Direct ring C–H bond activation to produce cresols from toluene and hydrogen peroxide catalyzed by framework titanium in TS-1

Article abstract of DOI:10.1016/j.jcat.2018.07.038

A series of TS-1 samples with varying Ti content was prepared and applied in the partial oxidation of toluene with H₂O₂ as oxidant. The TS-1 samples and relevant systems were characterized by XRD, FT-IR, in situ FT-IR, nitrogen adsorption/desorption, DR UV–vis, XPS, ICP-AES, EPR, SEM, TEM and TG. The results showed that the framework Ti species was responsible for the selective ring C–H bond activation, resulting in toluene hydroxylation; while the extra-framework Ti species was advantageous to the formation of side chain oxidation products and to the deep oxidation of both kinds of the products formed. Under the optimized reaction conditions, the toluene conversion reached 38.9% with 84.3% selectivity to cresols (TOF, 2.66 h^?1), while, the selectivity of para-cresol was 66.8%, which might be caused by the pore structure and steric hindrance of methyl group. The catalyst regenerated by calcination exhibited stable activity and selectivity after the second run.

Full text of DOI:10.1016/j.jcat.2018.07.038

Journal of Catalysis 366 (2018) 37–49
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Direct ring CAH bond activation to produce cresols from toluene and
hydrogen peroxide catalyzed by framework titanium in TS-1
⇑
⇑
Conglin Pang, Jiahui Xiong, Guiying Li , Changwei Hu
Key Laboratory of Green Chemistry and Technology, Ministry of Education, College of Chemistry, Sichuan University, Chengdu, Sichuan 610064, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of TS-1 samples with varying Ti content was prepared and applied in the partial oxidation of
toluene with H2O2 as oxidant. The TS-1 samples and relevant systems were characterized by XRD,
FT-IR, in situ FT-IR, nitrogen adsorption/desorption, DR UV–vis, XPS, ICP-AES, EPR, SEM, TEM and TG.
The results showed that the framework Ti species was responsible for the selective ring CAH bond
activation, resulting in toluene hydroxylation; while the extra-framework Ti species was advantageous
to the formation of side chain oxidation products and to the deep oxidation of both kinds of the products
formed. Under the optimized reaction conditions, the toluene conversion reached 38.9% with 84.3%
selectivity to cresols (TOF, 2.66 hꢀ1), while, the selectivity of para-cresol was 66.8%, which might be
caused by the pore structure and steric hindrance of methyl group. The catalyst regenerated by
calcination exhibited stable activity and selectivity after the second run.
Received 21 June 2018
Revised 28 July 2018
Accepted 30 July 2018
Keywords:
Toluene
Cresols
TS-1
Hydrogen peroxide
Ring-oxidation
Framework Ti
Ó 2018 Elsevier Inc. All rights reserved.
1. Introduction
mance. However, two types of Ti species exist in TS-1 zeolite, that
is, the [TiO4] species tetrahedrally coordinated in the framework
Selective oxidation of aromatic carbon-hydrogen (CAH) bonds
to form new CAC, CAO, CAN bonds by one step process has
attracted considerable attention for its wide application in industry
[1,2]. The investigation of the direct hydroxylation of arenes with
hydrogen peroxide under mild condition is mostly concentrated
to the preparation of phenol from benzene [3,4]. However, substi-
tuted benzene is quite different and interesting, because of the
challenge for getting high selectivity from competitive reactions
existed between ring sp2 CAH bond activation and side chain sp3
CAH activation [5,6], among them, ring CAH bond activation is
usually expected. Meanwhile, thanks to the directing groups, the
control of the region-selectivity in CAH activation was worth dis-
cussing [7].
Titanium silicalite-1 (TS-1) has received much attention and
found large potential application during the last decade for its
excellent catalytic activity in selective oxidation of organics using
aqueous H2O2 (30 wt%) as oxidant [8–10]. Typical reactions include
the oxidation of alkanes and alcohols [11], aromatic hydroxylation
[12–18], alkene epoxidation [19–22], as well as ketone ammoxi-
mation, etc [23,24]. It is well known that as the active component
in TS-1, titanium plays an important role for the catalytic perfor-
[25–29], and the octahedral anatase-like TiO2 species or amor-
phous Ti species in the extra-framework [27,30]. Therefore, it is
of paramount importance to study the influence of these two kinds
of titanium species on the reaction. Guang et al. [20] found that
among the two species, the framework Ti species was the active
center for selective oxidation of propylene. Feng et al. [31] also
thought that only isolated framework Ti(IV) species were responsi-
ble for the selective propene epoxidation with H2 and O2. Xia et al.
[32] noted that the framework Ti was the active sites for phenol
oxidation reaction, while the extra-framework Ti was unfavorable
for the reaction because of unwanted promotion of the decompo-
sition of H2O2.
Cresols are important intermediates for fine chemicals, which
are widely employed for preparing phenolic resins, insecticides,
herbicides and dyes [33–35]. Generally, worldwide industrial pro-
duction of cresols was carried out from toluene via a multistep
reaction processes including sulfonated alkali fusion [36], chlorina-
tion [35] or cymene process [33]. The multistep processes have
many disadvantages, such as, energy and time consuming with
harsh operation conditions, low atom efficiency, large amount of
by-products and serious environmental pollution [35,37]. Studies
on the hydroxylation of aromatic compounds with various metal-
modified catalysts have been reported in literatures [17,26,38].
The catalysts investigated in the direct oxidation of toluene are
⇑
Corresponding authors.
E-mail addresses: gchem@scu.edu.cn (G. Li), changweihu@scu.edu.cn (C. Hu).
https://doi.org/10.1016/j.jcat.2018.07.038
0021-9517/Ó 2018 Elsevier Inc. All rights reserved.

38
C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
mostly porous zeolite materials, such as SBA-15 [39–41], ZSM-5
[42–44], MCM-41 [45–47], activated carbon (AC) [48] and so on.
On this basis, the materials were usually modified by various tran-
sition metals such as Ti [45], Pt [39], Fe [48] and so on. Neverthe-
less, most of the products are side chain oxidation products. The
conventional impregnation method has the problems of energy
consuming and the catalysts obtained suffer from metal’s leaching,
which limited their application [49]. Among them, titanium-
modified materials (i.e, TS-1) have inestimable advantages, for
the Ti species predominantly in tetrahedral co-ordination could
be obtained by direct hydrothermal synthesis, getting rid of metal
leaching [50].
At present, some literatures on the influence of different tita-
nium environments in TS-1 catalyst for hydroxylation of aromatics
have been reported. Barbera et al. [4] evaluated the different tita-
nium species by UV–vis and IR spectroscopy, and reported that
the extra-framework Ti species did not contribute to benzene
hydroxylation. Luo et al. [3] described a TS-1 without extra-
framework Ti and got an excellent phenol yield of 39% with 72%
selectivity. Comparative study on the commercial TS-1 bearing
both framework Ti and extra-framework Ti, showed that frame-
work Ti species were responsible for the direct hydroxylation of
benzene to phenol, while the extra-framework TiO2 species caused
decrease of activity. However, the hydroxylation of toluene for ring
CAH activation was studied rarely in literatures [33]. Liu et al. pre-
pared TS-1/diatomite catalyst, which was treated with H2SO4/HF
mixed acid solution and calcined at 1000 °C, and then used for
toluene hydroxylation by H2O2 with acetone solvent in fixed-bed
reactor. This reaction got a high cresol selectivity of 97.1% with
the 14.3% conversion of toluene.
2.2. Characterization
2.2.1. X-ray diffraction (XRD)
XRD analyses of both fresh and used catalysts were carried out
on an LED DX-1000 CSC Diffractometer with a Cu K
matic X-ray radiation (k = 0.15406 nm), and the data were
collected over the 2 range of 5° to 35° with a step of 0.06°/min.
a monochro-
H
2.2.2. Fourier transform infrared spectroscopy (FT-IR)
Samples were diluted with KBr and pressed. The FT-IR data
were collected on a VERTEX 70 FT-IR spetrometer equipped with
an MCT detector in the region of 4000–400 cmꢀ1 with a resolution
of 4 cmꢀ1
.
In situ FT-IR spectra of the deactivated TS-1 catalyst were car-
ried out on a FT-IR spectrometer (Bruker VERTEX 70) with a MCT
detector cooled by liquid nitrogen. The instrument could obtain
the information over the frequency range of 4000–900 cmꢀ1, and
the spectra resolution was 4 cmꢀ1. Prior to the experiment, the
sample (about 20 mg) was pretreated at 100 °C under vacuum for
1 h to avoid the moisture. The background spectrum was collected
from 20 to 500 °C at a rate of 10°/min under Ar flowing rate of
30 mL/min, and the spectral acquisition was performed at the
same heating rate in a dry air atmosphere. The final spectra were
obtained by subtracting the spectrum at the same temperature
points in Ar flow.
2.2.3. Nitrogen adsorption/desorption
The Brunauer-Emmett-Teller (BET) surface area (SBET), the sin-
gle point adsorption total pore volume of pores (Vtotal), t-Plot
micropore volume (Vmic), mesopore volume (Vmeso) and the aver-
age pore diameter (Davg) of the catalysts were measured at ꢀ196
°C on a Micromeritics Tristar 3020 instrument. Peior to N2
physisorption, 0.10 g catalyst was pretreated at 150 °C for 2 h
and 300 °C for 2 h under vacuum.
However, some challenges still remains, for example, what kind
of titanium species playing a pivotal role in the hydroxylation of
substituted benzene such as toluene has not been reported. Thus,
in this work, various TS-1 samples were synthesized with different
Ti content, and applied for toluene partial oxidation. The roles of
different Ti species on toluene hydroxylation were researched,
and the reaction conditions for the hydroxylation of toluene were
optimized.
2.2.4. Ultraviolet–visible diffuse reflectance (DR UV–vis)
The DR UV–vis spectra were examined by a TU-1901 spectrom-
eter at the scan range from 200 to 800 nm with BaSO4 as the refer-
ence sample. All the UV–vis patterns of TS-1 were processed by the
software of Peakfit v4.12.
2. Experiment
2.1. Catalyst preparation
2.2.5. X-ray photoelectron spectroscopy (XPS)
XPS was used to provide information on the surface composi-
tion of the TS-1 with different Ti content, and it was performed
with an AXIS Ulttra DLD (KRATOS) spectrometer equipped with a
The TS-1 samples were synthesized following the procedures
reported in the literatures [12], which was a modified one of that
reported in Ref. [51]. The molar compositions of the components
were the following: 1.0 SiO2: 0.36TPAOH: xTiO2: 27H2O, among
them, TEOS (tetraethyl orthosilicate, 98%) and TiCl3 (titanium
trichloride, 15–20%) was used as SiO2 and TiO2 source, respectively.
The mixture was heated at 55 °C for 1 h and 85 °C for 8 h with con-
tinuous dripping of water to ensure the total volume invariant. The
resulted solution was cooled down to room temperature and kept
standing for about 12 h. After that, the mixed liquid was aged at
175 °C for 7 days in a 0.3 L autoclave under autogeneous pressure.
The crystallized solid was filtered, washed and dried at 110 °C.
Finally the sample was calcined at 550 °C for 10 h and the TS-1 cat-
alyst was obtained. The TS-1 with various controlled Ti contents
were denoted as TS-1a (x = 0, silicate-1), TS-1b (x = 0.89%), TS-1c
(x = 1.48%), TS-1d (x = 2.96%) and TS-1e (x = 5.92%).
monochromatic Al-Ka x-ray source (excitation energy = 1486.6
eV). Spectrum curve fitting was carried out using the Casa XPS
software.
2.2.6. Scanning electron microscopy (SEM)
The morphologies of all the catalysts were examined by SEM
(JSM-7500F, JEOL, Tokyo, Japan) with energy dispersive spec-
troscopy (EDX). The voltage was set at 5 KV. A thin film of gold
was applied to improve the conductivity of the samples.
2.2.7. Transmission electron microscopy (TEM)
TEM measurements of the TS-1 catalysts were carried out on
JEOL JEM-2100 electron microscope equipped with a field emission
source working at 200 KV.
The calcined TS-1a (S-1) sample was crushed and sieved, then
the particles were dipped in aqueous solution of active ingredient
TiCl3 (the molar ratio of titanium to silicon was the same as TS-1e),
then dried at 80 °C to remove the solvent, calcined at 550 °C to get
the Ti/S-1 catalyst.
2.2.8. Inductively coupled plasma-atomic emission spectrometry (ICP-
AES)
The catalyst samples were dissolved in HF solution in a Teflon
vessel for about 12 h, then the HF was evaporated and the solid
was dissolved in aqua regia to prepare samples for ICP-AES. The
ICP-AES measurement was performed to determine the actual Ti
The TiO2 sample (>99.0% CAS13463-67-7) was commercially
obtained from Kelong Chemical Reagent Factory, Chengdu, China.

C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
39
Selectiviry of H2O2ð%Þ
content on a VGPQExCell instrument (TIA, Boston, MA, USA). The Ti
content of fresh and spent TS-1 samples from ICP analyses were
listed in Tables 2 and 4, respectively.
moles of toluene ꢁ ½ðYcresols þ YBMÞ þ 3 ꢁ ðYHA þ YMQ Þꢂ
moles of initial H2O2 ꢁ CH2O2
¼
ꢁ 100;
2.2.9. Thermo-gravimetric Analysis (TG)
ꢀ
ꢁ
moles of cresols produced
moles of total Ti ꢁ reaction time
TOF hꢀ1
¼
:
TG was used to analyze the deposited carbon of the used cata-
lysts on a Thermo-gravimetric analysis apparatus (TG, NETZSCH TG
209F1). Approximately 10 mg sample was heated from 30 °C to
800 °C with the rate of 10 °C/min in dry air under the flowing rate
of 60 mL/min.
3. Results and discussion
3.1. Catalyst characterization
2.2.10. Electron Paramagnetic Resonance (EPR)
EPR spectra were recorded at ambient temperature on a Bruker
EMX plus spectrometer to probe the radical species formed in this
reaction system, the instrument parameters were as follows:
microwave frequency 9.87 GHz, microwave power 10 mW, modu-
lation frequency 100 kHz, scan time 2 min.
3.1.1. N2 adsorption-desorption isotherms
Fig. 1 displayed the N2 adsorption-desorption characterization
of the prepared TS-1 samples, which was basically characteristic
of TS-1 according to literature [52]. The controlled and actual Ti/
Si molar ratio corresponding to different TS-1 samples were listed
in Table 1, and their corresponding actual Ti contents (wt %) were
listed in Table 2. In the following description, all of the Ti/Si molar
ratio was referred to the actual Ti/Si molar ratio determined by
ICP-AES analysis. As shown in Table 1, when the actual Ti/Si molar
ratio on TS-1 was increased from 0 to 7.66%, the SBET first increased
until a maximum appeared for TS-1c (Ti/Si molar ratio = 1.75%),
then decreased with further increase of Ti concentration. The value
of total pore volume did not change with the increase of Ti/Si molar
ratio from 0 to 1.75%, and then decreased remarkably with Ti/Si
molar ratio from 1.75% to 7.66%, while the pore size decreased
monotonically with increasing titanium content. The pore size dis-
tribution curves showed that the pore size of the TS-1 samples
were concentrated at 15–25 Å, which was characteristic for meso-
porous materials. In addition, the pore size distribution tended to
shift to lower value with increasing titanium content. Contrary to
the tendency of pore size variation, the Vmic gradually increased
with the increase of titanium content. For the Nitrogen adsorp-
tion/desorption isotherms, TS-1d and TS-1e showed almost no dif-
ference between adsorption and desorption isotherms, which was
consistent with type II isotherms. However, for samples with low
titanium content, TS-1a, b and c, the type IV isotherms with H1
type hysteresis loop from P/P0 = 0.9 to 1.0 appeared, which sug-
gested a peculiar feature of mesoporous structure [53]. The results
were probably caused by the fact that moderate Ti might enter the
framework replacing the Si to make the molecular sieve structure
more regular, while excessive titanium might present as extra-
framework coated on the wall of the pores or cause meso-pore
blocking effect [54]. It was interesting to note that mesopores
2.3. Catalyst activity test
The partial oxidation of toluene was carried out in a single
necked flask reactor (50 mL) with reflux condenser at atmospheric
pressure in open air. In a typical run, designed amount of catalyst,
H2O, hydrogen peroxide (30 wt%) and toluene were loaded in the
flask in turn. Then the reaction was carried out with stirring when
the temperature of the reactor reached a constant. When the
designed reaction time was reached, the reaction was stopped
immediately and cooled down to room temperature. At last, the
products were qualitatively analyzed by GC-MS (Aglient, 5973
Nework6890N) and were quantified by HPLC (Aglient 1200)
equipped with a ZORBAX Eclipse XDB-C18 column and an UV
detector.
The yield and the selectivity of the products were calculated,
ignoring the polymerization products, ring-opening products and
tar, meanwhile the loss of volatility was taken into account in
the calculation [36,37]:
moles of products ðcresols; etc:Þ produced
Yieldð%Þ ¼
ꢁ 100;
moles of initial toluene
moles of cresols produced
moles of all products
Selectivity of cresolsð%Þ ¼
Conversion of H2O2ð%Þ ¼
ꢁ 100;
moles of consumed H2O2
moles of initial H2O2
ꢁ 100;
(B)
(A)
Ti O2
Ti/S-1
TS-1e
TS-1d
TiO2
Ti/S-1
TS-1c
TS-1e
TS-1d
TS-1c
TS-1b
TS-1b
TS-1a
TS-1a
0.0
0.2
0.4
0.6
0.8
1.0
10
20
30
40
50
60
70
80
90
100
Relative pressure (p/p0)
Pore Diameter (Å)
Fig. 1. (A) N2 adsorption-desorption isotherms of the samples; (B) The pore diameter distribution of the samples.

40
C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
Table 1
Physicochemical properties of the samples.
Sample
Ti/Si (%)a
Ti/Si (%)b
SBET (m2/g)
Vtotal (cm3/g)
Vmic (cm3/g)
Vmeso (cm3/g)
Davg (Å)
TS-1a
TS-1b
TS-1c
TS-1d
TS-1e
Ti/S-1
TiO2
–
–
383.04
419.55
463.54
399.48
388.18
399.68
10.65
0.36
0.35
0.34
0.24
0.25
0.32
0.04
0.13
0.14
0.16
0.16
0.17
0.13
–
0.23
0.21
0.18
0.08
0.08
0.19
0.04
27.29
25.24
23.11
20.19
19.62
31.79
47.79
0.89
1.48
2.96
5.92
5.92
ND
1.05
1.75
3.62
7.66
ND
ND
ND: not determined.
a
The controlled molar ratio (%) in the preparation experiment above.
The actual Ti/Si molar ratio (%) determined by ICP-AES analysis.
b
Table 2
The elemental analysis results of TS-1 samples with different Ti content.
Samples
ICP
XPS
I960/I800 (FT-IR)
Ti framework/Ti
(XPS)
total
Titotal (wt %)
Titotal (wt %)
TS-1a
TS-1b
TS-1c
TS-1d
TS-1e
–
–
–
–
0.83
1.36
2.75
5.54
0.10
0.19
0.44
0.53
0.10
0.19
0.20
0.18
1.00
1.00
0.45
0.34
volume (0.23 cm3/g) was observed for silicalite sample (TS-1a) in
comparison with the other samples, which might be caused by
extra-framework amorphous materials [55]. It could be speculated
that TS-1a might contain amorphous silica.
1225, 1100, 960, 800, 550 and 450 cmꢀ1) of the MFI type zeolite
structure, which was consistent with the XRD results. Among
them, a silanol group (SiAOH) peak at 930–970 cmꢀ1 could be
found for TS-1a, which indicated the presence of extra-
framework amorphous silica [58–60]. Combining the mesoporosity
in Table 1, we could conclude that silicalite-1 also contained extra-
framework Si species. The peak at 960 cmꢀ1 for TS-1b, c, d and e
was ascribed to the interaction between the stretching vibration
of [SiO4] units [61] and titanium ions in neighboring coordination
sites, which was an evidence of the vibration of SiAOATi bond in
the framework [17,19,62]. The peak at 800 cmꢀ1 was ascribed to
the symmetrical stretching vibration of SiAOASi bond in [SiO4]
[31,58]. The level of incorporation of Ti(IV) into the framework
was usually calculated by the band intensity ratio of I960/I800. The
higher the ratio, the more was the amount of framework Ti. The
ratio in Fig. 3(B). and Table 2 showed that the amount of frame-
work Ti had the following order TS-1a < TS-1b < TS-1c ꢃ TS-1d ꢃ
TS-1e. It could be observed that the amount of titanium
3.1.2. XRD
The XRD patterns of the catalysts with different Ti contents
were plotted in Fig. 2. The formation of TS-1 with MFI (Zeolite
Socony Mobil Five) structure was confirmed by the XRD character-
istic patterns, indicated by the peaks at 2
H of 7.9°, 8.8°, 23.1°,
23.9°, 24.4° [31,56]. The insertion of Ti in the framework of S-1
was indicated by the variation in Fig. 2 of the shape and size of
the peak at 24.4 2
the others [57].
H degree of sample TS-1a in comparison with
3.1.3. FT-IR
The FT-IR absorption spectra of TS-1 samples were shown in
Fig. 3(A). The TS-1 samples exhibited all the feature signals (i.e.
24.4°
TS-1e
TS-1d
TS-1c
TS-1b
22.5 23.0 23.5 24.0 24.5 25.0
TS-1a
2Θ (°)
5
10
15
20
25
30
35
2Θ (°)
Fig. 2. XRD patterns of TS-1 samples with different Ti content.

C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
41
(A)
TS-1e
TS-1d
(B)
TS-1e
TS-1d
TS-1c
TS-1b
TS-1a
TS-1c
960nm
800nm
TS-1b
TS-1a
1000
4000
3500
3000
2500
2000
1500
1000
500
1300
1200
1100
900
800
700
Wavenumber (cm-1)
Wavenumber (cm-1)
Fig. 3. The FT-IR spectra of TS-1 samples with different Ti content.
incorporated into the silicon framework has a threshold, over
which the excess titanium could not incorporate into the frame-
work, but existed as extra-framework titanium species. This result
again demonstrated that the decrease in SBET, Vtotal and Davg in the
BET characterization was resulted from the formation of
extra-framework titanium species.
3.1.4. DR UV–vis
Fig. 4 showed the UV–vis spectroscopy of TS-1 with different Ti
content. TS-1a did not exhibit any characteristic spectra because
no Ti was added. The bands at 213 nm and 313 nm corresponded
to framework Ti and extra-framework Ti respectively [32]. On
TS-1b and TS-1c, the bands around 213 nm, originated from Ti4+
-
O2ꢀ ? Ti3+Oꢀ of framework tetrahedral [TiO4] species, was
observed [63]. While for TS-1d and TS-1e, in addition to the peak
at 213 nm, a new peak appeared above 300 nm, which was attrib-
uted to extra-framework TiO2 [63]. Ti/S-1, which was obtained by
impregnation method, showed a huge peak near 313 nm with tiny
peak at 213 nm, and for TiO2, only an obvious peak at 313 nm was
found. The absence of the band above 300 nm for TS-1b and TS-1c
indicated that all of the titanium had incorporated into the frame-
work. The results of deconvolution were shown in Fig. 5(A). The
percentages of the two peaks were calculated and shown in
Fig. 5(B) to show the ratio of the area of deconvoluted peaks corre-
sponding to framework and extra-framework Ti species. It was
indicated that when the Ti content was less than 1.36 wt% (Table 2),
all titanium formed framework titanium species. When the tita-
nium content exceeded this value, the excess titania would gener-
ate extra-framework titanium and the amount of extra-framework
titanium species increased with increasing titanium content. That
is to say, extra-framework Ti species had been generated for
TS-1d and TS-1e. The results were consistent with those reported
by Millini et al. [57], where the insertion of increasing amount of
213nm 313nm
TiO2
Ti/S-1
TS-1e
TS-1d
TS-1c
TS-1b
TS-1a
200
300
400
500
600
700
800
Wavelength (nm)
Fig. 4. The DR UV–vis spectra of the samples.
120
(A)
313nm
213nm
(B)
100
80
60
40
20
0
TS-1b
TS-1c
600
200
300
400
500
600
700
800 200
300
400
500
700
800
TS-1d
600
TS-1e
600
200
300
400
500
700
800 200
300
400
500
700
800
TS-1c
TS-1d
TS-1b
TS-1e
Samples
Wavenumber (nm)
Wavenumber (nm)
Fig. 5. (A) DR UV–vis spectra of TS-1 samples with different Ti content after deconvolution; (B) Their corresponding quantified area ratio between framework and extra-
framework Ti.

42
C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
Ti in the MFI framework produced a linear variation of the unit cell
up to a concentration of Ti equal to 2.5 mol %, while the excess Ti
segregated as extra-framework TiO2 species.
3.2. Catalytic activity
3.2.1. The influence of Ti content on the activity
In the partial oxidation of toluene, the main products were cre-
sols, and the by-products were benzyl alcohol, benzoic acid, and
methyl-p-benzoquinone. Among them, cresols and BM was the
ring CAH bond activation and the side chain oxidation products
respectively. MQ and HA corresponded to the further deep oxida-
tion products of cresols and BM respectively. The catalytic activi-
ties of TS-1 with different Ti content and commercially available
titanium dioxide were shown in Table 3. The conversion of toluene
firstly increased with the increase of titanium content, and then
remained almost unchanged around 39%. The yield of cresols
increased from 8.7% (TS-1a) to 32.8% (TS-1c) firstly, and then
decreased to 22.3% (TS-1e) with increasing titanium loading. The
selectivity of the cresols showed the same tendency with the yield
and the maximum selectivity was 84.3% obtained on TS-1c. The
TOF was remained stable basically with low titanium content,
and then monotonically decreased with the increasing Ti loading,
where the optimal TOF was 2.66 hꢀ1 obtained on TS-1c, corre-
sponding to the highest cresols yield and selectivity. This might
be due to the fact that different titanium species had different con-
tributions to the activation of the ring and side chain CAH bonds.
That is, the framework titanium species might be responsible for
the hydroxylation of toluene to cresols by TiAOOH intermediate
[31]. The increased titanium content from TS-1a to TS-1c led to
increased framework active centers, resulting in the higher yield
of cresols. Then the excessive titanium produced extra-
framework titanium species which was unfavorable for the ring
CAH bond activation resulting in the reduced cresols yield for
TS-1d and TS-1e. Three possibilities existed, that is, the promoted
decomposition of H2O2 (Supplementary material, Fig. S1), side
chain oxidation and further deep oxidation of the cresols and side
chain oxidation products [33]. It could be observed that, the con-
version of toluene kept almost the same, while the selectivity of
cresols reduced significantly with the increase of extra-
framework Ti species, and the yield of side chain oxidation product
(BM and HA) increased obviously, indicating that the extra-
framework Ti was advantageous to the oxidation of side chain
under the conditions investigated in the present work. Further-
more, as shown in Table 3, the ratio of deep oxidation products,
that is, HA and MQ on TS-1d and TS-1e increased, indicating the
deep oxidation of both the products from the first step. To further
support this view, commercially available titanium dioxide had
also been used for this reaction, and 2% yield of side chain oxida-
tion products were observed due to the absence of framework Ti
species nor proper pore structure. The catalytic performance of
Ti/S-1 was also comparatively studied. The conversion of toluene
was only 3.7% with the 1.8% yield to side-chain products, and
1.7% ring oxidation products with the lower cresols selectivity of
35.1% was obtained for Ti/S-1. The fact that the additional Ti did
not result in enhancement of conversion of toluene on Ti/S-1 might
be due to its small micropore, which might have a significant
impact on catalytic performance for toluene oxidation. Addition-
ally, since Ti/S-1 was prepared by impregnation, the amount of Ti
species entered into the framework of S-1 might be limited, while
most of Ti species existed in extra-framework as revealed by DR
UV–vis, which might be the reason of low selectivity to cresols.
In addition, TS-1c demonstrated the highest catalytic performance,
due to the largest BET surface and most framework Ti species with
no extra-framework Ti species. The reason might be that the bigger
SBET was beneficial to the mass diffusion of regent together with
products, and can increase the utilization of active sites.
3.1.5. XPS
The XPS results of the Ti 2p3/2 peak were shown in Fig. 6. It can
be seen that only the peak at 460.0 eV was observed for TS-1b and
TS-1c, which was assigned to framework Ti species [26,30]. While
for TS-1d and TS-1e, besides the peak above, a new peak at 457.5
eV was found, which was assigned to extra-framework Ti [64,65].
The corresponding ratios of each species were illustrated in Table 2.
It was indicated that all of the Ti species in TS-1b and TS-1c were in
the framework with tetrahedral structure. In contrast, the addi-
tional extra-framework Ti was observed in TS-1d and TS-1e, and
increasing Ti content would lead to an increase in the percentage
of the extra-framework Ti. Overall, both UV–vis and XPS results
showed that with the introduction of Ti, it incorporated into the
framework firstly (TS-1a to TS-1c) with low loading, and when
the critical value of 1.36 wt% for total Ti content was reached,
the excess titanium generated extra framework TiO2 species
(TS-1d to TS-1e). This threshold is higher than that in some of
the literatures [20,26,66], which might be caused by different
preparation method used. In short, in our work, the TS-1c, TS-1d
and TS-1e actually contained almost the same amount of frame-
work titanium content, while the TS-1d and TS-1e samples also
had extra-framework titanium species.
3.1.6. TEM
To further study the structure of the catalysts, the TS-1
samples with different Ti content were characterized by TEM
technique, and the results were shown in Fig. 7. It was shown
that the TS-1a and TS-1b exhibited similar defined cube-like
morphologies with uniform size distributions. With the increase
of Ti content, the cube-shaped particles gradually decreased and
replaced by increasing oval grain gradually, which might be
constituted by agglomerate of smaller particles [67]. When the
titanium content exceeded 1.36 wt% (TS-1d and TS-1e), the petal
morphology was observed. What is more, it could be found that
the samples with titanium content below 2.75 wt% (TS-1a, b, c, d)
showed approximately the same crystals sizes of 100 nm roughly,
while the crystal diameter continued to increase to about 200 nm
for TS-1e.
457.5 eV 460.0 eV
TS-1e
TS-1d
TS-1c
TS-1b
TS-1a
454
456
458
460
462
464
466
468
Binding energy / eV
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.jcat.2018.07.038.
Fig. 6. The XPS results of Ti 2p3/2 spectra in TS-1 samples.

C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
43
60
50
40
30
20
10
0
50
100
150
200
250
Particle Size (nm)
70
60
50
40
30
20
10
0
50
100
150
200
250
Particle Size (nm)
70
60
50
40
30
20
10
0
50
100
150
200
250
Particle Size (nm)
50
40
30
20
10
0
50
100
150
200
250
Particle Size (nm)
50
40
30
20
10
0
50
100
150
200
250
Particle Size (nm)
Fig. 7. TEM of TS-1 samples with different Ti content: (A) TS-1a; (B) TS-1b; (C) TS-1c; (D) TS-1d; (E) TS-1e.

44
C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
Table 3
a
The effect of Ti content on the catalytic hydroxylation of toluene
.
Sample
Yield (mol%)
o-cresol
Ctoluene (%)
Ycresols (%)
Scresols (%)
CH
(%)d
SH
(%)
2 O2
TOF (hꢀ1
)
2O2
p-cresol
BMc
HAc
MQc
TS-1a
TS-1b
TS-1c
TS-1d
TS-1e
TiO2
1.5(12.7)b
4.0(16.7)
6.8(17.4)
6.2(15.8)
4.5(12.0)
–
7.2(61.0)
15.8(65.8)
26.0(66.8)
22.7(57.8)
17.8(47.5)
–
0.6
3.1
4.7
8.5
11.4
0.2
1.6
0.6
0.3
0.7
0.7
1.5
–
1.9
0.8
0.7
1.2
2.3
–
11.8
24.0
38.9
39.3
37.5
0.2
8.7
73.7
82.5
84.3
73.5
59.5
–
7.9
5.1
5.5
9.3
13.2
13.6
9.0
12.5
30.1
44.5
27.2
20.0
0.1
–
19.8
32.8
28.9
22.3
–
2.62
2.66
1.16
0.44
–
Ti/S-1
0.8(21.6)
0.5(16.1)
0.2
0.6
3.7
1.3
35.1
3.5
0.03
a
b
c
Reaction conditions: 0.5 g catalyst, 2.3 mmol toluene, 39.2 mmol H2O2, 30 mL H2O, 60 °C, 120 min.
The values in parentheses were the corresponding selectivity of o-cresol and p-cresol.
BM, HA, MQ represented benzyl alcohol, benzoic acid and methyl-p-benzoquinone, respectively.
Remained amount of H2O2 was determined by indirect iodimetry under acidic condition.
d
25
20
15
10
5
25
20
15
10
5
TS-1e (A)
TS-1c (A)
o-cresol
p-cresol
BM
HA
MQ
0
0
25
25
20
40
60
80
100
120
20
40
60
80
100
120
TS-1e (B)
TS-1c (B)
20
15
10
5
20
15
10
5
0
0
20
40
60
80
100
120
20
40
60
80
100
120
t (min)
t (min)
Fig. 8. Plots of products yield from toluene over TS-1c and TS-1e. (A) In the absence of isopropanol; (B) In the presence of isopropanol (2.6 mmol).
As described in related literatures, framework Ti in TS-1 inter-
acted with hydrogen peroxide giving rise to the formation of a
TiAOOH species that could perform selective oxidation of hydro-
carbons through an non-radical mechanism [68], while the extra-
framework Ti could promote extensive homolytic side reaction
including decomposition of hydrogen peroxide, which followed a
free-radical mechanism [57]. In order to clarify the reaction mech-
anism of toluene partial oxidation, the influence of isopropanol, a
known hydroxyl radical (ÅOH) scavenger [69], on the performance
of TS-1c and TS-1d was examined. The results were shown in
Fig. 8. It was indicated that hydroxyl radical species might play a
significant role in side-chain oxidation. Compared with the reac-
tions catalyzed by these two different kinds of TS-1 samples, TS-
1e was significantly affected by isopropanol, while TS-1c was not
very obvious. Detection of free radical was also carried by EPR to
further confirm the role of radical species in TS-1/H2O2/H2O/
toluene system. The reaction was performed under the conditions
of 0.5 g TS-1c or TS-1e catalysts, 2.3 mmol toluene, 39.2 mmol
H2O2, 30 mL H2O at room temperature. After 30 min, we pipetted
TS-1e
TS-1c
3470
3480
3490
3500
3510
3520
3530
3540
Magnetic Field (G)
Fig. 9. The EPR spectra of DMPO spin-trapping.

C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
45
50
40
30
20
10
MQ
(A)
(B)
HA
50
BM
p-cresol
o-cresol
40
30
20
10
0
0
50
40
30
20
10
0
0.1
0.3
0.5
0.8
1
60
90
120
180
300
50
m (g)
t (min)
(C)
(D)
40
30
20
10
0
13.1
19.6
39.2
58.8
78.4
30
40
50
60
70
H2O2 (mmol)
T (ºC)
Fig. 10. Optimization of reaction parameters: (A) The influence of the amount of TS-1c; (B) The influence of reaction time; (C) The influence of the H2O2 dosage; (D) The
influence of temperature.
0.1 mL the supernatant liquid and added 0.02 mL 5, 5-Dimethyl-
pyrroline-oxide (DMPO). The results were shown in Fig. 9. Clearly,
a quartet of signals with relative intensities of 1:2:2:1 from the
DMPO-OHÅ adducts were detected during the reaction process at
room temperature catalyzed by TS-1e, suggesting the presence of
homolysis of H2O2. However, this phenomenon was not observed
for the system catalyzed by TS-1c, with absence of extra-
framework Ti species. It appears clear that in toluene oxidation
with TS-1 and hydrogen peroxide the ring hydroxylation was cat-
alyzed by framework titanium-hydroperoxo sites which could not
be detected by DMPO spin-trapping, while the side chain oxidation
40
30
20
10
MQ
HA
BM
p-cresols
o-cresols
(A)
(B)
0
40
30
20
10
0
occurred through
framework Ti.
a radical mechanism promoted by extra-
Titanium effect was also observed in benzene hydroxylation,
Barbera et al. [4] found that the framework Ti species led to the
formation of Ti sites in tetrahedral coordination and TiAOOH
was the active intermediate which was responsible for hydroxyla-
tion of benzene to phenol. Similarly, the different Ti species was
also considered to play different roles in propylene epoxidation.
Xiong et al. [20] also proposed that TiAOOH was the active inter-
mediate for propylene epoxidation, and they thought that the
acid sites generated by extra-framework species led to further
oxidization of the propylene oxide. Thus, over the catalysts (TS-
1d and TS-1e) with higher extra-framework Ti species, the deep
oxidation was also more obvious. Combining the literatures
above, we thought that the partial oxidation of toluene was pro-
ceeded according to the following mechanism. Firstly, the frame-
work Ti was responsible for the ring CAH bond activation to
produce cresols by framework Ti-hydroperoxo sites, while the
side chain oxidation products (benzyl alcohol) occurred though
a radical mechanism promoted by extra-framework Ti species.
Then, the further deep oxidation of both ring CAH activation
products and side chain oxidation products further proceeded,
that is, the cresols were further oxidized to methyl-p-
benzoquinone, and the benzyl alcohol was further oxidized to
benzoic acid by extra-framework Ti.
1
2
3
4
5
Recycling number
Fig. 11. Activity test of the used catalyst: (A) The used catalysts were only dried; (B)
The used catalysts were dried and calcined.
3.2.2. Optimization of reaction parameters1
According to the above results obtained, TS-1c was selected to
optimize the reaction conditions, where the effect of the amount
of TS-1c, the reaction temperature, the amount of H2O2 and the
reaction time on the reaction was optimized. The results were
shown in Fig. 10. As could be seen, the highest yield of 32.8% to cre-
sols was obtained with the selectivity of 84.3%, under the opti-
mized conditions of 0.5 g TS-1c, 2.3 mmol toluene, 39.2 mmol
H2O2, 30 mL H2O, in 60 °C water-bath for 120 min. The stability
of the final observed products under the reactions was good, espe-
cially o-cresol, BM, HA and MQ (Supplementary material, Fig. S2).
Moreover, this reaction was clearly observed to show a high
1
Reaction condition: (A) 2.3 mmol toluene, 39.2 mmol H2O2, 30 mL H2O, 60 °C,
120 min; (B) 0.5 g TS-1c, 2.3 mmol toluene, 39.2 mmol H2O2, 30 mL H2O, 60 °C; (C)
0.5 g TS-1c, 2.3 mmol toluene, VH2O:VH2O2 = 15:2, 60 °C; 120 min; (D) 0.5 g TS-1c, 2.3
mmol toluene, 39.2 mmol H2O2, 30 mL H2O, 120 min.

46
C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
selectivity to para-isomer, which was considered to be characteris-
tic for process via electrophilic mechanism, being consistent with
the results in Ref [35,70]. Combining with steric hindrance of
methyl in toluene, the substrate was superior to the para-
position [14]. In addition, it was clear that the TS-1 catalysts pre-
pared by modified hydrothermal synthesis method exhibited obvi-
ous para-selectivity, while the Ti/S-1 obtained by impregnation
method showed the ortho-priority. This could be explained by
the pore size distribution. Obviously, the TS-1 samples had a smal-
ler average pore size of 19.62–27.29 Å, while the Ti/S-1 samples
showed the average pore size of 31.79 Å. In summary, the size of
para-position product might be more accessible to the smaller
channel and be in touch with the active titanium center in TS-1
samples. However, the ortho-hydroxylation occurred in the bigger
pores of Ti/S-1 samples. Furthermore, it was observed that the
products with larger molecular size, such as o-cresols (5.71 Å)
and methyl-p-benzoquinone (5.64 Å), were easier to generate in
larger pores. However, since the molecular dimensions of other
products were only 4.32 Å, it was dominant in smaller pores.
3.3. Recycle of the catalysts
The spent catalysts were reused to test the stability of the cat-
alysts. As can be seen in Fig. 11, when the used catalysts were only
dried at 100 °C overnight, both the conversion of toluene and the
yield of cresols showed a monotonic decreasing trend. However,
if the used catalysts were dried overnight and calcined at 550 °C
for 10 h, the conversion and yield remained essentially stable after
being reused for two runs, and 28.6% yield and 81.5% selectivity to
cresols were obtained under the optimal conditions with 0.5 g TS-
1c used for 5 runs. According to related reports in the literatures
[23,32,71], the deactivation of TS-1 was likely to be attributed to
following three reasons: dissolution of Ti species, migration of Ti
species and organic by-products filling in the micropores of TS-1
zeolite. The two former reasons caused irreversible deactivation
[30,32], while the latter could be regenerated by calcination or
washing with dilute hydrogen peroxide [72]. Firstly, based on
XRD (Fig. 12), no significant differences were observed when
TS-1c was used repeatedly for 5 times, indicating the fresh TS-1c
and deactivated sample had almost the same crystal and the
excellent MFI topological structure. Similarly, the catalysts were
characterized by DR UV–vis (Fig. 13) and XPS (Fig. 14), and both
fresh and deactivated catalysts showed the same Ti 2p3/2 spectra,
demonstrating that the titanium species did not migrate [32].
However, the spectra of C 1s showed obvious changes, for a new
band at 285.5 eV appeared in the deactivated TS-1c, in addition
to the graphltic carbon groups on every samples (peak at 284.5
eV), indicating that the carbon presented in alcohol, phenolic and
ether groups [33]. This might be caused by fouling by deposition
of heavy by-products. The surface area and pore volumes of these
samples were summarized in Table 4. It could be seen that the
specific surface area and the micropore volumes decreased
gradually with the increase of reuse times, indicating the channels
of the spent TS-1 catalysts were blocked by the above species.
Supplementary as shown in Fig. 18, if the deactivated TS-1c sample
(used 5 runs) was calcined at 550 °C, the SBET and Vmic could be
restored, which was consistent with the literature [32,72].
TS-1c
(used 5 runs)
TS-1c
(used 5 runs with calcination)
TS-1c
5
10
15
20
2Θ(°)
25
30
35
Fig. 12. XRD patterns of fresh and deactivated TS-1c sample.
TS-1c (used 5 runs)
TS-1c (used 5 runs with calcination)
TS-1c
200
300
400
500
600
700
800
Wavenumber (nm)
Fig. 13. DR UV–vis patterns of fresh and deactivated TS-1c sample.
284.5eV
(B)
(A)
285.5eV
TS-1c (used 5 runs)
TS-1c
(used 5 runs )
TS-1c
(used 5 runs with calcination)
TS-1c
(used 5 runs with calcination)
TS-1c
TS-1c
276
279
282
285
288
291
294
454
456
458
460
462
464
466
468
Binding energy / eV
Binding energy / eV
Fig. 14. The XPS patterns of fresh and deactivated TS-1c samples: (A) Ti 2p3/2 spectra of the samples; (B) C 1 s spectra of the samples.

C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
47
Table 4
The Physical property for fresh and used TS-1c samples.
Sample
SBET (m2/g)
Vtotal (cm3/g)
Vmic (cm3/g)
Vmeso (cm3/g)
Davg (Å)
TGa (wt%)
EDSb (wt%)
ICPc (wt%)
TS-1c
TS-1c (used once)
463.54
303.80
291.04
286.49
247.21
253.65
445.04
0.34
0.28
0.30
0.28
0.29
0.29
0.34
0.16
0.12
0.10
0.09
0.10
0.10
0.17
0.18
0.16
0.20
0.19
0.19
0.19
0.17
23.11
25.02
28.42
25.67
29.04
28.20
24.15
0.75
4.06
5.92
6.35
6.77
6.97
1.66
–
1.36
1.33
1.31
1.31
1.31
1.31
1.31
7.46
7.63
8.67
9.08
15.47
–
TS-1c (used 2 runs)
TS-1c (used 3 runs)
TS-1c (used 4 runs)
TS-1c (used 5 runs)
TS-1c (used 5 runs with calcination)
a
b
c
The weight loss rate of the samples in TG analysis.
The carbon content of the spent TS-1c samples measured by SEM-EDS.
The titanium content analyzed by ICP-AES.
To confirm the cause of catalyst deactivation by carbon deposi-
3600-3100cm -1
3000cm-1
tion, TG-DTG curves (Fig. 15) of the fresh and used TS-1 were car-
ried out. It was seen that the weight loss was increased with the
used times (Table 4). Nevertheless, the calcined used TS-1c was
similar to fresh one for almost no weight loss, which indicated
an easily firing species was formed over the deactivated catalyst.
TG-DTG curves of the fresh TS-1 catalysts were similar to the
regenerated one, which only had one stage of weight loss over
the low temperature range of about 100–200 °C. However, there
were two distinct stages in the deactivated catalysts, the first stage
of weight loss at the range of 100–250 °C was attributed to desorp-
tion of H2O and the organic by-products on the catalyst surface,
and the other at the range of 300–600 °C was attributed to the
oligomeric compounds inside the channel of the catalyst [72,73].
In order to investigate the species of the deposition on the deac-
tivated catalysts, the TS-1c (used 5 runs) was characterized by
in situ FT-IR, and the results were shown in Fig. 16. With increasing
temperature in dry air, obvious reduction had been observed for
the following peaks: the band at 3000 and 1680 cmꢀ1 attributed
to aromatic C@C stretching vibration [74], the peak at 1370 cmꢀ1
to methyl (CH3), the bands over the range of 3600–3100 cmꢀ1
and at 1200 cmꢀ1 to the stretching vibration of OAH bond [75],
the bands at 1900–1680 cmꢀ1 to C@O, and the region of 1850
cmꢀ1 and 1740 cmꢀ1 summarized as acid anhydride. These
changes might be caused by the removal of polymer species which
were adsorbed in the pores. This result was coincident with the
TG-DTG curves above.
-1 1200cm-1
1680cm
1850cm-1
1740cm-1
1370cm-1
4000
3600
3200
2800
2400
2000
1600
1200
800
Wavenumber (cm-1)
Fig. 16. The in situ FT-IR spectra of the deactivated TS-1c sample.
Ti
standard
O
C
Ti
Fig. 17 showed the EDS analysis of the used TS-1, and the car-
bon content was listed in Table 4. It indicated that with the
increasing of used times, higher carbon content has been found.
TS-1c (used 5 runs with calcination)
TS-1c (used 5 runs)
TS-1c (used 4 runs)
TS-1c (used 3 runs)
TS-1c (used 2 runs)
TS-1c (used once)
0.10
100
95
90
85
80
TS-1c (uesd 5 runs)
0.05
0.00
TS-1c
-0.05
-0.10
0
1
2
3
4
5
keV
100
95
90
85
80
0.09
TS-1c (uesd 5 runs with calcination) 0.06
Fig. 17. The SEM-EDS analyses of used TS-1 samples.
0.03
0.00
-0.03
0.12
All of these evidences showed that the deactivation was caused
by carbon deposition. Interestingly, the catalyst performance of
the calcined catalyst was not recovered completely during the first
two repeated runs. To explore the reason, the filtrate after reaction
was analyzed by ICP characterization, and the amounts of leached
Ti in the first two runs were 2.2 wt% and 1.5 wt% respectively. That
is, leaching of the framework titanium also contributed to the
deactivation of the TS-1 samples for the first two runs, which
was irreversible. After two runs, the catalyst became stable.
100
95
90
85
80
0.08
0.04
0.00
-0.04
TS-1c
100
200
300
400
500
600
700
800
Fig. 15. The TG-DTG characterization of the used TS-1 samples.

48
C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
(A)
(B)
TS-1c (used 5 runs with calcination)
TS-1c(used 5 runs with calcination)
TS-1c (used 5 runs)
TS-1c(used 5 runs)
TS-1c(used 4 runs)
TS-1c(used 3 runs)
TS-1c (used 4 runs)
TS-1c (used 3 runs)
TS-1c (used 2 runs)
TS-1c(used 2 runs)
TS-1c(used once)
TS-1c (used once)
TS-1c
TS-1c
0.0
0.2
0.4
0.6
0.8
1.0
10
20
30
40
50
60
70
80
90
100
Relative pressure (p/p )
Pore Diameter (Å)
0
Fig. 18. (A) N2 adsorption-desorption isotherms of the used TS-1; (B) The pore diameter distribution of the used TS-1 catalysts.
[7] S. Bag, R. Jayarajan, R. Mondal, D. Maiti, Angew. Chem. Int. Ed. 56 (2017) 3182–
3186.
4. Conclusions
[8] C.W. Yoon, K.F. Hirsekorn, M.L. Neidig, X. Yang, T.D. Tilley, ACS Catal. 1 (2011)
1665–1678.
[9] G. Bellussi, R. Millini, P. Pollesel, C. Perego, New J. Chem. 40 (2016) 4061–4077.
[10] C. Perego, R. Millini, Chem. Soc. Rev. 42 (2013) 3956–3976.
[11] W. Fan, R.-G. Duan, T. Yokoi, P. Wu, Y. Kubota, T. Tatsumi, J. Am. Chem. Soc. 130
(2008) 10150–10164.
[12] T. Yu, Q. Zhang, S. Xia, G. Li, C. Hu, Catal. Sci. Technol. 4 (2014) 639–647.
[13] B. Guo, Q. Zhang, G. Li, J. Yao, C. Hu, Green Chem. 14 (2012) 1880.
[14] C. Xia, L. Long, B. Zhu, M. Lin, X. Shu, Catal. Commun. 80 (2016) 49–52.
[15] X. Wang, Y. Guo, X. Zhang, Y. Wang, H. Liu, J. Wang, J. Qiu, K.L. Yeung, Catal.
Today 156 (2010) 288–294.
[16] S.-T. Tsai, P.-Y. Chao, T.-C. Tsai, I. Wang, X. Liu, X.-W. Guo, Catal. Today 148
(2009) 174–178.
[17] X. Wang, X. Zhang, Y. Wang, H. Liu, J. Qiu, J. Wang, W. Han, K.L. Yeung, Chem.
Mater. 23 (2011) 4469–4479.
[18] S. Xia, T. Yu, H. Liu, G. Li, C. Hu, Catal. Sci. Technol. 4 (2014) 3108–3119.
[19] M. Liu, H. Wei, B. Li, L. Song, S. Zhao, C. Niu, C. Jia, X. Wang, Y. Wen, Chem. Eng.
J. 331 (2018) 194–202.
[20] G. Xiong, Y. Cao, Z. Guo, Q. Jia, F. Tian, L. Liu, Phys. Chem. Chem. Phys. 18 (2016)
190–196.
[21] W.-S. Lee, M. Cem Akatay, E.A. Stach, F.H. Ribeiro, W. Nicholas Delgass, J. Catal.
287 (2012) 178–189.
[22] L. Wang, G. Xiong, J. Su, P. Li, H. Guo, J. Phys. Chem. C 116 (2012) 9122–9131.
[23] Y. Wang, S. Zhang, Y. Zhao, M. Lin, J. Mol. Catal. A: Chem. 385 (2014) 1–6.
[24] A. Zheng, C. Xia, Y. Xiang, M. Xin, B. Zhu, M. Lin, G. Xu, X. Shu, Catal. Commun.
45 (2014) 34–38.
[25] X. Wang, X. Zhang, Y. Wang, H. Liu, J. Qiu, J. Wang, W. Han, K.L. Yeung, ACS
Catal. 1 (2011) 437–445.
[26] G. Zou, W. Zhong, L. Mao, Q. Xu, J. Xiao, D. Yin, Z. Xiao, S.R. Kirk, T. Shu, Green
Chem. 17 (2015) 1884–1892.
In conclusion, the threshold of Ti to incorporate into the Si
framework was 1.36 wt% by a modified hydrothermal synthesis
method used in the present work. The two kinds of Ti species on
TS-1 had different contributions to the partial oxidation of toluene,
that is, the framework Ti was responsible for the ring CAH bond
activation to produce cresols, while the extra-framework Ti was
advantageous to the formation of side chain oxidation products
and to the deep oxidation of both kinds of the products formed
in the first step. In toluene oxidation with TS-1 and hydrogen per-
oxide the ring hydroxylation was performed by framework
titanium-hydroperoxo sites, while the side chain oxidation
occurred through
a radical mechanism promoted by extra-
framework Ti. The highest yield of 32.8% to cresols was obtained
with the selectivity of 84.3%, under the optimized reaction condi-
tion of 0.5 g TS-1c, 2.3 mmol toluene, 39.2 mmol H2O2, 30 mL
H2O, in 60 °C water-bath for 120 min. Para-cresol was produced
with high selectivity (66.8%), which might be caused by the pore
structure of the catalyst and steric hindering of methyl group.
The catalyst could be regenerated by calcination, and the regener-
ated catalyst exhibited stable activity and selectivity after the
second run.
[27] X. Lu, Y. Guan, H. Xu, H. Wu, P. Wu, Green Chem. 19 (2017) 4871–4878.
[28] N. Wilde, M. Pelz, S.G. Gebhardt, R. Gläser, Green Chem. 17 (2015) 3378–3389.
[29] R. Millini, G. Perego, K. Seiti, Stud. Surf. Sci. Catal. 84 (1994) 2123–2129.
[30] A.C. Alba-Rubio, J.L.G. Fierro, L. León-Reina, R. Mariscal, J.A. Dumesic, M. López
Granados, Appl. Catal. B 202 (2017) 269–280.
[31] X. Feng, N. Sheng, Y. Liu, X. Chen, D. Chen, C. Yang, X. Zhou, ACS Catal. 7 (2017)
2668–2675.
[32] C. Xia, M. Lin, A. Zheng, Y. Xiang, B. Zhu, G. Xu, X. Shu, J. Catal. 338 (2016) 340–
348.
[33] H. Liu, G. Li, C. Hu, J. Mol. Catal. A: Chem. 377 (2013) 143–153.
[34] Y. Zhong, G. Li, L. Zhu, Y. Yan, G. Wu, C. Hu, J. Mol. Catal. A: Chem. 272 (2007)
169–173.
[35] D. Zhang, L. Gao, Y. Wang, W. Xue, X. Zhao, S. Wang, Catal. Commun. 12 (2011)
1109–1112.
Acknowledgement
The authors gratefully acknowledged the financial supports
from NSFC (No. 21372167 No. 21172157) as well as the province
from Sichuan (2016JY0181) and the province supported by the
Fundamental Research Funds for the Central Universities
(SCU2018D003, SCU2017D007). The technical support was
provided by Analytical and Testing Center of Sichuan University.
Especially, we would like to thank Yuefei Tian of the Analytical
and Testing Center of Sichuan University for the XPS
characterization.
[36] A. Raba, M. Cokoja, W.A. Herrmann, F.E. Kuhn, Chem Commun. 50 (2014)
11454–11457.
[37] X. Liu, C. Yang, Y. Wang, Y. Guo, Y. Guo, G. Lu, Chem. Eng. J. 243 (2014) 192–
196.
[38] V.A. de la Pena O’Shea, M. Capel-Sanchez, G. Blanco-Brieva, J.M. Campos-
Martin, J.L. Fierro, Angew. Chem. Int. Ed. 42 (2003) 5851–5854.
[39] K. Bendahou, L. Cherif, S. Siffert, H.L. Tidahy, H. Benaïssa, A. Aboukaïs, Appl.
Catal. A 351 (2008) 82–87.
[40] Z. Qu, Y. Bu, Y. Qin, Y. Wang, Q. Fu, Chem. Eng. J. 209 (2012) 163–169.
[41] W. Zhong, S.R. Kirk, D. Yin, Y. Li, R. Zou, L. Mao, G. Zou, Chem. Eng. J. 280 (2015)
737–747.
[42] B. Du, S.-I. Kim, L.-L. Lou, A. Jia, G. Liu, B. Qi, S. Liu, Appl. Catal. A 425–426
(2012) 191–198.
References
[1] N.A. Romero, K.A. Margrey, N.E. Tay, D.A. Nicewicz, Science 349 (2015)
1326–1330.
[2] Q. Liu, L. Zhu, L. Li, B. Guo, X. Hu, C. Hu, J. Mol. Catal. A: Chem. 331 (2010)
71–77.
[3] Y. Luo, J. Xiong, C. Pang, G. Li, C. Hu, Catalysts 8 (2018) 49.
[4] D. Barbera, F. Cavani, T. D’Alessandro, G. Fornasari, S. Guidetti, A. Aloise, G.
Giordano, M. Piumetti, B. Bonelli, C. Zanzottera, J. Catal. 275 (2010) 158–169.
[5] M. Feller, A. Karton, G. Leitus, J.M.L. Martin, D. Milstein, J. Am. Chem. Soc. 128
(2006) 12400–12401.
[43] X. Li, B. Lu, J. Sun, X. Wang, J. Zhao, Q. Cai, Catal. Commun. 39 (2013) 115–118.
[44] B. Puértolas, L. García-Andújar, T. García, M.V. Navarro, S. Mitchell, J. Pérez-
Ramírez, Appl. Catal. B 154–155 (2014) 161–170.
[6] D. Wang, X. Yu, X. Xu, B. Ge, X. Wang, Y. Zhang, Chem. Eur. J. 22 (2016)
8663–8668.

C. Pang et al. / Journal of Catalysis 366 (2018) 37–49
49
[45] M. Popova, Á. Szegedi, P. Németh, N. Kostova, T. Tsoncheva, Catal. Commun. 10
(2008) 304–308.
[46] M. Popova, A. Szegedi, Z. Cherkezova-Zheleva, I. Mitov, N. Kostova, T.
Tsoncheva, J. Hazard. Mater. 168 (2009) 226–232.
[61] L. Balducci, D. Bianchi, R. Bortolo, R. D’Aloisio, M. Ricci, R. Tassinari, R.
Ungarelli, Angew. Chem. Int. Ed. 42 (2003) 4937–4940.
[62] M. Wang, J. Zhou, G. Mao, X. Zheng, Ind. Eng. Chem. Res. 51 (2012) 12730–
12738.
[47] X. Fu, Y. Liu, W. Yao, Z. Wu, Catal. Commun. 83 (2016) 22–26.
[48] Y. Zhong, G. Li, L. Zhu, D. Tang, C. Hu, J. Chinese U. 28 (2007) 1570–1572.
[49] Y. Qin, Z. Qu, C. Dong, N. Huang, Chin. J. Catal. 38 (2017) 1603–1612.
[50] B. Guo, L. Zhu, X. Hu, Q. Zhang, D. Tong, G. Li, C. Hu, Catal. Sci. Technol. 1 (2011)
1060.
[51] J.A. Martens, P. Buskens, P.A. Jacobs, Appl. Catal. A 99 (1993) 71–84.
[52] J. Li, J. Li, Z. Zhao, X. Fan, J. Liu, Y. Wei, A. Duan, Z. Xie, Q. Liu, J. Catal. 352 (2017)
361–370.
[63] D.-G. Huang, X. Zhang, B.-H. Chen, Z.-S. Chao, Catal. Today 158 (2010) 510–
514.
[64] G. Wu, J. Xiao, L. Zhang, W. Wang, Y. Hong, H. Huang, Y. Jiang, L. Li, C. Wang,
RSC Adv. 6 (2016) 101071–101078.
[65] C. Liu, J. Huang, D. Sun, Y. Zhou, X. Jing, M. Du, H. Wang, Q. Li, Appl. Catal. A 459
(2013) 1–7.
[66] X. Wang, P. Zhang, X. Liu, B. Zhang, Appl. Surf. Sci. 254 (2007) 544–547.
[67] G. Bellussi, R. Millini, Struct Bond, Spring International Publishing AG, 2017.
[68] T. Lu, J. Zou, Y. Zhan, X. Yang, Y. Wen, X. Wang, L. Zhou, J. Xu, ACS Catal. (2018)
1287–1296.
[53] G. Li, C. Zhang, Z. Wang, H. Huang, H. Peng, X. Li, Appl. Catal. A 550 (2018) 67–
76.
[54] M. Du, G. Zhan, X. Yang, H. Wang, W. Lin, Y. Zhou, J. Zhu, L. Lin, J. Huang, D. Sun,
L. Jia, Q. Li, J. Catal. 283 (2011) 192–201.
[55] R. Millini, G. Bellussi, Catal, Sci. Technol. 6 (2016) 2502–2527.
[56] R. Millini, G. Perego, D. Berti, W.O.P. Jr., A. Carati, G. Bellussi, Micropor.
Mesopor. Mater. 35–36 (2000) 387–403.
[69] M. Mrowetz, E. Selli, PCCP 7 (2005) 1100.
[70] A. Costine, T. O’Sullivan, B.K. Hodnett, Catal. Today 99 (2005) 199–208.
[71] X. Zhang, Y. Wang, F. Xin, Appl. Catal. A 307 (2006) 222–230.
[72] Q. Wang, L. Wang, J. Chen, Y. Wu, Z. Mi, J. Mol. Catal. A: Chem. 273 (2007) 73–
80.
[57] R. Millini, E.P. Massara, G. Perego, G. Bellussi, J. Catal. 137 (1992) 497–503.
[58] T. Yu, R. Yang, S. Xia, G. Li, C. Hu, Catal. Sci. Technol. 4 (2014) 3159–3167.
[59] P. Stathi, Y. Deligiannakis, M. Louloudi, Catal. Today 242 (2015) 146–152.
[60] J. Wallace, O. Parker, R. Millini, J. Am. Chem. Soc. 128 (2006) 1450–1451.
[73] H. Liu, G. Lu, Y. Guo, Y. Guo, J. Wang, Chem. Eng. J. 108 (2005) 187–192.
[74] Z. Niu, G. Liu, H. Yin, D. Wu, C. Zhou, Fuel 172 (2016) 1–10.
[75] Z. Niu, G. Liu, H. Yin, C. Zhou, D. Wu, B. Yousaf, C. Wang, Energy Convers.
Manage. 124 (2016) 180–188.