Direct room-temperature synthesis of methyl-functionalized Ti-MCM-41 nanoparticles and their catalytic performance in epoxidation

Article abstract of DOI:10.1016/j.jcat.2009.01.013

Methyl-functionalized Ti-MCM-41 nanoparticles with a size of 80 to 160 nm (Me-Ti-MCM-41 NP) were directly prepared via a dilute solution route by the co-condensation of tetraethoxysilane and methylalkoxysilanes in sodium hydroxide medium at room temperature. The characterization results showed the existence of ordered hexagonal mesoporous structure and tetrahedral Ti species in the nanoparticles. In the epoxidation of cyclohexene with tert-butyl hydroperoxide and aqueous H₂O₂, Me-Ti-MCM-41 NP samples displayed higher turnover frequencies (TOFs) for cyclohexene and initial reaction rates compared to Ti-MCM-41 and methyl-functionalized Ti-MCM-41 with normal particle size and to non-functionalized Ti-MCM-41 nanoparticles. Simultaneously, a higher selectivity for cyclohexene epoxide was observed in the case of aqueous H₂O₂, suggesting that the hydrolysis of cyclohexene epoxide with water is reduced on Me-Ti-MCM-41 NP samples. The improved catalytic behavior of Me-Ti-MCM-41 NP is discussed both in terms of the nanosize and methylation of the surface of the catalyst particles. The regeneration of Me-Ti-MCM-41 NP with tert-butyl hydroperoxide solution was evaluated via washing and calcination approaches.

Full text of DOI:10.1016/j.jcat.2009.01.013

Journal of Catalysis 263 (2009) 75–82
Contents lists available at ScienceDirect
Journal of Catalysis
www.elsevier.com/locate/jcat
Direct room-temperature synthesis of methyl-functionalized Ti-MCM-41
nanoparticles and their catalytic performance in epoxidation
Kaifeng Lin a, Paolo P. Pescarmona a,∗, Kristof Houthoofd a, Duoduo Liang b, Gustaaf Van Tendeloo b, Pierre A. Jacobs a
a
COK, K.U. Leuven, Kasteelpark Arenberg 23-bus 2461, 3001 Heverlee, Belgium
EMAT, UA, Groenenborglaan 171, 2020 Antwerpen, Belgium
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 8 July 2008
Revised 20 January 2009
Accepted 25 January 2009
Available online 13 February 2009
Methyl-functionalized Ti-MCM-41 nanoparticles with
a size of 80 to 160 nm (Me-Ti-MCM-41 NP)
were directly prepared via a dilute solution route by the co-condensation of tetraethoxysilane and
methylalkoxysilanes in sodium hydroxide medium at room temperature. The characterization results
showed the existence of ordered hexagonal mesoporous structure and tetrahedral Ti species in the
nanoparticles. In the epoxidation of cyclohexene with tert-butyl hydroperoxide and aqueous H2O2, Me-
Ti-MCM-41 NP samples displayed higher turnover frequencies (TOFs) for cyclohexene and initial reaction
rates compared to Ti-MCM-41 and methyl-functionalized Ti-MCM-41 with normal particle size and to
non-functionalized Ti-MCM-41 nanoparticles. Simultaneously, a higher selectivity for cyclohexene epoxide
was observed in the case of aqueous H2O2, suggesting that the hydrolysis of cyclohexene epoxide with
water is reduced on Me-Ti-MCM-41 NP samples. The improved catalytic behavior of Me-Ti-MCM-41 NP
is discussed both in terms of the nanosize and methylation of the surface of the catalyst particles. The
regeneration of Me-Ti-MCM-41 NP with tert-butyl hydroperoxide solution was evaluated via washing and
calcination approaches.
Keywords:
Methyl-functionalized Ti-MCM-41
nanoparticles
Dilute solution route
Shorter mesoporous channels
Epoxidation of cyclohexene
tert-Butyl hydroperoxide
Aqueous H2O2
© 2009 Elsevier Inc. All rights reserved.
1. Introduction
cohols [20], plant oils [21], and the oxidation of organic sulfides
[22]. In the oxidation of small reactant molecules, however, they
Since the reported synthesis of microporous TS-1 [1], titanium-
substituted zeolites have been attracting great attention because
of their remarkable catalytic performance for selective oxidations
of various organic substrates [2–8]. However, they cannot effec-
tively catalyze conversion of bulky molecules, which have no ac-
cess to the active sites located inside the micropores (0.7 nm). The
discovery of ordered mesoporous titanosilicates with wider pore
sizes (2–10 nm) [9–17] and mesoporous titanium-containing ze-
olite [18] overcomes this limitation and offers an opportunity to
use titanosilicates as versatile catalysts in the oxidation of bulky
reactant molecules. The most important features of the ordered
mesoporous titanosilicates are the high surface area, which poten-
tially allows an efficient dispersion of active sites, and the large
and uniform pore diameters in the mesopore range, which favor
the diffusion of bulky molecules. MCM-41, presenting a hexagonal
array of unidirectional tubular pores, is one of the most studied
ordered mesoporous structures [19]. Titanium-containing MCM-
41 were reported for the first time in 1994 [9,10]. Since then
Ti-MCM-41 materials have been studied as catalysts for various
reactions, such as the epoxidation of olefins [10], unsaturated al-
show much lower catalytic activity than TS-1 and Ti-beta probably
due to the lower hydrophobicity that promotes water adsorption,
which poisons the catalytically active centers [11]. Ti-MCM-41 also
suffers from leaching of the titanium active species in the pres-
ence of H2O, which causes the gradual deactivation of the catalyst.
Therefore, an enhancement of the hydrophobicity is considered
important to improve the activity of Ti-MCM-41 and retard the Ti-
leaching in liquid phase oxidation. Moreover, water removal from
the catalyst surface would also reduce the hydrolysis of the epox-
ide product and, therefore, would be beneficial for the selectivity
of the reaction. An increased hydrophobicity of the catalyst surface
is also expected to favor the diffusion of the rather apolar alkene
substrates inside the pores of the catalyst.
Two different successful approaches have been reported to en-
hance the surface hydrophobicity of Ti-MCM-41 by introducing
organic groups into Ti-MCM-41 mesostructures. The first one in-
volves postsynthetic silylation of the surface [23–26] and the sec-
ond one is based on direct organic functionalization in one-step
synthesis [27–32]. Both methods result in a remarkable enhance-
ment of the catalytic activities in the epoxidation. However, in both
cases, the synthesis step in which the mesoporous titanosilicate
structure is formed includes a hydrothermal procedure that re-
quires high temperature and pressure with the risk of formation
Corresponding author. Fax: +32 16 321998.
*
E-mail address: paolo.pescarmona@biw.kuleuven.be (P.P. Pescarmona).
0021-9517/$ – see front matter © 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2009.01.013

76
K. Lin et al. / Journal of Catalysis 263 (2009) 75–82
of the anatase phase due to the high synthesis temperature. To
our knowledge, no successful approach to directly prepare active
organically-functionalized Ti-MCM-41 at room temperature in one
step has been reported so far, although this would be an advanta-
geous route compared with the hydrothermal procedure. A possi-
ble reason hindering this approach is the incomplete hydrolytic co-
condensation of the different silicon and titanium sources at room
temperature [32]. Additionally, it has been proposed that the large
particle size of mesoporous titanosilicates is another major cause
responsible for their reduced activity compared to Ti-substituted
zeolites [30]. We recently prepared Ti-MCM-41 nanoparticles at
room temperature, showing increased activity in the epoxidation of
cyclohexene due to a decrease of particle size of Ti-MCM-41 into
the nanoscale [33]. Therefore, the combination of introducing or-
ganic groups into Ti-MCM-41 mesostructures with the decrease in
their particle size into the nanoscale is expected to create a meso-
porous titanosilicate with improved catalytic performances.
In this article we describe the room-temperature synthesis of
methyl-functionalized Ti-MCM-41 nanoparticles with short meso-
porous channels and highly active titanium species. These mate-
rials were used to catalyze the epoxidation of cyclohexene with
tert-butyl hydroperoxide and aqueous H2O2, which are frequently
employed as test reactions for the catalytic evaluation of titanosil-
icate catalysts.
a diffuse-reflectance accessory in the 200–800 nm region. In gen-
eral, the samples were not treated before measurement and the
UV–visible spectra were recorded in air. Selected samples were
◦
recorded also after in situ treatment in a U tube for 12 h at 350 C
in N2. The isotherms of nitrogen were measured at liquid nitro-
gen temperature using a Micromeritics TriStar 3000. The pore-size
distribution was calculated using the Barrett–Joyner–Halenda (BJH)
model. Scanning electron microscopy (SEM) and transition electron
microscopy (TEM) images were taken on a Philips XL30 FEG appa-
ratus and a Philips CM 20 electron microscope, respectively. Si/Ti
molar ratios and the presence of sodium in every sample were
monitored by means of EDX analysis on a Philips XL30 FEG; every
sample was measured three times and the average of the results
was calculated. The 29Si MAS NMR spectra were recorded on a
Bruker AMX300 spectrometer (7.0 T). The spinning frequency of
the rotor was 5000 Hz. Tetramethylsilane (TMS) was used as shift
reference. FT-IR spectra of Me-Ti-MCM-41 NP before the first cat-
alytic run and of the recycled sample after mild calcination were
recorded on a Nicolet 730 FT-IR spectrometer using self-supported
wafers of 6.7 mg/cm2. The Si–CH3 group is recognized by a band
−1
at ca. 1260 cm
[23]. Thermogravimetric analysis (TGA) was car-
◦
ried out under N2 at a ramp of 10 C/min with a TGA Q500 of TA
instruments equipped with a high-throughput sampling platform
designed for 16 samples.
The hydrophobicity of the samples was evaluated from TGA
measurements of adsorbed water. The samples were pretreated in
a desiccator in the presence of a saturated NH4Cl solution for 72 h
to reach the maximum level of water adsorption. The number of
water molecules adsorbed on the surface of each sample was cal-
2. Experimental
2.1. Preparation of samples
◦
Ti-MCM-41 nanoparticles (Ti-MCM-41 NP) were prepared from
a dilute solution route as previously reported [33]. Methyl-func-
tionalized Ti-MCM-41 nanoparticles were directly synthesized from
the dilute solution route with a mixture of two silicon sources,
tetraethyl orthosilicate (TEOS) and methyltriethoxysilane (MTES).
First, 3.5 mL of 2 M NaOH aqueous solution was added to a stirred
mixture of 480 mL of distilled water and 1.0 g of cetyltrimethyl-
ammonium bromide (CTAB). Secondly, titanium(IV) isopropoxide
(TIP), TEOS and MTES were mixed into a solution at room tem-
perature and then slowly added to the above-mentioned solution
under stirring at 700–900 rpm, yielding a white gel with the
following molar composition: 1197 H2O/0.31 NaOH/0.125 CTAB/1
TEOS/xMTES/0.025 TIP, where x is 0.115 and 0.230. After stirring
for 2 h at ambient temperature, the resulting solids were filtered,
washed with 1500 mL of distilled water on a Büchner funnel, and
culated from the weight loss between 25 and 150 C measured by
TGA, using the following equation
ꢀm
1
W =
NA,
M
ABETmi
O
H2
where W is the number of adsorbed H2O molecules per nm2;
ꢀm is the weight loss in grams at a given temperature (g);
mi is the initial weight of the sample (g); MH
is the molar
2O
mass of water (18.0153 g mol−1); NA is the Avogadro constant
(6.022 × 1023 mol−1); ABET is the surface area (nm2 g−1) [34].
2.3. Catalytic reactions
The epoxidation of cyclohexene with tert-butyl hydroperoxide
(TBHP) was carried out in a glass vial under vigorous stirring. In a
typical run, 4.5 mmol of cyclohexene, TBHP (2.25 mmol, ∼5.5 M in
decane) and 30 mg of catalyst were mixed in the vial and heated
◦
dried at 60 C. To remove the surfactant, the methyl-functionalized
Ti-MCM-41 nanoparticles were treated with a solution of 1 M HCl
in diethyl ether (solid:liquid = 1g:100 mL) for 24 h at room tem-
perature. The solids were separated by centrifugation, washed with
◦
to the desired temperature (35, 45, or 60 C). After reaction, 4.5
◦
mL of tetrahydrofuran was added to the reaction solution. The
catalysts were separated by centrifugation and the products were
analyzed on an Interscience Finnigan Trace GC Ultra equipped with
a RTX-5 fused silica column (5 m, 0.1 mm). The TBHP efficiency
was calculated as the ratio between the moles of TBHP used to
form the epoxide and the moles of TBHP converted to tBuOH dur-
ing the reaction [TBHP efficiency = (molepoxide/molBuOH) × 100%].
When aqueous H2O2 was used as an oxidant, the epoxidation of
ethanol four times, and dried at 60 C. The methyl-functionalized
Ti-MCM-41 nanoparticles prepared by this route are denoted as
10% Me-Ti-MCM-41 NP and 19% Me-Ti-MCM-41 NP, where the per-
centage in the sample notation refers to the total amount of Me–Si
in the starting gels. On the other hand, removal of the organics
from the unfunctionalized Ti-MCM-41 samples was achieved by
calcination in air at 550 C for 6 h. All the samples were kept in
capped plastic bottles (in air).
For comparison, Ti-MCM-41 and 19% Me-Ti-MCM-41 with con-
ventional large particle size (designated as Ti-MCM-41 LP and 19%
Me-Ti-MCM-41) were respectively synthesized using previously re-
ported procedures [26,27,33].
◦
◦
cyclohexene was carried out at 60 C. 4.5 mmol of cyclohexene,
4.5 mL of acetonitrile, aqueous H2O2 (2.25 mmol, 50%) and 30 mg
of catalyst were mixed in the vial. Each methylated Ti-MCM-41 NP
catalyst was tested three times (each run on a fresh sample): aver-
ages of conversions and of turnover frequencies (TOF) are reported.
To prepare TBHP in dichloromethane (DCM), required amounts
of TBHP (70% in water) were extracted into DCM using a separation
funnel. The water layer was removed and the recovered organic
layer was dried over 0.4 nm molecular sieves. The prepared TBHP
in DCM was stored in a refrigerator until use.
2.2. Characterization
X-ray diffraction (XRD) patterns were obtained with a STOE
STADI P instrument using CuKα radiation. UV–visible spectra were
measured with a Varian Cary 5 spectrophotometer equipped with

K. Lin et al. / Journal of Catalysis 263 (2009) 75–82
77
For the recycling of the catalysts, the reaction solution was re-
moved from each sample and 5 mL of ethanol was added to each
vial. After stirring for 10 min, the solids were settled by centrifu-
gation. Then, the supernatant ethanol solution was removed. The
washing procedure was repeated 4 times. Finally, the samples were
which can be attributed to the higher amount of methyl groups
of the former. It is worth noting that the BET surface areas of
Ti-MCM-41 NP and Me-Ti-MCM-41 NP samples from the dilute
solution route are higher than of Ti-MCM-41 LP. This is mainly
attributed to the smaller particle sizes of Ti-MCM-41 NP and Me-
Ti-MCM-41 NP samples.
◦
dried in an oven at 60 C for 16 h.
3.2. SEM and TEM images and EDX elemental analysis
3. Results and discussion
The morphology and structure of Ti-MCM-41 NP and Me-Ti-
MCM-41 NP samples are clearly revealed by SEM and TEM images.
Ti-MCM-41 LP and 19% Me-Ti-MCM-41 consist of irregular agglom-
erated particles (Figs. 3A and 3B). In contrast, SEM images of Ti-
MCM-41 NP and 10 and 19% Me-Ti-MCM-41 NP samples from the
dilute solution route (Figs. 3C–3E) show nanosized spherulitic par-
ticles. The size of the particles ranges from 80 to 160 nm. TEM
images of Ti-MCM-41 NP [33] and 10 and 19% Me-Ti-MCM-41 NP
samples (Figs. 3F and 3G) show the existence of ordered hexagonal
arrays and one-dimensional mesoporous parallel channels within
these nanoparticles. The TEM results confirm that the Ti-MCM-41
nanoparticles are pure phases with short and ordered mesoporous
channels. Although the synthesis was carried out in a Na-rich so-
lution, negligible amounts of Na were incorporated in Ti-MCM-41
NP and Me-Ti-MCM-41 NP, as proven by EDX analysis.
3.1. X-ray diffraction and nitrogen adsorption/desorption isotherms
The XRD patterns of Ti-MCM-41 NP and Me-Ti-MCM-41 NP
samples exhibit reflections corresponding to the (100), (110) and
(200) planes, indicating an ordered 2d-hexagonal (p6mm) arrange-
ment of channels (Fig. 1). Comparing the values of d100 of as-
synthesized Ti-MCM-41 NP, 10% Me-Ti-MCM-41 NP, and 19% Me-
Ti-MCM-41 NP at 4.3, 4.2, and 4.0 nm, respectively, it follows that
the unit cell size slightly decreases as the number of methyl groups
increase.
N2 adsorption/desorption isotherms of calcined Ti-MCM-41 NP,
calcined Ti-MCM-41 LP, and Me-Ti-MCM-41 NP samples all give
typical type-IV isotherms with a sharp inflection at p/p0 > 0.3
(Fig. 2). This is characteristic of capillary condensation, which
points to the uniformity of the mesopore size distribution. Table 1
lists textural properties of these samples. Calcined Ti-MCM-41 NP
and Ti-MCM-41 LP present similar pore sizes and pore volumes.
Considering the functionalized materials, 19% Me-Ti-MCM-41 NP
exhibits slightly smaller pore size than 10% Me-Ti-MCM-41 NP,
In previous studies, it has been reported that the synthesis at
room temperature for 3 days without hydrothermal treatment re-
Fig. 2. N2 adsorption/desorption isotherms of Ti-MCM-41 LP (A), Ti-MCM-41 NP
(B), 10% Me-Ti-MCM-41 NP (C), and 19% Me-Ti-MCM-41 NP (D). For sake of clar-
ity, spectra B, C and D are offset along the vertical axis by 200, 300 and 400 cm3/g,
respectively.
Fig. 1. X-ray diffraction pattern of Ti-MCM-41 NP (A), 10% Me-Ti-MCM-41 NP (B)
and 19% Me-Ti-MCM-41 NP (C).
Table 1
Chemical composition, textural and structural properties.
d100
(nm)
Pore sizea
(nm)
Pore volumea
Average
particle size
(nm)e
Surface area
Methyl-group
contentd
(mol%)
SiOH contentd
(mol%)
Si/Tib
(cm3/g)
(m2/g)
Ti-MCM-41 LP
Ti-MCM-41 NP
3.8
3.5
4.2
4.0
–
2.1
2.0
2.3
2.0
–
0.54
0.58
0.62
0.61
–
–
847
1013
1065
1323
–
–
–
6.4
12.1
11.2
–
42.8
42.3
54.4
45.0
55.5
56.7
65.8
47.4
129.1
99.1
94.7
76.8
–
154
135
117
–
–
–
10% Me-Ti-MCM-41 NP
19% Me-Ti-MCM-41 NP
19% Me-Ti-MCM-41
Uncalcined Ti-MCM-41 NP
Ti-MCM-41-20Me-0c
4.3
–
–
–
–
–
–
–
8.9
570
a
Pore-size distributions and pore volumes derived from N2 adsorption isotherms at 77 K.
Molar ratios measured by EDX.
b
c
From Ref. [32].
d
e
Calculated from 29Si MAS NMR spectra.
Estimated from 50 randomly selected particles.

78
K. Lin et al. / Journal of Catalysis 263 (2009) 75–82
Fig. 3. SEM images of Ti-MCM-41 LP (A), 19% Me-Ti-MCM-41 (B), Ti-MCM-41 NP (C), 10% Me-Ti-MCM-41 NP (D), and 19% Me-Ti-MCM-41 NP (E) and TEM images of 10%
Me-Ti-MCM-41 NP (F) and 19% Me-Ti-MCM-41 NP (G).
sults in an extremely high Si/Ti ratio (570) in methyl-functionalized
Ti-MCM-41 [32]. It was proposed that the co-condensation of Ti
and Si sources did not proceed effectively at room temperature
and therefore Ti species might be washed out from the silica net-
work during the template-extraction procedure using aqueous HCl.
This is considered to be one of main reasons limiting the synthe-
sis of organically functionalized titanium-containing mesoporous
silicates at room temperature. However, Me-Ti-MCM-41 NP sam-
ples give much lower Si/Ti molar ratio (99.1 and 94.7) than in
the reported procedure [32], indicating that the dilute solution
route facilitates the inclusion of Ti species into the silica matrix by
the effective co-condensation of Ti source and Si sources at room
temperature. This can be explained in terms of the parameters in-
fluencing the kinetics of hydrolytic condensation reactions, such
as the nature and the concentration of the Ti and Si sources, the
solvent (if any), the amount of H2O and the pH of the synthesis
mixture [36]. The dilute solution route reported here presents a
lower pH (12 instead of 13) and a lower concentration of Si and Ti
sources compared to the literature method.
3.3. UV–visible spectroscopy
Fig. 4 shows the UV–vis spectra of calcined Ti-MCM-41 LP and
Ti-MCM-41 NP, 19% Me-Ti-MCM-41 and Me-Ti-MCM-41 NP sam-
ples. All the samples exhibit a band centered at 217–224 nm
attributed to distorted tetrahedral Ti species in mesoporous ti-
tanosilicates [32,35]. This indicates that Ti species are successfully
incorporated into the silica framework via the dilute solution route.
The spectra of Ti-MCM-41 LP and Ti-MCM-41 NP present a shoul-
der band at around 290 nm that can be attributed to reversible

K. Lin et al. / Journal of Catalysis 263 (2009) 75–82
79
Fig. 5. 29Si MAS NMR spectra of Ti-MCM-41 LP (A), uncalcined Ti-MCM-41 NP (B),
Ti-MCM-41 NP (C), 19% Me-Ti-MCM-41 (D), 10% Me-Ti-MCM-41 NP (E), and 19% Me-
Ti-MCM-41 NP (F).
MeSi(OSi)3 (T 3) and Me(HO)Si(OSi)2 (T 2). These results indicate
that methyl groups can be incorporated well into the wall of Ti-
MCM-41 NP by the direct synthesis from the dilute solution route.
The organosiloxane incorporation, viz. methyl-group content, is de-
termined by the following formula
100(T 2 + T 3)
ꢀ
%
i(Q i + T i)
and given in Table 1.
Normally, the R-group content of the final material is greatly
affected by hydrolysis rates of organosilanes at room tempera-
ture, since at low temperature the hydrolytic condensation rates
of organosilicon species are rather low [32]. Some researchers
observed that the hydrolytic condensation rates of alkylalkoxysi-
Fig. 4. UV–visible spectra of 10% Me-Ti-MCM-41 NP (A), 19% Me-Ti-MCM-41 NP
(B), 19% Me-Ti-MCM-41 (C), Ti-MCM-41 NP (D), and Ti-MCM-41 LP (E). For sake
of clarity, spectra B, C and E are offset along the vertical axis by 0.15, 0.3 and 0.1,
respectively.
−
lanes catalyzed by OH are lower than those of TEOS due to the
electron-donating ability of alkyl groups bonded to Si (especially
in the case of MTES) [32]. It has been reported that such a low
hydrolytic condensation rate of MTES causes a much lower Me-
content in methyl-functionalized Ti-MCM-41 than that in the start-
ing gels. For instance, Ti-MCM-41 materials functionalized with
methyl groups synthesized at room temperature were found to
have Me-content of 8.9 mol% (mol%-MTES of total Si) from start-
ing gels with a Me-content of 20 mol% [32]. The present Me-Ti-
MCM-41 NP samples have Me-contents of 6 and 12 mol% from a
Me-content of 10 and 19 mol% in the starting gels, respectively.
This result suggests that the dilute solution route is a more ef-
fective method to incorporate methyl groups into the framework
of Ti-MCM-41 by a direct organic functionalization at room tem-
perature. This result can be explained with arguments similar to
those used to discuss the improved incorporation of Ti species
into the nanoparticles (see above). The synthesis of Me-Ti-MCM-41
NP occurs via a complex mechanism implying the hydrolysis and
successive condensation of three different substances (TIP, TEOS,
and MTES). The synthesis conditions used in this work seem to
be more efficient than previously reported methods in promoting
a concerted condensation of the Si and Ti sources at room tem-
perature, leading to Ti-MCM-41 with higher Ti content and higher
degree of functionalization.
penta-coordinated Ti species forming from the interaction of Ti
species with moisture and/or to polymerized hexa-coordinated Ti
species [35]. Such a shoulder band is also visible but much less
pronounced in the spectra of methyl-functionalized Ti-MCM-41
samples, indicating that these methylated materials have a larger
◦
fraction of Ti in tetrahedral sites. After in situ treatment at 350 C
for 12 h under N2, the spectrum of anhydrous Ti-MCM-41 NP
shows only a small shift of the shoulder band from 290 to 280 nm,
indicating that this signal is due to irreversible species and, there-
fore, should be assigned to partially polymerized hexa-coordinated
Ti species. The absence of a band at 330 nm indicates that anatase
is absent in all samples.
3.4. 29Si MAS NMR spectra and TGA
Fig. 5 shows the 29Si MAS NMR spectra of various meso-
porous titanosilicates. Ti-MCM-41 LP and Ti-MCM-41 NP exhibit
three peaks at −110, −100, and −90 ppm, which are attributed
to (OSi)4 (Q 4), HOSi(OSi)3 (Q 3), and (HO)2Si(OSi)2 (Q 2), respec-
tively [26]. Methyl-functionalized Ti-MCM-41 nanoparticles show
two extra peaks at −63 and −56 ppm, which are assigned to

80
K. Lin et al. / Journal of Catalysis 263 (2009) 75–82
Table 2
Catalytic performances in the epoxidation of cyclohexene with TBHP solution over various catalysts.a
−1
on CH)
c
−1
d
Si/Ti
Conv.b
(%)
TBHP efficiency
(%)
TOF (h
)
(based
Selectivity for CHE
(%)
TOF (h
)
(based on CHE)
Ti-MCM-41 NP
Ti-MCM-41 LP
10% Me-Ti-MCM-41 NP
19% Me-Ti-MCM-41 NP
19% Me-Ti-MCM-41
129.1
47.4
99.1
94.7
76.8
16.3
13.0
31.7
38.5
43.7
89.5
79.7
90.7
98.2
> 99
38.3
11.4
57.3
66.6
61.5
88.2
69.9
94.3
94.1
95.8
33.8
8.0
54.0
62.7
58.9
a
◦
Reaction conditions: 4.5 mmol of cyclohexene, 2.25 mmol of TBHP (∼5.5 M in decane), 30 mg of catalyst, 60 C, 5 h.
b
c
Based on cyclohexene.
Turnover frequency is moles of cyclohexene converted per mole Ti active site of the catalyst per hour.
Turnover frequency is moles of cyclohexene epoxide formed per mole Ti active site of the catalyst per hour.
d
Table 3
Catalytic performances in the epoxidation of cyclohexene in different TBHP solutions
and solvents at different temperature over 10% Me-Ti-MCM-41 NP catalyst.a
◦
Temperature ( C) TBHP solutions
Solvent
–
Conv. (%)b Selectivity
for CHE (%)
60
60
60
35
35
45
∼5.5 M in decane
31.7
31.0
94.3
90.2
88.1
78.6
95.2
83.0
∼5.5 M in decane 3 mL of MeCN
∼5.5 M in decane 3 mL of decane 16.3
∼5.5 M in decane
∼3 M in DCM
–
–
–
14.9
34.6
17.6
∼5.5 M in decane
a
Reaction conditions: 4.5 mmol of cyclohexene, 2.25 mmol of TBHP, 30 mg of
catalyst, 5 h.
b
Based on cyclohexene.
of cyclohexene with TBHP (∼5.5 M in decane) as an oxidant. With
every catalyst tested, the main product in the epoxidation of cyclo-
hexene was cyclohexene epoxide (CHE), with up to 96% selectivity.
All the catalysts display high TBHP efficiency, with higher values
(%) for higher epoxide yields. Among the studied catalysts, 19%
Me-Ti-MCM-41 NP displayed the highest TOF values based on both
cyclohexene and cyclohexene epoxide and the highest initial re-
action rate calculated by the production of cyclohexene epoxide
after 30 minutes of reaction (Fig. 6). The improved catalytic re-
sults obtained with this mesoporous titanosilicate are attributed
to the combined advantages of incorporated methyl groups and
nanosized particles of Ti-MCM-41: the comparatively higher hy-
drophobicity and shorter length of the mesoporous channels result
in the enhanced diffusion of both cyclohexene and TBHP to the ac-
tive Ti-sites. Ti-MCM-41 NP and 19% Me-Ti-MCM-41 NP displayed
higher TOF values based on cyclohexene compared with the cor-
responding materials with larger particles, viz. Ti-MCM-41 LP and
19% Me-Ti-MCM-41. This is attributed to the decreased particle size
of Ti-MCM-41 into nanometer scale, allowing an enhanced acces-
sibility of the reactants to the catalytic Ti species in the shorter
channels of Ti-MCM-41 NP and 19% Me-Ti-MCM-41 NP, which is
in good agreement with previous work [33]. The improved diffu-
sion in the shorter pore channels may also prevent clogging of the
pores by heavy products that would cause deactivation of the cat-
alysts. All the methyl-functionalized Ti-MCM-41 samples showed
higher catalytic activities than the non-functionalized Ti-MCM-41
samples. This is ascribed not only to the improved hydrophobic-
ity stemming from the methylation of the surface, but also to the
larger fraction of Ti as catalytically active tetrahedral species in
the methyl-functionalized samples (see Section 3.3). However, the
difference in activity between 10% Me-Ti-MCM-41 NP and 19% Me-
Ti-MCM-41 NP is attributed only to the effect of the hydrophobicity
since the two methyl-functionalized samples present similar distri-
bution of Ti species as proved by UV–vis spectroscopy.
Fig. 6. Initial rate after 30 min of cyclohexene epoxidation based on cyclohexene
epoxide over various catalysts with TBHP (∼5.5 M in decane). Sample 1-5: Ti-MCM-
41 LP, Ti-MCM-41 NP, 10% Me-Ti-MCM-41 NP, 19% Me-Ti-MCM-41 NP and 19% Me-
Ti-MCM-41. For the reaction conditions, see Table 2.
The percentage of SiOH groups on all Si atoms (SiOH content)
derived from the 29Si MAS NMR analysis are listed in Table 1,
which are calculated by the following equation
100(2Q 2 + Q 3 + T 2)
ꢀ
%.
i(Q i + T i)
The SiOH content of the samples decreases in the following order:
uncalcined Ti-MCM-41 NP > 19% Me-Ti-MCM-41 > 10% Me-Ti-
MCM-41 NP > 19% Me-Ti-MCM-41 NP > Ti-MCM-41 LP > Ti-
MCM-41 NP. The lowest SiOH contents are observed in calcined
Ti-MCM-41 NP and Ti-MCM-41 LP samples due to the condensation
of SiOH groups upon calcination. For the other samples prepared
at room temperature without calcination, Me-Ti-MCM-41 NP sam-
ples show lower SiOH content than uncalcined Ti-MCM-41 NP due
to the incorporation of methyl groups in the framework.
The hydrophobicity of the Ti-MCM-41 NP samples was evalu-
ated by TGA from the amount of adsorbed water, which is calcu-
◦
lated from the weight loss of the samples between 25 and 150 C.
The number of adsorbed water molecules per nm2 of Ti-MCM-
41 NP, 10% Me-Ti-MCM-41 NP and 19% Me-Ti-MCM-41 NP is 3.1,
2.5 and 1.6, respectively, indicating that the hydrophobicity is en-
hanced by the incorporation of Me groups on the surface.
Table 3 lists the catalytic performances of 10% Me-Ti-MCM-
41 NP in different TBHP solutions at different temperatures
with/without co-solvents. When acetonitrile was used as a co-
3.5. Catalytic results
◦
Table 2 summarizes the catalytic activities of the inorganic and
organically functionalized Ti-MCM-41 catalysts in the epoxidation
C, the catalytic activity and selectivity for CHE
solvent at 60
were similar as those when no co-solvent was used. However,

K. Lin et al. / Journal of Catalysis 263 (2009) 75–82
81
Table 4
Catalytic performances in the epoxidation of cyclohexene with aqueous H2O2 over various catalystsa.
−1
c
−1
d
Si/Ti (molar)
Conv.
(%)b
TOF (h
)
Selectivity for CHE
(%)
TOF (h
)
(based on CH)
(based on CHE)
Ti-MCM-41 NP
Ti-MCM-41 LP
10% Me-Ti-MCM-41 NP
19% Me-Ti-MCM-41 NP
19% Me-Ti-MCM-41
129.1
47.4
99.1
94.7
76.8
4.9
5.7
8.4
8.9
8.5
115.0
50.0
151.9
153.8
119.5
12.2
10.3
13.4
14.0
11.0
14.0
5.2
20.3
21.5
13.1
a
◦
Reaction conditions: 4.5 mL of MeCN, 4.5 mmol of cyclohexene, 2.25 mmol of H2O2 (50 wt% in water), 30 mg of catalyst, 60 C, 0.5 h.
Based on cyclohexene.
Turnover frequency is moles of cyclohexene converted per mole Ti active site of the catalyst per hour.
Turnover frequency is moles of cyclohexene epoxide formed per mole Ti active site of the catalyst per hour.
b
c
d
the activity decreased when additional decane was used, sug-
gesting that a long linear alkane as decane can adsorb on the
methyl-functionalized catalyst surface, thus retarding the contact
of cyclohexene and TBHP with the titanium sites. To further eval-
uate the effect of solvents, a new TBHP solution (∼3 M in DCM)
was prepared and used as an oxidant. A much higher catalytic
◦
activity at 35 C was detected with the new kind of TBHP solu-
tion than with TBHP in decane solution. These results confirm the
detrimental effect of decane molecules on the catalytic activity in
such reaction.
When aqueous H2O2 was used as an oxidant, four different
products of the oxidation of cyclohexene were detected: cyclohex-
ene epoxide (CHE), 1,2-cyclohexanediol (CHD), 2-cyclohexene-1-ol
(CH–OH) and 2-cyclohexene-1-one (CH–ONE). It has been revealed
that the O–O bond of Ti-oxo species (Ti-peroxo, Ti-hydroperoxo or
Ti-superoxo) generated on titanosilicate molecular sieves by con-
tact with H2O2 cleaves either heterolytically or homolytically. The
heterolytic cleavage leads to the epoxide product (CHE), while the
homolytic cleavage leads to the allylic oxidation products (CH–OH
and CH–ONE) [37]. CHD is formed by catalytic hydrolysis of the
epoxide ring of cyclohexene oxide in the presence of acid sites.
Table 4 presents the catalytic performances of the inorganic and
organically functionalized Ti-MCM-41 catalysts in the epoxidation
of cyclohexene with aqueous H2O2 in acetonitrile. The conversion
of cyclohexene in the presence of H2O2 with no catalyst and with
Si-MCM-41 is 0.5 and 0.6%, respectively, indicating that the self-
oxidation and/or radical oxidation play a minor role in the activity
measured with the Ti-catalysts. Ti-MCM-41 NP and 19% Me-Ti-
MCM-41 NP displayed higher TOF value and selectivity for CHE
than Ti-MCM-41 LP and 19% Me-Ti-MCM-41, respectively. This re-
sult points to the advantages of the enhanced accessibility to the
catalytic Ti species for the reactants and of the shorter residence
time of CHE in the shorter channels of Ti-MCM-41 NP and 19%
Me-Ti-MCM-41 NP samples, in agreement with earlier reports [33].
When compared with Ti-MCM-41 NP, Me-Ti-MCM-41 NP samples
exhibited higher activities and selectivities for CHE. The higher
selectivities are ascribed to a reduced hydrolysis of cyclohexene
epoxide with H2O: the enhanced surface hydrophobicity generated
by the methylation (as estimated by TGA) favors the removal of
H2O formed from H2O2 during the epoxidation. The higher ac-
tivities are attributed to the improved surface hydrophobicity fa-
voring the access of cyclohexene to the active sites and reducing
the poisoning effect of water, as well as to the larger fraction
of Ti in the form of tetrahedral species in Me-Ti-MCM-41 NP, as
mentioned before. The slightly higher activity displayed by 19%
Me-Ti-MCM-41 NP compared with 10% Me-Ti-MCM-41 NP1 is as-
Fig. 7. Catalytic activity (2 and Q) and selectivity for CHE (" and a) of repeated
use of 10% Me-Ti-MCM-41 NP in the epoxidation of cyclohexene with TBHP (∼5.5 M
in decane) by washing (2 and ") and mild calcination (Q and a) approaches.
cribed uniquely to the increase in hydrophobicity with the degree
of methyl-functionalization, as discussed above for the epoxidation
with TBHP.
In order to evaluate the recyclability potential of the samples,
10% Me-Ti-MCM-41 NP was washed and reused in several 5-h cat-
alytic cycles with TBHP solution (∼5.5 M in decane). Fig. 7 shows
the results of the repeated use of Ti-MCM-41 NP after four room
temperature washings with ethanol (Runs 2 and 3) and calcina-
tion (Run 4). In the washing case, the cyclohexene conversion and
selectivity for CHE gradually decreased during subsequent runs,
probably to be attributed to the deposition of reaction residue
around the catalytically active sites on the surface and/or the
leaching of titanium during the reaction as observed for the classi-
cally prepared Ti-MCM-41. An alternative approach on regenerabil-
ity of 10% Me-Ti-MCM-41 NP was via intermediate mild calcination
◦
at 300 C for 2 h under N2. Under these conditions, the methyl
groups on the catalyst were mostly preserved, as indicated by the
detection of the peaks due to the Si–CH3 group [23] in the FT-IR
spectra of the material both before the first catalytic run and after
mild calcination. The almost complete reactivation of the used 10%
Me-Ti-MCM-41 NP was achieved by this calcination procedure (Run
4), giving comparable conversion as that in Run 1. It can be con-
cluded that the gradual decrease in cyclohexene conversion is due
to heavy product deposition because adsorbed reaction residues
that hinder the access of the reagents to the active centers would
be removed during the intermediate calcination. However, the se-
lectivity for CHE slightly decreases, which might be attributed to
1
Given the small difference in activity and selectivity between 10% Me-Ti-MCM-
41 NP and 19% Me-Ti-MCM-41 NP, each sample was tested three times (each run on
a fresh sample) and the averages of activities were reported. In each run, 19% Me-
Ti-MCM-41 NP displayed higher activity and selectivity than 10% Me-Ti-MCM-41 NP,
proving that the difference in the reported values is meaningful, although small.

82
K. Lin et al. / Journal of Catalysis 263 (2009) 75–82
the aggregation of Ti-MCM-41 NP particles and/or the change of Ti
coordinated environment from their original location during calci-
nation.
References
[1] M. Taramasso, G. Perego, B. Notari, US Patent 4410501, 1983.
[2] A. Tuel, Zeolites 15 (1995) 236.
[3] D.P. Serrano, H.X. Li, M.E. Davis, Chem. Commun. (1992) 745.
[4] A. Corma, M.A. Camblor, P. Esteve, A. Martínez, J. Pérez-Pariente, J. Catal. 145
(1994) 151.
4. Conclusions
[5] T. Tatsumi, N. Jappar, J. Phys. Chem. B 102 (1998) 7126.
[6] A. Corma, P. Esteve, A. Martínez, J. Catal. 161 (1996) 11.
[7] B. Notari, Adv. Catal. 41 (1996) 253.
[8] P. Ratnasamy, D. Srinivas, H. Knözinger, Adv. Catal. 48 (2004) 1.
[9] P.T. Tanev, M. Chibwe, T.J. Pinnavaia, Nature 368 (1994) 321.
[10] A. Corma, M.T. Navarro, J. Pérez-Pariente, Chem. Commun. (1994) 147.
[11] T. Blasco, A. Corma, M.T. Navarro, J. Pérez-Pariente, J. Catal. 156 (1995) 65.
[12] K.A. Koyano, T. Tatsumi, Chem. Commun. (1996) 145.
[13] O. Franke, J. Rathousky, G. Schulz-Ekloff, J. Starek, A. Zukal, Stud. Surf. Sci.
Catal. 84 (1994) 77.
[14] I.F. Vankelecom, N.M. Moens, K.A. Vercruysse, A.L. Karen, R.F. Parton, P.A. Ja-
cobs, Stud. Surf. Sci. Catal. 108 (1997) 437.
[15] D.T. On, M.P. Kapoor, P.N. Joshi, L. Bonneviot, S. Kaliaguine, Catal. Lett. 44 (1997)
171.
[16] T. Maschmeyer, F. Rey, G. Sankar, J.M. Thomas, Nature 378 (1995) 159.
[17] K.A. Vercruysse, D.M. Klingeleers, T. Colling, P.A. Jacobs, Stud. Surf. Sci.
Catal. 117 (1998) 469.
[18] I. Schmidt, A. Krogh, K. Wienberg, A. Carlsson, M. Brorson, C.J.H. Jacobsen,
Chem. Commun. (2000) 2157.
[19] P. Selvam, S.K. Bhatia, C.G. Sonwane, Ind. Eng. Chem. Res. 40 (2001) 3237.
[20] C. Berlini, M. Guidotti, G. Moretti, R. Psaro, N. Ravasio, Catal. Today 60 (2000)
219.
[21] L.A. Rios, P. Weckes, H. Schuster, W.F. Hoelderich, J. Catal. 232 (2005) 19.
[22] A. Corma, M. Iglesias, F. Sanchez, Catal. Lett. 39 (1996) 153.
[23] J. Bu, H.-K. Rhee, Catal. Lett. 65 (2000) 141.
[24] M. Guidotti, I. Batonneau-Gener, E. Gianotti, L. Marchese, S. Mignard, R. Psaro,
M. Sgobba, N. Ravasio, Microporous Mesoporous Mater. 111 (2008) 39.
[25] T. Tatsumi, K.A. Koyano, N. Igarashi, Chem. Commun. (1998) 325.
[26] K. Lin, L. Wang, F. Meng, Z. Sun, Q. Yang, Y. Cui, D. Jiang, F.-S. Xiao, J. Catal. 235
(2005) 423.
[27] A. Corma, J.L. Jordá, M.T. Navarro, F. Rey, Chem. Commun. (1998) 1899.
[28] N. Igarashi, S. Kidani, R. Ahemaito, T. Tatsumi, Stud. Surf. Sci. Catal. 129 (2000)
163.
Methylated Ti-MCM-41 nanoparticles (Me-Ti-MCM-41 NP) with
highly active titanium species were prepared at room temperature
from a dilute solution route. Characterization of these nanoparti-
cles indicates the formation of ordered hexagonal arrays and one-
dimensional mesoporous parallel channels and the incorporation
of the methyl groups. The dilute solution route reported here not
only is a successful method to prepare Ti-MCM-41 nanoparticles,
but is also a more effective method to incorporate methyl groups
into the framework of Ti-MCM-41 by a direct organic functional-
ization at room temperature.
In the epoxidation of cyclohexene with TBHP, 19% Me-Ti-MCM-
41 NP showed the highest TOF and initial reaction rate among
various mesoporous titanosilicates. When aqueous H2O2 was used
as the oxidant, 19% Me-Ti-MCM-41 NP was again the best catalyst,
displaying the highest TOF and selectivity based on cyclohexene
epoxide.
The improved catalytic performance of 19% Me-Ti-MCM-41 NP
is ascribed to a combined effect of its shorter mesoporous channels
and of the improved hydrophobicity generated by the incorpora-
tion of the methyl groups.
The gradual decrease in cyclohexene conversion in the regener-
ation of the methyl-functionalized catalysts via washing approach
is ascribed to heavy product deposition on the catalyst surface. The
other possible regeneration approach is an intermediate mild cal-
cination resulting in almost complete reactivation.
[29] N. Igarashi, S. Kidani, R. Ahemaito, T. Tatsumi, Zeolitic Mater. 1 (2001) 15.
[30] A. Bhaumik, T. Tatsumi, J. Catal. 189 (2000) 31.
Acknowledgments
[31] A. Bhaumik, T. Tatsumi, Catal. Lett. 66 (2000) 181.
[32] N. Igarashi, S. Kidani, R. Ahemaito, K. Hashimoto, T. Tatsumi, Microporous
Mesoporous Mater. 81 (2005) 97.
[33] K. Lin, P.P. Pescarmona, H. Vandepitte, D. Liang, G. Van Tendeloo, P.A. Jacobs,
J. Catal. 254 (2008) 64.
[34] R. Rinaldi, U. Schuchardt, J. Catal. 227 (2004) 109.
[35] E. Gianotti, C. Bisio, L. Marchese, M. Guidotti, N. Ravasio, R. Psaro, S. Coluccia,
J. Phys. Chem. C 111 (2007) 5083.
[36] P.P. Pescarmona, T. Maschmeyer, Aust. J. Chem. 54 (2001) 583.
[37] V.N. Shetti, D. Srinivas, P. Ratnasamy, Z. Phys. Chem. 219 (2005) 905.
K.L. is grateful for a Center of Excellence grant (CECAT) from
K.U. Leuven. P.P.P. acknowledges FWO for a postdoc grant. The au-
thors acknowledge sponsoring in the frame of the following re-
search programs: CECAT (K.U. Leuven, Flemish Government), SBO-
BIPOM (IWT), IAP-PAI (BELSPO, Federal Government), NoE IDECAT
(EU) and GOA (Flemish Government). We thank Gina Vanbutsele
for assistance in the measurement of the N2 isotherms.