-TS-1 displayed comparable conver-
sion and TOF value to those over TS-1, indicating the presence of
highly active Ti sites. In the epoxidation of a bulkier substrate as
cyclohexene, the catalytic conversion, TOF and epoxide yield over
HPB-TS-1 are much superior to those over TS-1, thus underlining
the importance of the small size of the TS-1 nanocrystals, which
are accessible through the meso/macropores network. Finally, an
important asset of HPB-TS-1 is its straightforward separation due
to the bead format, which allows recovering the catalyst without
centrifugation or filtration. This feature is very attractive for
possible industrial applications.
The undesired decomposition of hydrogen peroxide (2H
2 2
O ?
2
H
2
O + O ) is often observed as a side reaction in oxidations
2
catalyzed by titanosilicates. It has been generally accepted that
this disproportionation reaction occurs more readily on extra-
framework Ti sites [53]. Accordingly, a much higher degree of
decomposition was observed over HPB-TS-1AS (Table 2), which is
the only material among those presented here that displays a sig-
nificant population of Ti sites with higher coordination (vide
supra). It has been proposed that molecular oxygen is involved in
the homolytic radical pathways through which allylic oxidation
products are formed in the oxidation of cyclohexene [54,55].
2 2
Therefore, the occurrence of the disproportionation of H O can
also be correlated to the higher selectivity toward allylic oxidation
products observed over the bead catalysts (Table 2). However,
other factors play a role in the observed formation of these prod-
ucts over the bead catalysts, because the selectivity is similar over
HPB-TS-1 and HPB-TS-1AS
, but much less decomposition of

1
48
W. Cheng et al. / Journal of Catalysis 333 (2016) 139–148
[
[
24] K. Moller, B. Yilmaz, U. Muller, T. Beins, Chem. Mater. 23 (2011) 4301–4310.
25] L.F. Wang, Z. Zhang, C.Y. Yin, Z.C. Shan, F.S. Xiao, Microporous Mesoporous
Mater. 131 (2010) 58–67.
Acknowledgments
The authors acknowledge sponsoring in the framework of the
following research programs: Natural Science Foundation of China
[26] M. Choi, K. Na, J. Kim, Y. Sakamoto, O. Terasaki, R. Ryoo, Nature 461 (2009)
246–249.
[
27] L.H. Chen, X.Y. Li, G. Tian, Y. Li, J.C. Rooke, G.S. Zhu, S.L. Qiu, X.Y. Yang, B.L. Su,
Angew. Chem., Int. Ed. 50 (2011) 11156–11161.
(
51472062, 21303031, 21273057), Natural Science Foundation of
Heilongjiang Province of China (B201010), Fundamental Research
Funds for the Central Universities (HIT. IBRSEM. 201326), and
Talent Program for Innovation of Science and Technology of Harbin
[
28] W.H. Wang, G. Li, W.G. Li, L.P. Liu, Chem. Commun. 47 (2011) 3529–3531.
[29] K.F. Lin, O.I. Lebedev, G. Van Tendeloo, P.A. Jacobs, P.P. Pescarmona, Chem. Eur.
J. 16 (2010) 13509–13518.
[
30] Horv Aacute, G. Th, Eacute Za, K. Kawazoe, J. Chem. Eng. Jpn. 16 (1983) 470–
75.
[31] E.P. Barrett, L.G. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373–380.
(
2013RFQXJ004). Dr. Jun Yue (University of Groningen) is kindly
4
acknowledged for useful discussion.
[
[
32] D. Cani, P.P. Pescarmona, J. Catal. 311 (2014) 404–411.
33] L. Tosheva, V. Valtchev, J. Sterte, Microporous Mesoporous Mater. 35–6 (2000)
621–629.
Appendix A. Supplementary material
[
[
[
[
[
[
34] L.F. Wang, K.F. Lin, Y. Di, D.L. Zhang, C.J. Li, Q. Yang, C.Y. Yin, Z.H. Sun, D.Z. Jiang,
F.S. Xiao, Microporous Mesoporous Mater. 86 (2005) 81–88.
35] A. Zecchina, S. Bordiga, G. Spoto, L. Marchese, G. Petrini, G. Leofanti, M.
Padovan, J. Phys. Chem. 96 (1992) 4985–4990.
36] D. Scarano, A. Zecchina, S. Bordiga, F. Geobaldo, G. Spoto, G. Petrini, G. Leofanti,
M. Padovan, G. Tozzola, J. Chem. Soc., Faraday Trans. 89 (1993) 4123–4130.
37] D. Trong On, S.V. Nguyen, V. Hulea, E. Dumitriu, S. Kaliaguine, Microporous
Mesoporous Mater. 57 (2003) 169–180.
38] C. Danumah, S. Vaudreuil, L. Bonneviot, M. Bousmina, S. Giasson, S. Kaliaguine,
Microporous Mesoporous Mater. 44 (2001) 241–247.
39] E. Gianotti, C. Bisio, L. Marchese, M. Guidotti, N. Ravasio, R. Psaro, S. Coluccia, J.
Phys. Chem. C 111 (2007) 5083–5089.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2015.09.017.
References
[
[
[
[
1] L. Luo, C. Dai, A. Zhang, J. Wang, M. Liu, C. Song, X. Guo, RSC Adv. 5 (2015)
9509–29512.
2] W.B. Fan, R.G. Duan, T. Yokoi, P. Wu, Y. Kubota, T. Tatsumi, J. Am. Chem. Soc.
30 (2008) 10150–10164.
3] S. Kwon, N.M. Schweitzer, S. Park, P.C. Stair, R.Q. Snurr, J. Catal. 326 (2015)
07–115.
4] W.-S. Lee, M.C. Akatay, E.A. Stach, F.H. Ribeiro, W.N. Delgass, J. Catal. 313
2014) 104–112.
2
1
[
[
[
40] P. Ratnasamy, D. Srinivas, H. Knozinger, Adv. Catal. 48 (2004) 1–169.
41] R.A. Sheldon, Chem. Soc. Rev. 41 (2012) 1437–1451.
42] W. Lueangchaichaweng, N.R. Brooks, S. Fiorilli, E. Gobechiya, K.F. Lin, L. Li, S.
Parres-Esclapez, E. Javon, S. Bals, G. Van Tendeloo, J.A. Martens, C.E.A.
Kirschhock, P.A. Jacobs, P.P. Pescarmona, Angew. Chem., Int. Ed. 53 (2014)
1
(
[
[
[
[
[
5] L. Xu, J. Ding, Y. Yang, P. Wu, J. Catal. 309 (2014) 1–10.
6] A. Thangaraj, R. Kumar, P. Ratnasamy, J. Catal. 131 (1991) 294–297.
7] C.B. Khouw, C.B. Dartt, J.A. Labinger, M.E. Davis, J. Catal. 149 (1994) 195–205.
8] P. Wu, T. Tatsumi, T. Komatsu, T. Yashima, J. Catal. 202 (2001) 245–255.
9] A. Corma, P. Esteve, A. Martinez, J. Catal. 161 (1996) 11–19.
1585–1589.
[
[
[
[
43] F.S. Xiao, J.M. Sun, X.J. Meng, R.B. Yu, H.M. Yuan, J.N. Xu, T.Y. Song, D.Z. Jiang, R.
R. Xu, J. Catal. 199 (2001) 273–281.
44] U. Wilkenhoner, G. Langhendries, F. van Laar, G.V. Baron, D.W. Gammon, P.A.
Jacobs, E. van Steen, J. Catal. 203 (2001) 201–212.
45] X.Y. Yang, Y. Han, K.F. Lin, G. Tian, Y.F. Feng, X.J. Meng, Y. Di, Y.C. Du, Y.L. Zhang,
F.S. Xiao, Chem. Commun. 22 (2004) 2612–2613.
46] K.F. Lin, Z.H. Sun, L. Sen, D.Z. Jiang, F.S. Xiao, Microporous Mesoporous Mater.
[
10] V. Hulea, F. Fajula, J. Bousquet, J. Catal. 198 (2001) 179–186.
[
11] H.G. Peng, L. Xu, H.H. Wu, K. Zhang, P. Wu, Chem. Commun. 49 (2013) 2709–
2
711.
[
[
[
[
[
12] Y.S. Tao, H. Kanoh, L. Abrams, K. Kaneko, Chem. Rev. 106 (2006) 896–910.
13] K. Moller, T. Bein, Chem. Soc. Rev. 42 (2013) 3689–3707.
14] C.M.A. Parlett, K. Wilson, A.F. Lee, Chem. Soc. Rev. 42 (2013) 3876–3893.
15] C. Perego, R. Millini, Chem. Soc. Rev. 42 (2013) 3956–3976.
16] J. Perez-Ramirez, C.H. Christensen, K. Egeblad, C.H. Christensen, J.C. Groen,
Chem. Soc. Rev. 37 (2008) 2530–2542.
72 (2004) 193–201.
[
[
47] S.T. Oyama, X.M. Zhang, J.Q. Lu, Y.F. Gu, T. Fujitani, J. Catal. 257 (2008) 1–4.
48] C. Cativiela, J.M. Fraile, J.I. Garcia, J.A. Mayoral, J. Mol. Catal. A: Chem. 112
(
1996) 259–267.
[
49] I. Schmidt, A. Krogh, K. Wienberg, A. Carlsson, M. Brorson, C.J.H. Jacobsen,
Chem. Commun. 21 (2000) 2157–2158.
[17] N.D. Petkovich, A. Stein, Chem. Soc. Rev. 42 (2013) 3721–3739.
[18] Y.C. Tong, T.B. Zhao, F.Y. Li, Y. Wang, Chem. Mater. 18 (2006) 4218–4220.
[19] X.X. Wang, G. Li, W.H. Wang, C.Z. Jin, Y.Y. Chen, Microporous Mesoporous
Mater. 142 (2011) 494–502.
[
[
[
50] T. Blasco, A. Corma, M.T. Navarro, J.P. Pariente, J. Catal. 156 (1995) 65–74.
51] J. Bu, H.K. Rhee, Catal. Lett. 65 (2000) 141–145.
52] L. Xu, D.-D. Huang, C.-G. Li, X. Ji, S. Jin, Z. Feng, F. Xia, X. Li, F. Fan, C. Li, P. Wu,
Chem. Commun. 51 (2015) 9010–9013.
[
20] M. Reichinger, W. Schmidt, V.V. Narkhede, W.P. Zhang, H. Gies, W. Grunert,
Microporous Mesoporous Mater. 164 (2012) 21–31.
[
[
53] W. Fan, P. Wu, T. Tatsumi, J. Catal. 256 (2008) 62–73.
54] J.M. Fraile, J.I. García, J.A. Mayoral, E. Vispe, Appl. Catal., A: Gen. 245 (2003)
[21] D. Verboekend, J. Perez-Ramirez, ChemSusChem 7 (2014) 753–764.
[22] Y. Cheneviere, F. Chieux, V. Caps, A. Tuel, J. Catal. 269 (2010) 161–168.
[23] J. Song, D.S. Jensen, D.N. Hutchison, B. Turner, T. Wood, A. Dadson, M.A. Vail, M.
R. Linford, R.R. Vanfleet, R.C. Davis, Adv. Funct. Mater. 21 (2011) 1132–1139.
363–376.
[
55] E. Jorda, A. Tuel, R. Teissier, J. Kervennal, J. Catal. 175 (1998) 93–107.