Highly selective 1-pentene epoxidation over Ti-MWW with modified microenvironment of Ti active sites

Article abstract of DOI:10.1039/d0cy00478b

A titanosilicate/H₂O₂catalytic system was applied to process the liquid-phase selective epoxidation of 1-pentene to 1,2-epoxypentane (EP). The effects of titanosilicate topology (MWW, MFI, MSE, MEL, MOR, and *BEA), solvent, H₂O/H₂O₂ratio, catalyst amount, reaction temperature, pressure, and time on the EP production were investigated systematically. The Ti-MWW/H₂O₂/acetonitrile system exhibited the highest 1-pentene conversion of 72.9% together with high EP selectivity of 99.9% and H₂O₂utilization efficiency of 91.5%. Moreover, it was proved that the Ti active sites located inside the intralayer 10-membered ring sinusoidal channels catalyzed the epoxidation process primarily owing to their supplying more steric fitness for 1-pentene molecules. A piperidine (PI)-assisted structural rearrangement of Ti-MWW was performed to further enhance the catalytic activity, almost doubling the turnover number value. The evolution of the microenvironment of Ti active sites in this structural rearrangement process was carefully investigated, revealing the coordination of N atoms in PI molecules to the Ti atoms. More importantly, we identified that the hexa-coordinated Ti sites with the PI molecules as ligand could significantly accelerate H₂O₂activation, the effect of which far exceeded the inhibition effect caused by the electronegativity increase of Ti active sites.

Full text of DOI:10.1039/d0cy00478b

View Article Online
View Journal
Catalysis
Science &
Technology
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. Yin, H. Xu, B.
Wang, W. Tian, J. Yin, J. Jiang and P. Wu, Catal. Sci. Technol., 2020, DOI: 10.1039/D0CY00478B.
Volume
Number
21 January 2017
Pages 313-534
7
2
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Catalysis
Science &
Technology
rsc.li/catalysis
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 2044-4761
PAPER
Chaoqiu Chen, Yong Qin et al.
Highly dispersed Pt nanoparticles supported on carbon
nanotubes produced by atomic layer deposition for hydrogen
generation from hydrolysis of ammonia borane
rsc.li/catalysis

Page 1 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
Highly selective 1-pentene epoxidation over Ti-MWW with modified
micro-environment of Ti active sites
Jinpeng Yin, Hao Xu*, Bowen Wang, Wenwen Tian, Jianyong Yin, Jingang Jiang, Peng Wu*
Shanghai Key Laboratory of Green Chemistry and Chemical Processes, School of Chemistry and
Molecular Engineering, East China Normal University, Shanghai 200062, China
Tel: +86-21-6223-2292
Fax: +86-21-6223-2292
E-mail address: hxu@chem.ecnu.edu.cn (H. Xu), pwu@chem.ecnu.edu.cn (P. Wu).

Catalysis Science & Technology
Page 2 of 49
View Article Online
DOI: 10.1039/D0CY00478B
Abstract
The titanosilicate/H2O2 catalytic system was applied to process the liquid-phase
selective epoxidation of 1-pentene to 1,2-epoxypentane (EP). The effects of
titanosilicate topology (MWW, MFI, MSE, MEL, MOR, and *BEA), solvent,
H2O/H2O2 ratio, catalyst amount, reaction temperature, pressure, and time on the EP
production were investigated systematically. The Ti-MWW/H2O2/acetonitrile system
exhibited the highest 1-pentene conversion of 72.9% together with high EP selectivity
of 99.9% and H2O2 utilization efficiency of 91.5%. Moreover, it was proved that the Ti
active sites located inside the intralayer 10-membered ring sinusoidal channels
catalyzed the epoxidation process primarily owing to supplying more steric fitness for
1-pentene molecules. A piperidine (PI)-assisted structural rearrangement of Ti-MWW
was performed to further enhance the catalytic activity, almost doubling the turnover
number value. The evolution of the micro-environment of Ti active sites in this
structural rearrangement process was carefully investigated, revealing the coordination
of N atoms in PI molecules to the Ti atoms. More importantly, we identified that the
hexa-coordinated Ti sites with the PI molecules as ligand could significantly accelerate
H2O2 activation, the effect of which far exceeded the inhibition effect caused by the
electronegativity increase of Ti active sites.

Page 3 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
1. Introduction
1,2-Epoxypentane (EP), as one of the important oxygen-containing derivatives
of 1-pentene, serves as the raw material for producing many high value-added chemical
products, such as cosmetics additives, polyester fiber, and pharmaceutical intermediates.
Among them, 1,2-pentanediol (PD) is now the most important downstream product of
EP, which is the indispensable intermediate for propiconazole synthesis with high
global consumption, expensive price currently and increasing demand year by year [1-
4]. Meanwhile, a large number of C5 components are generated in the petrochemical
processes. For instance, the production of ethylene from petroleum cracking in China
produced 2500 kt/a C5 distillates in 2015, among which the amount of 1-pentene
accounted for a large proportion. However, most of these C5 hydrocarbons are
commonly used as additives in high-octane gasoline. Considering the comprehensive
utilization of resources, it is obviously a waste to burn 1-pentene directly as fuel.
Therefore, it is quite competitive to develop efficient process of 1-pentene to EP.
Industrially, the chlorohydrin process, initially conducted by Wurtz for the production
of propylene oxide (PO) in 1859, has been used as the main route for EP synthesis for
many decades. However, it suffers serious problems of waste pollution and equipment
corrosion. The peroxy organic acid oxidation method is gradually replacing the
traditional chlorohydrin process [5-7]. Nevertheless, the peroxy acid oxidant is quite
explosive and corrodes the equipment severely. Furthermore, the two industrialized
processes both have low atomic economy.
In 1983, the first generation titanosilicate TS-1 was reported by Taramasso et al,

Catalysis Science & Technology
Page 4 of 49
View Article Online
DOI: 10.1039/D0CY00478B
which was later used as a highly effective and eco-friendly catalyst for the selective
oxidation reactions with H2O2 as an oxidant [8]. The TS-1/H2O2 system allows the
reactions operated under mild conditions, yielding only H2O as the main by-product.
Subsequently, its success in developing industrial applications for manufacturing
various oxygen-containing bulk chemicals, especially PO, marks an important
milestone in zeolite catalysis [9-11]. With the purpose to develop highly efficient and
clean selective oxidation processes using titanosilicate/H2O2 catalytic system,
numerous studies have been carried out in the respective of pore architecture of
titanosilicates and micro-environment of Ti active sites.
The pore structure mainly affects the diffusion of substrate molecules and the
accessibility of active sites. Due to the restriction of its medium 10-membered ring (MR)
pores, TS-1 always exhibits a low catalytic ability in the epoxidation of bulky molecules.
Therefore, other titanosilicates with larger pores were developed successively, such as
Ti-MWW [12], Ti-Beta [13], Ti-ITQ-7 [14], Ti-MCM-68 [15] and Ti-MCM-41 [16].
Especially, the layered Ti-MWW, with two independent 10-MR channels, 12-MR side
cups and supercages, shows high reaction activity in the epoxidation of both small and
bulky substrates. In addition, introducing mesopores even macropores, changing the
morphology of titanosilicate and tuning the structural hydrophilicity/hydrophobicity are
also regarded as effective strategies for enhancing the catalytic activity in epoxidation
reactions, especially involving the bulky substrates [17-19].
In terms of the reaction mechanism of the epoxidation reaction catalyzed by
titanosilicate/H2O2 system, it is commonly believed that the isolated Ti (Ⅳ) centers

Page 5 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
firstly activate H2O2 molecules to form the intermediates hydroperoxo (Ti-Oα-Oβ-H)
and peroxo (Ti-(η2-O2)) species (Scheme S1 † ), and the two intermediates were
interconvertible in the presence of H2O molecule [20]. Subsequently, numerous studies
revealed that the Ti-Oα-Oβ-H intermediate, which interact with the protic solvent
molecules of alcohols or water via hydrogen bonding, was responsible for epoxidation
reaction on Ti-zeolite, rather than the Ti-(η2-O2) species [10, 21, 22]. Then, the
electron-rich alkene substrates attack the intermediates through nucleophilic reaction to
transfer the oxygen atoms into the product [23-25]. Thus, modulating the electronic
environment of Ti active sites is an alternative way to improve the catalytic activity of
titanosilicates. Considering the more electronegative Oα directly coordinated to the Ti
site attacks the C=C bond in the alkene epoxidation process, enhancing the
electropositivity of Ti centers would accelerate the nucleophilic reaction and improve
the catalytic activity. The catalytic performance of Ti-MWW was significantly
enhanced by introducing fluorine atoms into the framework to create the SiO3/2F species
coordinated with Ti species [26-28]. The Ti species in titanosilicate catalysts mainly
exist as framework TiO4 species [29], framework TiO5 species [30], framework TiO6
species [31], non-framework TiO6 species [32] and anatase [33], among which only the
framework Ti species can catalyze the epoxidation effectively. And it is commonly
accepted that framework Ti species with higher coordination number exhibited the
better catalytic activity in epoxidation reactions [30, 34]. For a long time, abundant
efforts have been devoted to construct the highly active hexa-coordinated Ti species
with H2O as the ligand (Ti(OSi)2(OH)2(H2O)2) via chemical post-modification and

Catalysis Science & Technology
Page 6 of 49
View Article Online
DOI: 10.1039/D0CY00478B
adding crystallization-supporting agent in synthesizing titanosilicates [5, 34, 35].
Recently, we communicated an organic-inorganic hybrid MWW-type titanosilicate
with a two-dimensional (2D) layered structure exhibited greatly enhanced activity in
the cyclopentene epoxidation reaction, which was constructed by structural
arrangement of 3D Ti-MWW zeolite and possessed hexa-coordinated Ti species with
piperidine (PI) as the ligand (Ti(OSi)2(OH)2(H2O)PI) besides the framework TiO4
species [36]. However, the exact reason for the enhanced activity by introducing the PI
ligand is still unclear.
Herein, the epoxidation of 1-pentene catalyzed by titanosilicate with H2O2 have
been systemically studied to produce EP selectively and efficiently. By comparison to
the other five representative titanosilicates, Ti-MWW showed the highest catalytic
activity with acetonitrile (MeCN) as the solvent. The effects of the reaction conditions
on the 1-pentene epoxidation over Ti-MWW were investigated in detail. The activity
was further enhanced by introducing PI as the ligand to construct open Ti sites. The
reasons were discussed in depth for the outstanding activity of PI-modified Ti-MWW
titanosilicate despite a partial blockage of micropores aroused by PI coordination.
2. Experimental
2.1 Catalyst preparation
2.1.1 Synthesis of different titanosilicates
Six titanosilicates with different topologies, including Ti-MWW, TS-1, TS-2, Ti-
MOR, Ti-MCM-68 and Ti-Beta, were employed to catalyze the liquid-phase 1-pentene
epoxidation reaction. According to the procedures reported by previous literature [12],

Page 7 of 49
Catalysis Science & Technology
View Article Online
Ti-MWW was hydrothermally synthesized with piperidine (PI) as the sDtrOuI:c1t0u.1r0e3-9/D0CY00478B
directing agent (SDA) and using boric acid as the crystallization-supporting agent. The
synthetic mixture gel with a molar composition of 1.0 SiO2: 0.03 TiO2: 1.4 PI: 0.67
B2O3: 19 H2O was hydrothermally crystallized at 443 K for 7 days under a rotation rate
of 10 rpm. After crystallization, the product was filtered, washed with deionized water
and dried at 373 K overnight, giving 2D Ti-MWW lamellar precursor denoted as P-Ti-
MWW. Subsequently, P-Ti-MWW was treated by refluxing in 2.0 M HNO3 at 413 K
for 4 h to remove extra-framework Ti species and a part of framework B species as
well. The acid-treated product was filtered, washed with deionized water, dried at 373
K overnight, and calcined at 823 K for 6 h subsequently to obtain the 3D Ti-MWW
catalyst. In addition, this matrix suffered the above-mentioned acid-treatment without
calcination was named as A-Ti-MWW. TS-1 was hydrothermally synthesized by
classic method [8]. The as-synthesized TS-1 was treated with 1 M HCl solution to
remove extra-framework Ti species before calcination. TS-2 with MEL topology was
hydrothermally synthesized following the procedure reported by R. Kumar et al [37].
Ti-MOR was obtained by atom-planting method with the gas-solid reaction between
deeply dealuminated mordenite and vapor TiCl4 at elevated temperature [38]. The Ti-
MCM-68 zeolite with MSE topology was also synthesized by performing the gas-solid
reaction over the dealuminated MCM-68 zeolite [15]. Ti-Beta was synthesized in
fluoride medium via the structural reconstruction method [39].
2.1.2 Structural rearrangement of Ti-MWW catalyst
According to the previous report [36], the transformation of 3D Ti-MWW to 2D

Catalysis Science & Technology
Page 8 of 49
View Article Online
lamellar structure was carried in the solution containing PI with the molar ratDioOI:o1f0.110.309/D0CY00478B
SiO2: 0.4 PI: 10 H2O at 443 K for 1 day under a rotation speed of 10 rpm. The resultant
product was collected by filtration, washed and dried at 373 K overnight. Then, the 2D
Ti-MWW containing PI molecules in the channels was obtained, which was denoted as
R-Ti-MWW. As a reference, R-Ti-MWW was further calcined at 823 K for 6 h to
remove the organic species, denoted as R-Ti-MWW-cal.
2.2 Characterization methods
The X-ray diffraction (XRD) patterns were collected on a Rigaku Ultima IV
diffractometer using Ni-filtered CuKα radiation (λ = 0.1541 nm). The voltage and
current were 35 kV and 25 mA, respectively. The amount of Si and Ti were quantified
by Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP-AES) on a
Thermo IRIS Intrepid II XSP atomic emission spectrometer. The UV-Vis spectra were
collected on a PerkinElmer UV-Vis Lambda 35 spectrophotometer using BaSO4 as a
reference. UV resonance Raman spectra (UVRRS) were collected on a triple
spectrograph Raman system UV-Raman-100 with the excitation line at 244 nm and
spectral resolution of 3 cm-1. The SEM images were taken on a Hitachi S-4800 scanning
electron microscope. The FT-IR spectra were measured using the self-supported wafer
by a Nicolet Nexus 670 FT-IR spectrometer at a resolution of 2 cm-1. The spectra in the
framework vibration region (500 - 1300 cm-1) were measured using KBr pellet
technology. In order to eliminate the influence of absorbed water, all the samples were
evacuated at 723 K for 3 h before measurement. Thermogravimetric (TG) analysis was
carried out in a Netzsch Sta 4049 F3 apparatus in air with a heating rate of 10 K min-1

Page 9 of 49
Catalysis Science & Technology
View Article Online
in the temperature range of 473-1073 K. The solid-state MAS NMR specDtrOaI: 1w0.1e0r3e9/D0CY00478B
recorded on a VARIAN VNMRS-400 MB NMR spectrometer using a 7.5 mm T3HX
29
probe and single-pulse method. The Si MAS NMR was measured at a frequency of
100.54 MHz and spinning rate of 3 KHz with [(CH3)3SiO]8SiO12 as the chemical shift
reference, while the 13C MAS NMR was measured at a frequency of 79.43 MHz and a
spinning rate of 5 KHz and with adamantane as the chemical shift reference. The
surface and pore volume of the calcined titanosilicates were determined by N2 physical
sorption at 77 K using a BEL SORP instrument after degassing in vacuum at 473 K for
3 h, while the textural properties of the PI-containing Ti-MWW were determined by Ar
physical adsorption at 87.3 K on a Micromeritics ASAP2020 adsorption instrument
after evacuation at 413 K for 3 h. The X-ray photoelectron spectroscopy (XPS) was
measured on the ESCALAB 250 XI equipment. The element analysis was carried out
with an Elementar Vario Ⅲ analyzer.
2.3 Liquid-phase epoxidation of 1-pentene
The liquid-phase 1-pentene epoxidation was carried out in an autoclave reactor
equipped with a Teflon-inner. In a typical run, 50 mg catalyst, 10 mL solvent, 10 mmol
1-pentene and 10 mmol H2O2 (30 wt.%) were added into the reactor, and the reaction
pressure was then adjusted to 0.5 MPa by N2, after the air inside the autoclave was
exhausted by N2 exchange for 3 times. After stirring vigorously at 333 K for 2 h, the
reactor was cooled down with ice water and then depressurized slowly. The products
were analyzed by a Gas Chromatograph (Shimadzu 2014, FID detector) equipped with
Rtx-Wax capillary column, and the generated products were confirmed by a GC-MS

Catalysis Science & Technology
Page 10 of 49
View Article Online
DOI: 10.1039/D0CY00478B
(Agilent 6890 series GC system, 5937 network mass selective detector). The remaining
amount of H2O2 was determined by the titration method with 0.05 M Ce(SO4)2 aqueous
solution.
3. Results and discussion
3.1 Characterization of various titanosilicates
The XRD patterns confirmed that the six titanosilicates synthesized by
hydrothermal synthesis or post-synthesis processes all were highly crystalline materials
without impurity and possessed the topologies of MSE, MWW, MFI, MEL, *BEA, and
MOR, respectively (Fig. 1). UV-Vis and FT-IR spectra were employed to investigate
the coordination states of Ti species. As shown in Fig. 2A, the predominate band at
200-220 nm, assigned to the charge transfer from O2- to Ti4+ in isolated Ti(OSi)4 or
Ti(OSi)3(OH) species, was observed for all the titanosilicates, indicating that Ti species
were mainly tetrahedrally coordinated in the framework [27, 40, 41]. And negligible
bands at 260 nm and 330 nm ruled out the existence of extra-framework amorphous Ti
species and anatase, respectively. Consistently, all the IR spectra showed obvious
characteristic bands at 960 cm-1 (Fig. 2B), which was normally taken as fingerprint of
the incorporation of Ti species in the zeolite framework [10, 42], although the precise
attribution of this IR band is still controversial. Therefore, combining the UV-Vis and
IR results, it could be concluded that Ti species in titanosilicates were mainly
tetrahedrally coordinated in the frameworks. The crystal morphologies of titanosilicates
were provided by the SEM images (Fig. 3). Ti-MCM-68 possessed the aggregates of
nanosized cubic crystals with a particle size of 50 - 100 nm (Fig. 3a). Ti-MWW showed

Page 11 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
a uniform plate-shaped morphology with a thickness of ~100 nm (Fig. 3b). TS-1
possessed unique nanosized sphere crystals with a size of ~ 400 nm (Fig. 3c), while TS-
2 was composed of rod-like nanocrystals (Fig. 3d). Ti-Beta was of nanosized crystalline
particles with a particle size in the range of 50 - 150 nm (Fig. 3e). The irregularly shaped
crystals packed closely were observed for Ti-MOR (Fig. 3f). The textural properties of
these titanosilicates are listed in Table S1†. The titanosilicates possessed the Langmuir
specific surface areas in the range of 514 - 615 m2 g-1, and micropore volume of 0.12 -
0.19 cm3 g-1, respectively, indicating that they were well crystallized and characterized
by microporous structures. Among them, Ti-Beta exhibited the highest surface area,
while Ti-MCM-68 had the highest micropore volume, which were in good accordance
with their morphologies and pore structures.
3.2 Liquid-phase epoxidation of 1-pentene over various titanosilicates
3.2.1 A comparison of catalytic performance among various titanosilicates
The six titanosilicates were applied to catalyze the 1-pentene epoxidation and the
catalytic results are summarized in Table 1. They all exhibited an extremely high EP
selectivity (> 95%), whether in aprotic solvent of MeCN for Ti-MCM-68, Ti-MWW,
Ti-Beta and Ti-MOR or in protic solvent of MeOH for TS-1 and TS-2. And no
significant difference was observed in the crystallinity and the coordination state as
well as the amount of Ti active center between the fresh and spent catalysts (Fig. S1-
6†), indicating the excellent stability of all the Ti-zeolites. However, these
titanosilicates showed different 1-pentene conversion obviously, with the order of Ti-
MWW > TS-1 ≈ Ti-MCM-68 > Ti-Beta > TS-2 > Ti-MOR. Ti-MWW showed the

Catalysis Science & Technology
Page 12 of 49
View Article Online
highest 1-pentene conversion of 72.9%. Considering the different Ti contentsDiOnI: 1t0h.e10s3e9/D0CY00478B
titanosilicates, turnover number (TON) was calculated to compare the specific catalytic
activities of these titanosilicates fairly. The TON value decreased in the order of Ti-
MWW > Ti-MCM-68 > TS-1 > Ti-Beta > TS-2 > Ti-MOR. Ti-MWW exhibited the
highest TON value of 390, while Ti-MOR had the lowest TON value of only 12. TON
value of Ti-MCM-68 was 352, close to that of Ti-MWW (390) but much higher than
that of TS-1 (183). Ti-MOR with 12-MR pores showed the worst activity for 1-pentene
epoxidation. In previous reports, the catalytic activities of Ti-MOR for the epoxidation
of other alkenes were also very low [43, 44], indicating Ti-MOR may be not a suitable
epoxidation catalyst, although it showed excellent catalytic performance in the
ammoximation reactions [45, 46]. For Ti-Beta with 3D 12-MR pores, a higher
hydrophilicity aroused by the intergrowth of polymorph A and B probably induced a
lower activity than TS-1. Although TS-2 possessed the 10-MR pores very similar to
TS-1, its reaction activity was lower, which was consistent with previous results in
catalyzing the epoxidation of other alkenes [47, 48]. It could be deduced that the pore
size of Ti-zeolites is not the only parameter that determines the catalytic performance
in the 1-pentene epoxidation reaction. For the three titanosilicates of TS-1, Ti-MCM-
68 and Ti-MWW with outstanding catalytic activity, the solvent effects were further
investigated in detail.
3.2.2 Influence of solvents
The solvents generally imposed significant influence on the performance of
titanosilicate in catalyzing the epoxidation reactions. The aprotic/protic nature and

Page 13 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
polarity of solvent, solubility of reactants and products, and the effect of diffusion are
all needed to be considered. In addition, the reactants, intermediate species or the
products might have an interaction with the solvent so that the reaction pathway was
also affected possibly [49].
Table 2 compared the catalytic performance of Ti-MWW, Ti-MCM-68 and TS-1
in the 1-pentene epoxidation reaction in various solvents. For Ti-MWW, all the solvents
showed extremely high EP selectivity (＞98%) except H2O, indicating that EP was
relatively stable and not tended to suffer the ring-opening side reaction in non-aqueous
solvent. Ti-MWW exhibited a superior catalytic performance in aprotic solvents of
MeCN and acetone in comparison to protic solvents of MeOH, t-BuOH and H2O. And
MeCN was the most favorable solvent for Ti-MWW, in which the highest 1-pentene
conversion (72.9%), EP selectivity (99.9%), and H2O2 utilization efficiency (91.5%)
were achieved. In protic solvent system, the catalytic activity was higher in t-BuOH
than in MeOH and H2O in despite of the larger diameter of the former one. Nevertheless,
in comparison to the solvent t-BuOH, MeOH showed a relatively higher PD selectivity
of 1.1%, and small amount of pentanediol ethers (PE) were also detected, indicating
that the hydrolysis and solvolysis occurred more easily in MeOH solvent. Ti-MWW
zeolite, originated from the layered precursor, inevitably contains some defect sites
such as interlayer silanol groups due to the incomplete condensation of neighboring
layers and intralayer hydroxyl nests [12]. The defect sites favor the adsorption of protic
molecules, such as alcohol and H2O, which would impose steric hindrance and
diffusion constrains. Thus, Ti-MWW exhibited a higher reaction activity in aprotic

Catalysis Science & Technology
Page 14 of 49
View Article Online
solvent of acetone and MeCN than in protic solvents. It was noteworthy that DdOioI: l10w.10a3s9/D0CY00478B
predominately produced over Ti-MWW catalyst with H2O as the solvent, in which PD
selectivity was up to 98.5%, but the 1-pentene conversion and H2O2 utilization
efficiency were extremely low. On the one hand, the solubility of reactants and products
was much lower in H2O comparing to other investigated organic solvents. On the other
hand, the generated EP and H2O solvent in the system were successively hydrolyzed to
produce by-product PD, which may easily adsorb on the silanol groups to cause
micropore blockage.
Similar to Ti-MWW, Ti-MCM-68 also contains some defect sites owing to the
dealumination process before Ti insertion. Thus, Ti-MCM-68 could have similar
solvent effects as Ti-MWW. The 1-pentene conversion decreased gradually in the order
of MeCN > acetone >MeOH > t-BuOH > H2O, indicating that the aprotic solvent was
more suitable for Ti-MCM-68. Meanwhile, Ti-MCM-68 exhibited the lowest EP
selectivity (89%) in protic solvent MeOH, and a certain amount of PD was also detected,
indicating that EP was hydrolyzed much easier over Ti-MCM-68 in the solvent of
MeOH than that over Ti-MWW. Ti-MCM-68 showed a higher EP selectivity of 98.9%
in solvent H2O, whereas the opposite result was observed for Ti-MWW. It might be
due to the Brønsted acid sites (BAS) derived from the residual framework Al atoms in
Ti-MCM-68. The BAS are effective in catalyzing the epoxide hydrolysis, but the
yielding diol had a strong interaction with the BAS [50], which would cover the BAS
and inhibit the further conversion of epoxide to diol.
In the case of the hydrophobic TS-1 with less defects, MeOH was the most

Page 15 of 49
Catalysis Science & Technology
View Article Online
effective solvent from the viewpoints of reaction activity listed in Table 2D.OAI: 1n0d.103i9t/D0CY00478B
showed a product distribution similar to Ti-MWW and Ti-MCM-68 in the case of
MeOH solvent, meaning the occurrence of solvolysis. TS-1 was almost inactive in the
solvent of H2O apparently. Nevertheless, the aprotic solvent MeCN, with the 1-pentene
conversion low to 8.5%, was proved unsuitable for TS-1 in catalyzing the 1-pentene
epoxidation. In addition, a plenty of diols were produced in the solvent of acetone and
t-BuOH, indicating the ring-opening side reaction occurred obviously over TS-1 along
with the main epoxidation reaction.
Based on the above results, the reaction pathway of the 1-pentene epoxidation
reaction was verified and depicted in Scheme 1. The main product of the 1-pentene
epoxidation was EP, which may be hydrolyzed to form by-product PD on the Lewis
acid centers provided by Ti-O-O-H in titanosilicate-H2O2 system [50]. With MeOH as
the solvent, EP also possibly further underwent solvolysis to produce PE by-products.
After the detail investigation of solvent effect, the three titanosilicates of Ti-MWW,
TS-1, and Ti-MCM-68, were evaluated in the respective of reaction time under their
most favorable solvent (Fig. S7†). Over the whole reaction time courses, Ti-MWW
exhibited the highest catalytic activity, and Ti-MCM-68 was slightly inferior to it, but
the former two titanosilicates were both much higher than TS-1. All the above-
mentioned results proved that Ti-MWW could serve as a highly efficient catalyst for
the 1-pentene epoxidation in solvent of MeCN.
3.2.3 Investigation of the location of the active Ti sites that catalyzed
For Ti-MWW, the active Ti sites are generally located in three different places,

Catalysis Science & Technology
Page 16 of 49
View Article Online
near 12-MR side cups on the external surface of crystal, within the intralayer siDnOuI:s1o0i.1d0a39l/D0CY00478B
10-MR channels and supercages (Scheme S2A†). To gain insight into the specific Ti
location which contributed more to the catalytic activity, the selective poisoning
experiment was performed over Ti-MWW. The poisoning reagent tripropylamine (TPA)
with small molecule size was able to poison all the Ti sites, while the bulky 2,4,6-
collidine (COD) hardly entered 10-MR pores and thus selectively poisoned the Ti sites
within the 12-MR side cups on the external surface of the Ti-MWW (Scheme S2B†).
According to the Ti amount (0.02 mmol) in Ti-MWW catalyst that added into the
reaction system, 2 mmol of the two amines was sufficient to poison the accessible Ti
active sites. As listed in Table S2†, when TPA was added, the 1-pentene conversion
exhibited a dramatic decrease as expected. However, the bulky COD had little impact
on the catalytic activity, suggesting that most of Ti species located at the external 12-
MR side cups did not involve in catalyzing the 1-pentene epoxidation. It was indicated
that although the Ti sites on the external 12-MR side cups had the better accessibility,
the exposed Ti sites might lack steric matching with the substrate molecules. Thus, the
1-pentene epoxidation mainly occurs in the intralayer sinusoidal 10-MR channels and
supercages.
Bearing the main reaction places of Ti-MWW in mind, the reason why Ti-MWW
exhibited superior catalytic performance than TS-1 and Ti-MCM-68 could be discussed
in detail. The activities of Ti-MWW and Ti-MCM-68 with 12×18 MR supercages were
better than that of TS-1, indicating that the appropriately sized supercage structure
might be conductive to the 1-pentene epoxidation owing to the steric restriction. It is

Page 17 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
worth noting that Ti-MWW, Ti-MCM-68 and TS-1 all contain 10-MR pore channels.
However, the catalytic activity of Ti-MWW with 2D 10-10-MR pore structure was
higher than that of TS-1 with 3D 10×10-MR pore structure and that of Ti-MCM-68
with 3D 12×10×10-MR pore structure. The different activity was speculated to be
related to the pore architecture of the three titanosilicates with different topologies. The
intralayer sinusoidal 10-MR channels of Ti-MWW was more restricted and elliptical
(4.05.9 Å) than that of TS-1 (5.15.5 Å) and Ti-MCM-68 (5.25.8 Å and 5.25.2 Å),
which might supply better steric matching between the linear 1-pentene and
intermediate Ti-O-O-H complex without affecting the diffusion of substrates. This
result also highlighted the significance of taking into account the relationship between
the shape of substrate molecule and the topology of titanosilicate for developing a
highly active titanosilicate-H2O2 catalytic system.
3.3 Catalytic performance of Ti-MWW in 1-pentene epoxidation
3.3.1 Influence of the H2O/H2O2 molar ratio
H2O was inevitably introduced into the catalytic system when aqueous H2O2 was
used as the oxidant for epoxidation reactions. Hence, the effect of H2O/H2O2 molar
ratio on catalytic performance was investigated for Ti-MWW catalyzed 1-pentene
epoxidation. As depicted in Fig. 4, both the conversions of H2O2 and 1-pentene were
linearly increased with the H2O/H2O2 molar ratio increased from 4.4 to 7.4, whereas no
obvious effect was observed with the H2O/H2O2 further increasing. H2O2 efficiency
was not significantly affected by the addition of H2O in the reaction system. However,
EP selectivity was gradually decreased along the increase of H2O/H2O2. Thus, in the

Catalysis Science & Technology
Page 18 of 49
View Article Online
Ti-MWW/H2O2/MeCN catalytic system, the increasing of H2O/H2O2 couldDOslI:i1g0h.1t0l3y9/D0CY00478B
promote H2O2 and 1-pentene conversion, while it induced negative effect on the
epoxide selectivity.
3.3.2 Influence of reaction temperature and time
The reaction temperature had conspicuous effect on catalytic performance of Ti-
MWW in the 1-pentene epoxidation reaction. For both H2O2 and 1-pentene conversions,
higher reaction temperature favored higher conversion (Fig. 5A and C). They increased
rapidly at the beginning of catalytic reaction, and then the speed gradually slowed down
with further prolonging the reaction time. As shown in Fig. 5B, H2O2 utilization
efficiency decreased slowly with increase of reaction temperature, indicating higher
reaction temperature promoted the non-productive decomposition of H2O2. Meanwhile,
the EP selectivity always maintained at > 99% (Fig. 5D), which again confirmed the
effective inhibition of secondary reaction of hydrolysis using MeCN as the solvent. All
these results indicated the 1-pentene epoxidation should be proceeded at a relatively
low reaction temperature. Considering both the H2O2 utilization efficiency and catalytic
activity, 333 K is considered to be an optimal reaction temperature.
3.3.3 Influence of reaction pressure
Since 1-pentene is gas phase at the reaction temperature of 333 K, reaction
pressure would have an important influence on the gas-phase involved catalytic system.
As depicted in Fig. 6, when the reaction pressure increased from 0.1 to 0.5 MPa, H2O2
and 1-pentene conversion as well as the H2O2 utilization efficiency increased
proportionally. Further increasing the pressure, the catalytic performance levelled off.

Page 19 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
Notably, EP selectivity was almost unaffected by the reaction pressure. Increasing the
reaction pressure could improve the solubility of 1-pentene in MeCN, so as to enhance
the conversion of 1-pentene.
3.3.4 Influence of Ti-MWW catalyst amount
The catalyst amount was also investigated and the results were shown in Fig. 7.
Reasonably, larger amount of catalyst favored the conversion of H2O2 and 1-pentene.
The H2O2 utilization efficiency slightly increased with the increase of Ti-MWW
amount, whereas the opposite result was found for EP selectivity. The H2O2 conversion,
efficiency and 1-pentene conversion reached the maximum of 94.6%, 96.3% and 91.7%,
respectively, when 200 mg Ti-MWW was used. However, the EP selectivity decreased
to 97.6%. Therefore, although increasing the Ti-MWW amount lead to the higher
reactant conversion and oxidant utilization, the hydrolysis of epoxide was also
promoted due to the introduction of more Lewis acid sites.
3.4 Ti-MWW modification by structural rearrangement with piperidine
3.4.1 A comparison of the catalytic performance of Ti-MWW and R-Ti-MWW in 1-
pentene epoxidation reaction
To further improve the catalytic performance, the PI-assisted structural
rearrangement was performed over Ti-MWW, yielding an organic-inorganic hybrid
material of R-Ti-MWW. The XRD pattern and 13C MAS NMR confirmed its 2D MWW
structure and the occlusion of organic PI molecules in MWW framework (Fig. S8†).
The PI treatment hardly affected the Ti content and the Ti/N molar ratio was 0.25, but
the TON value of R-Ti-MWW catalyst increased twice as much as that of the 3D Ti-

Catalysis Science & Technology
Page 20 of 49
View Article Online
DOI: 10.1039/D0CY00478B
MWW (Table 3). Meanwhile, the H2O2 utilization efficiency was also improved from
90.2% to 94.6%. Namely, this organic-inorganic hybrid Ti-MWW catalyst, with
partially blocked pores containing PI molecules, exhibited superior catalytic activity
than Ti-MWW with totally open channels in the 1-pentene epoxidation. The apparent
activation energies of Ti-MWW and R-Ti-MWW were calculated by using the
Arrhenius equation for 1-pentene epoxidation and they were 22.0 and 16.3 kJ mol-1,
respectively (Fig. 8), indicating that the PI-assisted structural rearrangement reduced
the apparent activation energy. The values of the activation energies (< 20-30 kJ mol-1)
indicate that the reaction rate is controlled by the diffusion of reactants and not the
chemical interaction itself [51]. Therefore, according to the results of apparent
activation energies, it could be inferred that the reaction rate of R-Ti-MWW was
controlled by the diffusion of reactants, which was different from that of original Ti-
MWW controlled by the chemical reaction itself.
3.4.2 Discussion on PI-incorporation for catalytic activity improvement
More experiments were performed to reveal the reason for higher activity of R-Ti-
MWW. When the PI-containing Ti-MWW was used as the catalyst, the products were
analyzed by GC-MS and no acetamide was detected. Therefore, the contribution of
Payne oxidation to the whole oxidation process could be basically excluded under this
condition, even though the basic PI might favor the Payne oxidation. Once the same
amount PI was added in the Ti-MWW catalyzed system, an obvious decrease of the 1-
pentene and H2O2 conversion was observed (Table S3†), indicating the introduction of
PI into this catalytic system directly suppressed the reaction activity. Thus, it is

Page 21 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
speculated that PI molecules might need to be embedded in the Ti-MWW framework
to enhance the catalytic activity.
P-Ti-MWW and A-Ti-MWW, both with 2D MWW structure and PI molecules
inside the framework (Fig. S8† and Table 4) were applied to catalyze the 1-pentene
epoxidation. However, neither of them was active during the whole reaction time course
(Fig. 9). P-Ti-MWW exhibited an extremely low total surface area (53 m2 g-1) and
micropore volume (0.001 cm3 g-1), indicating its pore channels were almost fully
blocked by PI. The micropore volume and total surface area of R-Ti-MWW was 0.04
cm3 g-1 and 187 m2 g-1, which were close to those of A-Ti-MWW (0.05 cm3 g-1 and 217
m2 g-1), implying their pore channels were partially open. The liquid-phase 1-pentene
adsorption experiment further proved that the substrate molecules could diffuse into the
partially open micropores of R-Ti-MWW and A-Ti-MWW (Fig. S10†), implying the
accessibility of the active sites inside A-Ti-MWW and R-Ti-MWW. Generally, the
weight loss in the range of 473-1073 K was ascribed to the removal of the PI molecules
in MWW zeolite. More specifically, the weight loss in the lower temperature region of
473-643 K was attributed to the decomposition of the PI molecules occluded in the
interlayer space, whereas the decomposition of PI molecules in the intralayer 10-MR
channels occurred in the higher temperature region of 643-1073 K [52]. The weight
loss of P-Ti-MWW, A-Ti-MWW and R-Ti-MWW in the range of 473-643 K was 7.0%,
2.6% and 6.9%, respectively (Fig. S11† and Table 4). However, in the range of 643-
1073 K, the weight loss value of P-Ti-MWW, A-Ti-MWW and R-Ti-MWW were 8.9%,
4.7% and 4.8%, respectively. For the 2D lamellar MWW structure, only the intralayer

Catalysis Science & Technology
Page 22 of 49
View Article Online
DOI: 10.1039/D0CY00478B
sinusoidal 10-MR pores and exterior 12-MR side pockets exist and the supercages
cannot form due to the disconnection of neighboring layers (Scheme S3†). Therefore,
the TG analysis indicated that the opening degree of intralayer sinusoidal 10-MR
micropores of A-Ti-MWW and R-Ti-MWW was similar, which was obviously higher
than that of P-Ti-MWW. The substrate diffusion in the interlayer space of P-Ti-MWW
and R-Ti-MWW was more constrained than that of A-Ti-MWW.
The coordination states of Ti active sites in these PI-containing Ti-MWW zeolites
were analyzed by UV-Vis, UV-Raman and XPS spectra. P-Ti-MWW displayed a main
absorption band at 260 nm, attributed to the amorphous non-framework TiO6 species
[12], together with a weak shoulder band around at 210 nm resulting from the
framework TiO4 species (Fig. 10A). A-Ti-MWW sample only exhibited the band at 210
nm, indicating the non-framework TiO6 species were mostly removed via the acid
washing. For R-Ti-MWW, apart from the TiO4 species, an obvious shoulder bands at
280 nm was also observed, due to the novel hexa-coordinated Ti species owing to the
reinsertion of PI molecules in the structural rearrangement. The structure of this new
hexahedral coordinated Ti species was verified to be the open Ti sites with PI as ligand
in our previous work (Scheme S4c†) [36]. In the UV-Raman spectrum of P-Ti-MWW
(Fig. 10B), the extremely strong band at 700 cm-1 were assigned to the Ti-O-Ti
stretching in the non-framework TiO6 oligomers (Scheme S4, d and e†). Besides, a very
weak band at 1108 cm-1 was also observed, attributed to the symmetric vibration of the
tetrahedron TiO4 in the framework of titanosilicates [32]. For A-Ti-MWW, the band at
700 cm-1 disappeared whereas the bands at 1108 cm-1 was significantly enhanced,

Page 23 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
demonstrating the elimination of non-framework TiO6 species upon acid treatment [53].
R-Ti-MWW showed a new weak Raman band at 724 cm-1, in addition to the band at
1108 cm-1. The Raman bands in the region of 600-800 cm-1 were generally attributed to
the hexahedral Ti species [32, 53]. Thus, the hexahedral Ti species in R-Ti-MWW were
entirely different from those in P-Ti-MWW zeolites, which was consistent with the
observation in UV-Vis spectra, indicating the formation of an unusual hexahedral Ti
species in R-Ti-MWW. Furthermore, since the PI-assisted structural rearrangement
process was performed in alkaline system, slight desilication inevitably occurred and
resulted in the transformation of Si micro-environment, which in turn would affect the
micro-structure of Ti species. The increase in Q3 and the presence of additional Q2 sites
in R-Ti-MWW indicated that a portion of Q4 Si species was transferred to Q3 and Q2
ones (Fig. S12†) and more hydroxyl groups (Si-OH and Ti-OH) generated in the PI-
assisted structural rearrangement, which also implied the transformation of some close
TiO4 species into the open TiO6 species (Table 4). In addition, the presence of more
hydroxyl groups might also facilitate the epoxidation by promoting the enrichment of
H2O2 nearby the Ti sites to improve the activation of H2O2 [19].
In C 1s XPS spectra, two signals with the binding energy (BE) values of 284.7 and
286.4 eV were observed for all the samples (Fig. 11A), which were ascribed to the sp2
bond C and its satellite peak due to energy loss processes, respectively [54], suggesting
the coordination state of carbon atom in the PI molecules was the same for the three PI-
containing Ti-MWW zeolites. However, in the N 1s XPS spectra, apart from the BE
value at 401.7 eV attributed to the C-N bond [55], R-Ti-MWW exhibited a new signal

Catalysis Science & Technology
Page 24 of 49
View Article Online
DOI: 10.1039/D0CY00478B
at 399.0 eV in comparison to the other two Ti-MWW zeolites (Fig. 11B), ascribed to
the N atom in N-Ti-O group [56]. This indicated that R-Ti-MWW contained a new kind
of coordinated N species, due to the coordination of N atoms in the PI with Ti atoms.
In Ti 2p XPS spectra, the signals at 465.3 and 459.5 eV of A-Ti-MWW, corresponding
to Ti 2p3/2 and 2p1/2, respectively, were attributed to the framework TiO4 species [57].
For P-Ti-MWW, the signals at lower BE values of 463.6 and 458.0 eV were observed
due to the large amount of amorphous extra-framework Ti species. Nevertheless, for R-
Ti-MWW, the BE valued of Ti 2p3/2 and 2p1/2 shifted from 459.5 to 459.3 eV and from
465.3 to 464.7 eV compared to A-Ti-MWW, respectively. This indicated that the
density of outer electron cloud of the Ti species in R-Ti-MWW was increased in
comparison with that of tetrahedral Ti species in A-Ti-MWW. Combining with the N
1s XPS spectra, this kind of more negatively charged Ti species in R-Ti-MWW was
supposed to be those coordinated by N atom in the ligand PI molecules (denoted as PI-
TiO6 species).
Based on above characterizations of the three PI-containing Ti-MWW zeolites, P-
Ti-MWW exhibited no reaction activity in 1-pentene epoxidation mainly due to the
blocked pore channels and the enrichment of amorphous extra-framework TiO6 species
that are generally considered to play unfavorable roles in alkene epoxidation [32]
(Scheme 2B). Although the Ti species of A-Ti-MWW were all tetrahedrally
coordinated in the framework and had space for the diffusion of substrate molecules,
A-Ti-MWW hardly catalyze the 1-pentene epoxidation. It was indicated that the
residual uncoordinated PI molecules that embedded within the 10-MR pores may

Page 25 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
poison the active centers, similar to the results of adding PI molecules directly into the
reaction system (Table S3†). The 2D lamellar R-Ti-MWW, with partially blocked pore
channels and micropore volume close to that of A-Ti-MWW, exhibited the superior
catalytic performance than 3D Ti-MWW with totally open pore channels (Fig. 9). Thus,
the fantastic catalytic performance of R-Ti-MWW was positively attributed to Ti
species with unique micro-environment, through the incorporating of the organic ligand
PI. A hot catalyst filtration test was also performed for R-Ti-MWW (Fig. S13†), and
no Si and Ti element was detected in the filtrate, which proved the heterogeneous
catalytic nature of R-Ti-MWW.
Besides, when TBHP was used as oxidant, which arouse great diffusion constrains
due to the formation of bulky active intermediates of Ti alkylperoxo species, the
catalytic activity of R-Ti-MWW decreased significantly (Fig. 9). It could be speculated
that even if there were highly active hexahedral Ti species on the external 12-MR side
cups, the epoxidation still mainly occurred in the intralayer 10-MR micropores, which
further suggested that the degree of steric fitness between substrates and active centers
might have the significant influence on the reaction activity. Additionally, the epoxide
yield of R-Ti-MWW was slightly lower than that of Ti-MWW at the beginning of
reaction, and then it started to increase more rapidly (Fig. S14†), possibly because of
the diffusion constrains in R-Ti-MWW aroused by the coordinated PI molecules.
It is commonly accepted that the epoxidation process in titanosilicate-H2O2 system
is divided into two steps (Scheme 2Aa). One was activation of H2O2 to generate the Ti-
Oα-Oβ-H intermediate, which was thought to the rate determining step, and the other

Catalysis Science & Technology
Page 26 of 49
View Article Online
DOI: 10.1039/D0CY00478B
one was the effective transfer of reactive O to form epoxide. And the Oα species
coordinated directly with Ti sites are more electrophilic and attract the C=C in alkene,
manifesting that increasing the electropositivity of the active Ti sites would be
beneficial to enhance the ability of transferring reactive O. However, according to the
Ti 2p XPS spectra, the BE value decreased after PI coordination (Fig. 11C), indicating
the electronegativity of active Ti sites improved. Namely, taking only the reactive O
transferring in account, the PI-TiO6 species would decrease the activity of R-Ti-MWW
in catalyzing epoxidation reaction. Therefore, the increased epoxidation activity
definitely resulted from that the PI-TiO6 species could significantly accelerate the step
of H2O2 activation, the effect of which far exceeded the inhibition effect caused by the
increasing electronegativity of Ti active sites (Scheme 2Ab).
In addition, as a reference, the R-Ti-MWW was further calcined to remove the
organic species to obtain the R-Ti-MWW-cal, during which the structure converted
from 2D back to 3D and the PI-TiO6 species transformed into the hexa-coordinated Ti
species with H2O as the ligand, denoted as H2O-TiO6 species (Fig. S15†).
As expected, the TON value of R-Ti-MWW-cal in 1-pentene epoxidation was
512, which was obviously higher than 329 of the Ti-MWW matrix, but lower than 663
of the R-Ti-MWW (Table 3). And the apparent energies of R-Ti-MWW-cal was 19.2
kJ mol-1, greater than 16.3 kJ mol-1 of R-Ti-MWW but less than 22.0 kJ mol-1 of Ti-
MWW (Fig. 8). The difference in apparent energies of Ti-MWW and R-Ti-MWW-cal
also implied that the rate-determining step also changed from the reaction itself to the
reactants diffusion. Considering the different micro-environments of the Ti active

Page 27 of 49
Catalysis Science & Technology
View Article Online
center, it could be inferred that the open TiO6 species constructed by sDtrOuI:c1t0u.1r0a39l/D0CY00478B
rearrangement, including the PI-TiO6 and H2O-TiO6 species, were more active than the
close TiO4 species, which enhanced the reaction rate and resulted in the diffusion as the
rate-determining step. Moreover, the crystallinity and the coordination state as well as
the amount of Ti active sites of R-Ti-MWW-cal had not changed after use (Fig. S16†).
The 3D R-Ti-MWW-cal exhibited much lower activity in comparison to the 2D R-Ti-
MWW, although it possessed the completely open pores. This result revealed that the
micro-environment of Ti sites might be more important to the epoxidation activity than
the diffusion limitation of substrates. Notably, for R-Ti-MWW-cal, the BE valued of
2p1/2 shifted from 464.7 to 465.0 eV compared to R-Ti-MWW, while the BE valued of
2p3/2 did not change, still for 459.3 eV (Fig. S17†). This meant that in comparison to
PI-TiO6 species, H2O-TiO6 species was slightly less electronegative and thus promoted
the transfer of reactive O in the epoxidation process, which should theoretically lead to
a higher catalytic activity. Therefore, it indicated that the effect of PI-TiO6 species on
the rate-determine step of H2O2 activation was stronger than that of H2O-TiO6 species
in the 1-pentene epoxidation (Scheme 2Ac).
3.4.3 Recycling tests of the modified Ti-MWW
The stability and reusability of Ti-MWW, R-Ti-MWW and R-Ti-MWW-cal were
compared in the 1-pentene epoxidation (Fig. 12). To better observe the difference of
stability, the initial epoxide yield of all the catalysts was controlled to ~ 100% by
adjusting the reaction conditions. The used catalysts were centrifuged, washed by
MeCN and then dried before being reused in the first six cycles. After six runs, R-Ti-

Catalysis Science & Technology
Page 28 of 49
View Article Online
MWW still showed higher EP yield of 73.6%, while the EP yield decreased tDoO2I: 410..110%39/D0CY00478B
for Ti-MWW and 37.7% for R-Ti-MWW-cal, respectively, which strongly proved an
outstanding stability of R-Ti-MWW catalyst. Then, the deactivated catalysts were
further washed by MeCN/H2O2 at 353 K and dried to reuse. All the three catalysts
regained their activity to a certain degree, signifying the promising feasibility of the in-
situ regeneration of deactivated catalyst by the reaction mixture after separating the
main product in industry. Furthermore, in order to regenerate the three catalysts, they
were calcined at 823 K to burn the organic molecules. For R-Ti-MWW, it was treated
by PI aqueous solution again after calcination. As expected, all the catalysts exhibited
similarly high activity to that of fresh catalysts, verifying the excellent reusability of
these catalysts. It also indicated that the deactivation of Ti-MWW zeolites in 1-pentene
epoxidation mainly caused by the deposition of heavy components like PD inside the
micropores and further covered the active Ti sites. Moreover, no significant difference
was observed in the structure, the state of active center and Ti amount among the fresh,
used and regenerated R-Ti-MWW (Fig. S18†), indicating the excellent stability and
recycling performance of R-Ti-MWW. In addition, the EP hydrolysis experiment was
also operated on the three catalysts. The EP conversion of R-Ti-MWW with H2O2 was
16.7%, far lower than 62.6% of Ti-MWW and 78.5% of R-Ti-MWW-cal (Fig. S19†).
It indicated that the basic PI molecules embedded in titanosilicate could significantly
suppress the Lewis acidity aroused by Ti-O-O-H in the catalytic system, and thus inhibit
the EP hydrolysis and reduce the generation of heavy components.
Conclusions

Page 29 of 49
Catalysis Science & Technology
View Article Online
DOI: 10.1039/D0CY00478B
Highly efficient liquid-phase 1-pentene epoxidation was successfully carried out in
the titanosilicates-H2O2 catalytic system. Among them, Ti-MWW catalyst possessed
the superior catalytic performance in solvent MeCN due to the better steric matching
between pore structure and 1-penten molecules. An organic-inorganic hybrid 2D
structural Ti-MWW that possessed the framework PI-TiO6 species was constructed by
PI-assisted structural rearrangement. In spite of introducing diffusion constrains, the
PI-TiO6 species in the framework could facilitate the formation of the intermediate Ti-
O-O-H by accelerating the H2O2 activation and then improve the reaction activity of
Ti-MWW for the liquid-phase 1-pentene epoxidation. In addition, the basic PI
molecules embedded in Ti-MWW significantly resisted the catalyst deactivation in the
1-pentene epoxidation by suppressing the formation of heavy by-products. This work
shows that a suitable modification of the micro-environment of Ti centers may
efficiently improve the catalytic performance, which is even superior to pore size
enlargement in some cases.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
The authors gratefully acknowledge the financial support from NSFC of China
(21872052, 21533002, 21972044), China Ministry of Science and Technology
(2016YFA0202804).
Notes and References
[1] P. G. Chen, CN Patent 102584802-A, 2012.