A safer and greener chlorohydrination of allyl chloride with H2O2 and HCl over hollow titanium silicate zeolite

Article abstract of DOI:10.1016/j.apcata.2017.06.004

Industrial production of dichloropropanols through chlorohydrination of allyl chloride suffers from a series of disadvantages such as use of hazardous Cl₂, low atom economy, low dichloropropanol concentration and serious pollution. In this work, a safer and greener route for chlorohydrination of allyl chloride with H₂O₂ and HCl over hollow titanium silicate (HTS) at mild condition is developed. Unlike the traditional Cl₂-based chlorohydrination, this novel method is initiated via synergistic effect of Lewis acidity (HTS) and Br?nsted acidity (HCl) to promote occurrence of oxidation, protonation and nucleophilic reaction of allyl chloride simultaneously and hence dichloropropanols are generated. Owing to a completely different reaction route, the formation of 1,2,3-trichloropropane by-product is depressed and the content of dichloropropanol exceeded 22?wt%, which increase by about 4 times compared with traditional Cl₂-based chlorohydrination (the content of dichloropropanol is below 4?wt%). At the optimized conditions, both of the allyl chloride conversion and dichloropropanol selectivity could approach 99% simultaneously and the waste is minimized. What's more, the HTS was reusable. Concentrated HCl solution treatment was adopted to test HTS's stability. The characterization and catalytic evaluation results reveal that, although parts of the framework Ti species have transformed into non-framework Ti and then leached into the solution, HTS remains structural stable, and the allyl chloride conversion and dichloropropanol selectivity didn't decrease obviously during the treatment.

Full text of DOI:10.1016/j.apcata.2017.06.004

AppliedCatalysisA,General543(2017)17–25
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
A safer and greener chlorohydrination of allyl chloride with H2O2 and HCl
over hollow titanium silicate zeolite
Xinxin Penga, Changjiu Xiaa, Min Lina,⁎, Xingtian Shua, Bin Zhua, Baorong Wangb, Yao Zhanga,
Yibin Luoa, Xuhong Mua
a
State Key Laboratory of Catalytic Materials and Reaction Engineering, Research Institute of Petroleum Processing, SINOPEC, 100083 Beijing, PR China
National Engineering Research Center for Carbohydrate Synthesis, Jiangxi Normal University, 330027 Nanchang, PR China
b
A R T I C L E I N F O
A B S T R A C T
Keywords:
Industrial production of dichloropropanols through chlorohydrination of allyl chloride suffers from a series of
disadvantages such as use of hazardous Cl2, low atom economy, low dichloropropanol concentration and serious
pollution. In this work, a safer and greener route for chlorohydrination of allyl chloride with H2O2 and HCl over
hollow titanium silicate (HTS) at mild condition is developed. Unlike the traditional Cl2-based chlorohydrina-
tion, this novel method is initiated via synergistic effect of Lewis acidity (HTS) and Brønsted acidity (HCl) to
promote occurrence of oxidation, protonation and nucleophilic reaction of allyl chloride simultaneously and
hence dichloropropanols are generated. Owing to a completely different reaction route, the formation of 1,2,3-
trichloropropane by-product is depressed and the content of dichloropropanol exceeded 22 wt%, which increase
by about 4 times compared with traditional Cl2-based chlorohydrination (the content of dichloropropanol is
below 4 wt%). At the optimized conditions, both of the allyl chloride conversion and dichloropropanol se-
lectivity could approach 99% simultaneously and the waste is minimized. What’s more, the HTS was reusable.
Concentrated HCl solution treatment was adopted to test HTS’s stability. The characterization and catalytic
evaluation results reveal that, although parts of the framework Ti species have transformed into non-framework
Ti and then leached into the solution, HTS remains structural stable, and the allyl chloride conversion and
dichloropropanol selectivity didn’t decrease obviously during the treatment.
Titanium silicate
Synergistic effect
Oxidation
Nucleophilic reaction
Chlorohydrination
Dichloropropanol
1. Introduction
one chlorine atom of one Cl2 molecule can be transferred into the DCP,
however the other chlorine atom is transformed into unrecyclable Cl−
.
As important intermediate chemicals, dichloropropanols (DCP, in-
cluding 1,3-dichloro-2-propanol and 2,3-dichloro-1-propanol) are
mainly used to produce epichlorohydrin (ECH), the third largest pro-
duction capacity of epoxides, which ranks behind ethylene oxide and
propylene oxide. The ECH, global consumption of which is about 1.2
million tons per year with the annual growth rate around 5%, is widely
used for the manufacturing of epoxy resins, synthetic glycerol, and so
on [1].
Although almost 95% of ECH worldwide is produced by Cl2-based
chlorohydrination process [2], which has been accused of several
drawbacks [3,4]. First of all, the utilization of poisonous and corrosive
liquid chlorine decreases its intrinsic safety and increases the cost of
investment and operation. Another problem is low atom economy of
Cl2-based chlorohydrination process. In the traditional process, Cl2 is
firstly dissolved and transformed into HClO and HCl in water, and then
the HClO reacts with allyl chloride (AC) to obtain DCP. Therefore, only
The atom economy of conventional Cl2-based route is about 78%. Due
to the massive employment of Cl2, the chlorination of AC to form low-
valued TCP is inevitable, which accounts for approximately 7% of AC
conversion. The most accused issue is seriously environmental pollu-
tion. In order to inhibit chlorination, tremendous water is used to dilute
the reactants, which in return results in a very low DCP concentration
(below 4 wt%) and large amount of Cl− contained in water. It is esti-
mated that 40 tons of such waste water is formed corresponding to per
one ton of ECH production.
Alternative approaches have been explored, even industrialized.
Take glycerol chlorination for example [5,6]. With less waste release,
HCl gas is adopted as the chlorination reagent to replace liquid Cl2,
however the complexity of the process and an unsteady supply of gly-
cerol are the major challenges to its application. Direct epoxidation of
AC is regarded as the green process for ECH production [7–9]. How-
ever, owing to the electron withdrawing effect of the Cl, epoxidation of
⁎
Corresponding author.
E-mail address: minlin@tom.com (M. Lin).
http://dx.doi.org/10.1016/j.apcata.2017.06.004
Received 18 March 2017; Received in revised form 28 May 2017; Accepted 1 June 2017
Availableonline13June2017
0926-860X/©2017ElsevierB.V.Allrightsreserved.

X. Peng et al.
AppliedCatalysisA,General543(2017)17–25
AC is not as easy as propylene conversion, and the conversion of AC is
relatively low. In addition, a large amount of methanol is used as sol-
vent and the methanol oxidation involved side reactions make the
solvent purification process complicate.
from a Hitachi 4800 microscope (20 kV). NMR spectra were collected
on Bruker AVANCE III 500 WB. Products qualitalive analysis were
conducted on Agilent 7890A/5977C GC–MS.
Due to the excellent oxidative activity, titanium silicate (TS-1) has
been widely used in ketone ammoximation [10], phenol hydroxylation
and olefin epoxidation [11,12], etc. SINOPEC has developed a new
brand of titanium silicate, hollow titanium silicate (HTS) with better
accessibility of active centers, which has already been commercially
applied in cyclohexanone-ammonium oximation [13] and propylene
epoxidation process [14], etc. In order to overcome the drawbacks of
the Cl2-based chlorohydrination process, especially the poisonous Cl2,
low atom economy of conventional Cl2-based route and low DCP con-
centration problems, herein, we propose a simple, safe, green and ea-
sily-established method for DCP synthesis via the chlorohydrination of
AC with aqueous H2O2 and HCl over HTS under mild reaction condi-
tions [15]. In this reaction system, the synergistic effect of Lewis acidity
of Ti species within HTS and Brønsted acidity of HCl was observed,
through which AC is oxidized by H2O2 over the catalysis of HTS, and
the intermediate product immediately react with HCl to form DCP. To
the best of our knowledge, catalytic oxidation under strongly acidic
reaction condition, to produce DCP has not been fully studied before.
Therefore, a detailed investigation of reaction conditions, especially the
reusability of HTS under appropriate reaction condition and the
leaching effect on the HTS zeolite which were not published before are
discussed in this article. Meanwhile, many characterization methods
were employed to analyze the chemicals that were trapped inside the
zeolite channels, as well as to investigate the physicochemical states
and catalytic properties transformation from framework Ti species to
non-framework Ti species under strong acidic environment.
2.3. Procedure for chlorohydrination of allyl chloride
Chlorohydrination of AC with H2O2 and HCl were evaluated over
different samples. Unless otherwise indicated, a typical reaction was
conducted as follows. 2.6 g catalyst, 0.23 mol allyl chloride and 83.86 g
HCl (10 wt%) was mixed in a three-necked flask equipped with a
condenser and a magnetic stirrer. The molar ratio of AC to HCl was 1:1.
0.23 mol H2O2 was added into the flask by peristaltic pump with a flow
rate of 0.65 mL/min. The total time of H2O2 feeding was about 40 min.
The reaction temperature was controlled at 30 °C. The reaction time
was 2 h including H2O2 feeding time. After the reaction, the product
was filtrated and catalyst was washed by 50 mL ethanol. Products were
analyzed by Agilient 7890A/5977C GC–MS and the quantitative ana-
lysis was approached by Agilent 6890N GC.
3. Results and discussion
3.1. Chlorohydrination of AC over different catalysts
Chlorohydrination of AC with H2O2 and aqueous HCl were con-
ducted over different Ti-containing catalysts homogeneously and het-
erogeneously under mild conditions (Fig. 1). Despite of high acid con-
centration, no obvious decomposition of H2O2 was observed because of
very low feeding rate at 0.65 mL/min. We carefully analyzed all the
products over different catalysts by GC–MS, however, only 1,3-di-
chloro-2-propanol (1,3-DCP), 2,3-dichloro-1-propanol (2,3-DCP), 1,2,3-
trichloropropane (TCP), epichlorohydrin (ECH), 3-chloro-1,2-propane-
diol (CPD) and 1,3-dichloro-acetone (DCA), these six kinds of com-
pounds, were found. As shown in Fig. 1, non-catalytic epoxidation,
chlorohydrination and chlorination can be observed in blank test. Ex-
cept for TS-2, TS-1 and HTS, other catalysts do not show obvious im-
provement in AC conversion compared with blank test. However, it is
observed that the TCP selectivity for the first nine Ti-containing cata-
lysts are significantly improved compared with the blank test, which
suggests an enhanced Cl2 generation catalyzed by the Ti species in
presence of HCl and H2O2 [16]. As for the titanium silicate with fra-
mework Ti species (TS-2, TS-1 and HTS), catalytic reactions are sig-
nificantly improved, and moreover, AC conversion and DCP selectivity,
especially the 1,3-DCP selectivity were greatly enhanced.
2. Experimental
2.1. Materials and catalyst preparation
All the materials were used as available. TiCl4 (AR), tetrabutyl ti-
tanate (CP), tetraethyl orthosilicate (AR), tetrapropylammonium hy-
droxide (20 wt%), tetrabutylammonium hydroxide (25 wt%), ce-
trimonium bromide (AR), NaOH (AR), HCl (37 wt%), ally chloride
(99.8 wt%), H2O2 (30 wt%) were purchased from Sinopharm Chemical
Reagent Beijing Co., Ltd. S-1, SBA-15, TS-1 and HTS zeolites were
supplied by SINOPEC Catalyst Co., Ltd. The preparation method of the
oxides and other zeolites was provided in the supplementary informa-
tion. For the HTS treatment by HCl solution experiment, 10 g HTS
zeolite was soaked in 50 g 37 wt% HCl solution in static state. After
treated for certain days, the mixture was then filtered, washed and
dried at 120 °C for 6 h, and then calcined at 550 °C for 6 h. The samples
are designated as HTS-0, HTS-10, HTS-25, HTS-90, HTS-150 depended
on the treating days.
2.2. Characterization
The X-ray diffraction (XRD) patterns were collected on a Philips
Panalytical X'pert diffractometer with nickel-filtered Cu Kα radiation
(40 kV, 250 mA). The 2θ angle scanning ranged from 5° to 35°.
Chemical composition of catalysts was analyzed by X-Ray Fluorescence
(XRF) on a Rigaku 3721E spectrometer running at 40 kV. The X-ray
photoelectron spectroscopy (XPS) spectra were recorded on a Thermo-
Fischer-VG ESCALAB250 with Al Kα radiation. N2 adsorption–desorp-
tion isotherms were collected at 77 K on a Micromeritics ASAP 2405
apparatus. The samples were previously dried under vacuum (0.1 Pa) at
300 °C for 6 h. The surface properties were derived from the isotherms
using BET and t-plot methods. The UV–vis (UV–vis) spectra were re-
corded on a JASCO UV–vis 550 spectrometer. The samples were pressed
into a self-supported wafer, and the spectra was recorded from 200 nm
to 800 nm. Scanning electron microscopy (SEM) images were taken
Fig. 1. AC chlorohydrination catalyzed by different Ti-containing catalysts. Reaction
conditions: HCl:H2O2:AC = 1:1:1 (mol); concentration of HCl solution, 10 wt%; tem-
perature, 30 °C; catalyst concentration, 2 wt%; time, 2 h. Notes: Amor-TiO2, amorphous
TiO2; A-TiO2, anatase; R-TiO2, rutile; B-TiO2, brookite; SiO2-TiO2, amorphous silicon ti-
tanium oxide; TiO2/S-1, S-1 zeolite loaded with TiO2; AC, allyl chloride; 1,3-DCP, 1,3-
dichloro-2-propanol; 2,3-DCP, 2,3-dichloro-1-propanol; TCP, 1,2,3-trichloropropane;
ECH, epichlorohydrin; CPD, 3-chloro-1,2-propanediol and DCA, 1,3-dichloro-acetone.
18

X. Peng et al.
AppliedCatalysisA,General543(2017)17–25
For the traditional process, 2,3-DCP is the major product and TCP is
and CPD as the temperature increased because cyclic ether is proto-
nated in acid solution and thus very active to be attacked by nucleo-
philic reagents, eg Cl−, H2O, alcohols etc (Scheme S3). Although H2O is
a less active nucleophilic reagent than Cl− ion, CPD selectivity is sig-
nificantly enhanced with the rising of temperature. Owing to the dif-
ferent steric hindrance and charge density of two cyclic ether carbon
atoms, Cl− ion would favorably attacks the side carbon atoms rather
than the middle ones [21,22]. For the two possible DCP isomers formed
in ring-opening reaction of ECH, 1,3-DCP is more favored. When Cl2 is
used as oxidant, AC would firstly form a chloronium ion and due to
electron-withdrawing effect of the chlorine substituent on the AC, the
second chlorine would be more favored to bond with the other side
carbon atoms. Meanwhile, hydroxyl group with electron-donating ef-
fect would bond with the middle carbon atoms which promotes the
high selectivity of 2,3-DCP.
The great improvement in DCP content is crucial because it saves a
large amount of energy in the saponification to convert it into ECH and
reduces waste water. In order to minimize the TCP by-product forma-
tion from direct chlorination of molecule Cl2 and AC and maintain a
high DCP selectivity, Cl2-based chlorohydrination prefers to introduce
in large amount of water to dilute reactants. Therefore, the con-
centration of DCP product is usually maintained bellow 4 wt%, which
burdens the saponification process and increases the waste water.
Owing to a different reaction route, the novel chlorohydrination we
proposed can be operated at lower water content reaction conditions
and a higher DCP content can be obtained.
the by-product, while for the novel chlorohydrination catalyzed by
HTS, 1,3-DCP is the predominant product with CPD as the by-product.
This is in accordance with the ECH hydrolysis experimental result as
shown in Fig. S1, that ECH hydrolyzed in HCl solution and gave almost
the same product distribution compared with those over HTS in Fig. 1.
It is known that the 4-coordinated framework Ti species is the active
center for olefins epoxidation [17,18], therefore, it is probable that AC
was firstly oxidized and then attacked by HCl to form DCP. Given the
strong diffusion capacity of H+ and Cl− in aqueous solution, it is also
highly possible that these ions are involved in the AC oxidation process.
Although olefins with electronic withdrawing groups are more difficult
to be epoxidized and to obtain a relatively high conversion than non
substituted ones, for example, propylene, perhaps there existed a sy-
nergistic effect between framework Ti of HTS and HCl through which
AC was oxidized by H2O2 over HTS and meanwhile HCl attacked the
intermediate product to generate DCP. The synergistic effect of Lewis
acidity and Brønsted acidity kept high rates of AC consuming and DCP
formation and hence promoted the AC chlorohydrination in this reac-
tion system.
Based on the above results, possible reaction mechanism of oxida-
tion-chlorohydrination route is proposed and illustrated in Scheme 1
and S1. In which a plausible DCP formation route is suggested that, AC
is firstly oxidized due to the superior epoxidation activity of titanium
silicate, then HCl attacks the intermediate product through protonation
and nucleophilic reaction to form of DCP and to leave 1,3-DCP as a
major product (Scheme S1, Eqs. 3 and 4). Since HTS exhibited the best
chlorohydrination performance among the selected catalysts, effects of
reaction conditions were further investigated and demonstrated in the
following sections.
3.3. Influence of HCl content
The influences of HCl concentration and HCl to H2O2 molar ratio on
AC chlorohydrination were investigated (Fig. 3). As the concentration
of HCl solution increases from 10 wt% to 30 wt%, the AC conversion
and DCP selectivity both increase to above 99% and 91% respectively
(Fig. 3a and b). Although the content of DCP is also increased appar-
ently, this is mainly contributed by the decrease of reactants amount
brought by the increase of HCl concentration, rather than the increase
of DCP yield. As observed in the experiments, the solubility of ECH in
water is about 12 wt% at room temperature. So the liquid product was
supposed to be separated into two phases, including DCP phase and
water phase. As the increasing concentration of HCl, the content of
3.2. Influence of reaction temperature
The reaction temperature of AC chlorohydrination is important
because in our reaction system, increasing reaction temperature pro-
motes both AC oxidation[19,20] and ring-opening reactions (Scheme
S1, Eqs. 1–4) as well as H2O2 decomposition reaction (Scheme S2).
Nevertheless, most parts of AC can be converted at temperature above
20 °C with high DCP selectivity and content (exceeding 75% and 17%,
respectively, Fig. 2). Meanwhile, ECH is easily to be converted into DCP
Scheme 1. Mechanism comparison between Cl2-based and novel chlorohydrination.
19

X. Peng et al.
AppliedCatalysisA,General543(2017)17–25
Fig. 2. The effect of reaction temperature on chlorohydrination. (a) The AC conversion and DCP content and (b) products selectivity at different reaction temperature. Reaction
conditions: HCl:H2O2:AC = 1:1:1 (mol); concentration of HCl solution, 10 wt%; HTS concentration, 2 wt%; time, 2 h.
water phase was decreasing, and during the reaction the insoluble DCP
phase kept increasing. Because AC and Cl2 generated from the oxidation
of HCl and H2O2 (Scheme S1, eq. 6) would be more likely to be dis-
solved in DCP phase rather than water phase. Therefore, the chlorina-
tion of AC was accelerated, which promoted the TCP selectivity slightly.
We also notice that the DCP selectivity is increasing slowly and it is
inferred that, the increasing HCl has indeed promoted HCl involved
nucleophilic reaction which is competing with the H2O involved hy-
drolysis reaction (Scheme 1), and thus the selectivity of DCP is in-
creased and the generation of CPD is inhibited.
Increasing the HCl amount may accelerate two pathways of the reaction
system. One is that HCl reacts with oxidation intermediate product to
generate 1,3-DCP as the main product. The other is large amount of HCl
promotes the generation of Cl2 through H2O2 and HTS participated Cl−
oxidation, and further promoted chlorohydrination and leaving 2,3-
DCP and TCP as product. However, according to the experimental re-
sults, more 1,3-DCP is formed (Fig. 3d), which suggests that the former
route is obviously accelerated. In addition, both the AC conversion and
DCP selectivity reach above 99% through increasing the concentration
of HCl solution (Fig. 3c and d). Due to the chlorination reaction and
esterification of Cl2-based chlorohydrination route, TCP and other
heavy products such as ethers are inevitable; and this causes the
As shown in Fig. 3b and d, increasing the molar ratio of
HCl:H2O2:AC from 1:1:1 to 2:1:1 obtains the increase in DCP selectivity.
Fig. 3. The effect of HCl content on chlorohydrination reaction. The AC conversion and DCP content and products selectivity with the increasing HCl concentration at the molar ratio of
HCl:H2O2:AC = 1:1:1 ((a) and (b)) and 2:1:1 ((c) and (d)),respectively. Reaction conditions: temperature,30 °C; HTS concentration, 2 wt%; time, 2 h.
20

X. Peng et al.
AppliedCatalysisA,General543(2017)17–25
Fig. 4. Reusability of HTS for AC chlorohydrination in HCl solution. (a) The AC conversion and DCP content and (b) products selectivity on consecutive runs. Reaction conditions: HCl:
H2O2:AC = 2:1:1 (mol); concentration of HCl solution, 20 wt%; temperature, 30 °C; HTS concentration, 2 wt%; time, 2 h.
decrease of AC effective utilization rate. By reacting AC directly with
H2O2, HCl and HTS, the atom economy is increased from 78% to 88%,
and AC and H2O2 can be almost fully transferred into target chemical
compounds at optimized conditions with H2O as the only reduction
product, while the extra HCl could be easily separated and recycled.
Therefore, the atom economy can be significantly improved. This result
is very promising because high conversion and selectivity could be si-
multaneously achieved in a simple process, especially for the product
purification and waste disposal, as well as saves a large amount of in-
vestment and energy consuming.
S1. During the whole analyzing process, the ion current signals mainly
appear from the temperature of 150 °C (about the 15th scanning cycle)
to 500 °C (about the 50th scanning cycle) with the summit at the
temperature of 220 °C (about the 22nd scanning cycle) and 400 °C
(about the 40th scanning cycle). Therefore, the ion current signals of
these two cycles are further analyzed. For the ionic signals at 220 °C,
the mass spectra analysis result suggests that, the exothermic peak
before 250 °C is probably generated by desorption of water and ad-
sorbed chemicals such as 1-chloro-2-propanone, 1,3-dichloro-2-pro-
panol and acetate, 2,3-dichloro-1-propanol, etc (Fig. S3 and S4). While
for the ionic signals at 400 °C, the remaining incompletely desorbed
chemicals are combusted, hence intensive signals of H2O, CO2, HCl and
other smaller ion fragments are detected (Figs. S5 and S6).
3.4. Reusability of HTS
Solid state 13C CP/MAS NMR was used to further discriminate the
adsorbed species on the 6 runs reused HTS zeolite (Fig. 5d). According
to the NMR spectrum of the zeolite, there are five identifiable peaks
while two of them (chemical shift at around 72 and 47) could be as-
signed to 1,3-DCP, which confirms the existence of 1,3-DCP in the re-
used HTS zeolite. A slight difference between the NMR results of the 6
runs reused HTS and 1,3-DCP can be attributed to a different chemical
environment during the NMR analysis procedure. Moreover, based on
the integration result of the NMR peaks (Table 2), it is inferred that at
least half of the residues could be identified as 1,3-DCP. Based on the
analysis combining with catalytic activity data, it is safe to conclude
that the HTS zeolite is far from deactivated.
The HTS zeolite reusability are shown in Fig. 4. There are no ob-
vious loss in conversion and selectivity during six consecutive runs.
What’s more, the selectivity of CPD is decreased and the selectivity of
TCP is slightly increased, which indicates that the competing reaction of
Cl2-promoted chlorohydrination is slightly strengthened after several
runs. This pheromone is interesting and the explanation is provided in
the following part.
In order to investigate physical and chemical changes between the
fresh and reused HTS, the fresh and reused zeolite after 6 runs were
characterized by XRD, BET, TG-MS and 13C CP/MAS NMR. XRD ana-
lysis results revealed that although the characteristic peaks between 22°
to 25° and the peaks at higher degree are not influenced obviously, the
intensities of part diffraction peaks especially the peaks before 20° are
decreased obviously (Fig. 5a). Although the relative crystallinity (RC) is
slightly deceased to 93%, the topology structure of the zeolite remains
well.
The isotherms and textural information of the fresh and reused HTS
zeolite are shown in Fig. 5b and Table 1. It is observed that the iso-
therms of 1 run and 6 runs reused HTS overlapped. Although it was
reused for several runs, the isotherms remain IV type with only slight
decrease in the adsorption and desorption volume. In addition, the
micropore areas as well as the micropore volume decrease by about
40 m2/g and 0.016 mL/g, respectively. The tiny change in micropore
specific area and pore volume suggests pore block, which is consistent
with the XRD result.
In order to find out reasons for blockage, TG-MS was adopted to
analyze the residues within reused HTS after 6 runs. The analysis results
are shown in Figs. 5 c and S2 to S6. The analysis was conducted with
the temperature rising from room temperature to 650 °C in an air at-
mosphere. The weight loss relative to the reused zeolite is about 12.5%
throughout the whole analysis process; meanwhile, the residues on the
HTS zeolite generate two main exothermic peaks on the DTA curve
during the heating procedure, which represent different desorption and
combustion situation (Fig. 5c). The ion current intensity of desorption
and combustion was recorded and analyzed by Mass as shown in Fig.
3.5. Influence of HCl aqueous solution treatment on HTS
Given the strong acid of HCl solution, the acid etching of the zeolite,
the stability of zeolite framework structure as well as the catalytic ac-
tivity are major points to be concerned. Dilute HCl solution is usually
employed for washing off the non-framework Ti species since it does
not destroy the framework structure of the zeolite [23]. However,
things could be different when concentrated HCl solution is employed.
The topology structure and relative crystallinity of HTS treated in
HCl solution were characterized by XRD (Fig. 6a). Since the treatment
was carried out at room temperature, the topology structure of the HTS
was retained with the characteristic peaks from 22° to 25° clearly re-
cognized. Perhaps owing to the non-framework Ti cleaning effect of HCl
to the zeolite pores, the relative crystallinity of HTS is slightly improved
with the increasing treatment time.
However, the HCl solution does have effects on the composition of
the zeolite (Table 3). During the first 25 days of treatment, the bulk Si/
Ti molar ratio of the HTS is only slightly increased. When the treatment
time is prolonged to 90 or 150 days, the bulk Si/Ti molar ratio is in-
creased to above 23. Meanwhile, after interaction with HCl for several
months, the surface Si/Ti molar ratio of HTS is remarkably increased
from 74.36 of HTS-0 to 118.12 of HTS-25, which indicates the loss of Ti
21

X. Peng et al.
AppliedCatalysisA,General543(2017)17–25
Fig. 5. Analysis results for the fresh and 6 runs reused HTS. (a) XRD; (b) N2 adsorption-desorption isotherms; (c) TGA and DTA curves of 6 runs reused HTS obtained by TG-MS; 13C CP/
MAS NMR spectra of 6 runs reused HTS and 1,3-DCP, solvent for 1,3-DCP analysis was D2O.
easily. Nevertheless, TieOeSi bond is much stronger than AleOeSi
bond since the AleOeSi bond cannot last for 2 h until it is fully hy-
drolyzed when it was treated under the same conditions (Fig. S7).
Heteroatom molecular sieve exhibits its fantastic structure stability
under strong acidic environment, and it is this stability that helped to
retain a certain content of Ti species. The structural stability combining
with a certain amount of active centers play a vital role in maintaining
high catalytic activity of AC chlorohydrination under strong acid con-
ditions.
Table 1
Textural properties of fresh and reused HTS.
Sample
Surface area/(m2/g)
Pore volume/(mL/g)
SBET
Sext
Smicro
Vtotal
Vmicro
Fresh HTS
1 Run reused HTS
6 Runs reused HTS
412
374
374
36
38
37
376
336
337
0.281
0.260
0.263
0.169
0.153
0.153
Although the HTS zeolite was treated for several months, its iso-
therm remains the IV type and is almost the same as fresh HTS (Fig. 6b).
Perhaps due to the cleaning effect of HCl solution, the micropore spe-
cific areas as well as micropore volume slightly increased (Table 4).
UV–vis spectroscopy and XPS were conducted to analyze the co-
ordination states of Ti species in the bulk phase and on the zeolite
surface (Fig. 7 and Table 2). UV–vis spectroscopy is very sensitive for
characterization of different Ti species [24,25]. Usually, the 4-co-
ordinated framework Ti species show absorbance at the wavelength of
210 nm, and the absorbance band ranging from 260 to 290 nm is as-
signed to 6- coordinated non-framework Ti species, while the absor-
bance band at about 330 nm is assigned to anatase species [26]. The
result from Fig. 7a suggests that the distribution of Ti species were not
obviously changed until the treatment was prolonged from one month
to three months. As the treatment advances, it is also observed that the
proportion of 4-coordinated Ti species is decreased whereas 6-co-
ordinated Ti species is increased. In other words, part of the framework
Ti species were eroded by the acid and slowly transferred from 4-co-
ordinated to 6-coordinated state. The non-framework Ti species on the
zeolite were continually leached by the acid and the fragments of newly
transformed combining with the already existed extra-framework Ti
species went into the liquid and thus they could be detected by ICP.
Table 2
13C NMR spectrum peak areas distribution of 6 runs reused
HTS.
Chemical shift
Peak area/%
82
72
60
47
34
Sum
3.48
18.90
6.86
36.70
34.06
100.00
species. This phenomenon is also evidenced by the ICP analysis of HCl
aqueous solution mixture. Despite of the strong acid environment, there
is almost no Si detection in HCl aqueous solution mixture, whereas the
Ti species is significantly dissolved by the HCl leaching. For the sample
of HTS-10 or HTS-25, there is only about 200 mg/L Ti species detected
in the filtrate, however, this value is remarkably increased by 3 times
for the sample of HTS-90 and it kept growing as increasing treatment
time. These results suggest that, compared with SieOeSi bond,
TieOeSi bond could be attacked and broke down by acid solution more
22

X. Peng et al.
AppliedCatalysisA,General543(2017)17–25
Fig. 6. XRD spectra (a) and N2 adsorption-desorption isotherms (b) of fresh and HCl treated HTS.
These two species can be easily distinguished from each other by XPS-
peak-fitting analysis. According to the result of Fig. 7b, the peak areas
of 4-coordinated Ti species are decreased while the peak areas of 6-
coordinated Ti species are increased. The relative value of the two peak
areas shown in Table 3 reconfirms the decrease of 4-coordinated Ti
species. The analysis of surface Ti species content and proportion of Ti
species with different coordination states indicate that, owing to the
erosion of HCl solution to the HTS zeolite, surface framework Ti-O-Si
bonds are slowly hydrolyzed and broke apart. Therefore, the co-
ordination states of framework Ti species are transferred from 4-co-
ordinated into 6-coordinated, and meanwhile, as the treatment went
on, these Ti species were fully dissociated from the framework and
became non-framework Ti (Scheme S4). Parts of these non-framework
Ti species would further transfer from the zeolite surface into the liquid,
which resulted in the decrease of surface Ti content.
The change in chemical composition of the zeolite was also recorded
by 29Si CP/MAS NMR and FT-IR (Fig. 8). As for the HTS-25 and HTS-
150, there are no obvious change of the Q4 species [29,30] peak on the
29Si CP/MAS NMR spectra. While the peak intensity for Q3 species in-
creased remarkably for HTS-150, which could be explained by the
cleavage of TieOeSi bond and leaving a Q3 species. Furthermore, if the
Ti species are fully dissociated and left 4 Q3 species, then a hydroxyl
nest is formed. FT-IR is very useful to detect the hydroxyl groups. The
sharp peak around 3740 cm−1 represents terminal silanol and the
broad band around 3530 cm−1 represents nest silanol. As illustrated in
Fig. 8b, the peak area of terminal silanol groups especially the nest
silanol groups for HTS-150 increases significantly, which confirms the
dissociation of the active centers evidenced by 29Si CP/MAS NMR
spectra.
Table 3
Structural and composition of fresh and HCl treated HTS.
Sample
Si/Ti
Bulka
Concentration/(mg/L)c
6-/4-b,e
Surfaceb
Ti
Si
HTS-0
19.91
20.39
20.85
23.50
23.95
74.36
75.97
118.12
96.52
100.71
< 1.0d
198
1.1d
1.4
0.58
0.85
0.75
1.20
1.43
HTS-10
HTS-25
HTS-90
HTS-150
218
826
< 1.0
1.4
1003
2.1
Notes: (a) XRF; (b) XPS; (c) ICP; (d) treated by water; (e) 6-coordinated Ti/4-coordinated
Ti.
Table 4
Structural parameters of fresh and HCl treated HTS calculated from N2 adsorption-des-
orption isotherms.
Sample
Surface area/(m2/g)
Pore volume/(mL/g)
SBET
Sext
Smicro
Vtotal
Vmicro
HTS-0
416
417
415
413
411
66
35
37
37
37
350
381
378
376
374
0.272
0.288
0.301
0.277
0.277
0.153
0.172
0.171
0.171
0.170
HTS-10
HTS-25
HTS-90
HTS-150
Anatase species are inertia and uneasy to react with HCl, so they were
less supposed to be converted into other forms.
The XPS results supplies further information of Ti species transfor-
mation (Fig. 7b). The binding energy of 4-coordinated Ti species is
about 460 eV while 6-coordinated Ti species is about 459 eV [27,28].
The catalytic performance of HCl solution treated HTS was eval-
uated and the results are shown in Fig. 9. Although the HTS zeolite
Fig. 7. Chemical state analysis of Ti species on fresh and HCl treated HTS. (a) UV–vis; (b) XPS.
23

X. Peng et al.
AppliedCatalysisA,General543(2017)17–25
Fig. 8. Analysis results of Si chemical state for the fresh and HCl treated HTS. (a) 29Si NMR; (b) FT-IR.
endured HCl treatment as long as five months, and loss of parts fra-
mework Ti species is also confirmed by various characterization
methods, the catalyst still maintains high catalytic performance. The AC
conversion doesn’t decrease obviously until the HTS is treated for 5
months, which reflects the strong ability of HTS zeolite toward re-
sistance to leaching and maintaining high catalytic activity. The DCP
content and selectivity did not significantly change along with the
treatment time. However, the increase in TCP selectivity for HTS-90
and HTS-150 should be associated with the transformation of Ti species,
as 6-coordinated Ti species are observed and favored to promote
chlorination reaction while 4-coordinated ones promotes the oxidation
reaction in the other way. The slight TCP increase for the reusability
test of HTS might also be explained by the increased 6-coordinated Ti
species during the consecutive runs. Due to the completely different
mechanism and products distribution between traditional and this
novel chlorohydrination, a parameter of which a ratio of 1,3-DCP se-
lectivity to 2,3-DCP selectivity is introduced to distinguish novel
chlorohydrination and traditional chlorohydrination process in this
reaction system. For example, the ratio for HTS-0 and HTS-150 calcu-
lated from Fig. 9(b) are 34 and 25, respectively, which represents for
the HTS-150, the Cl2-based chlorohydrination route is strengthened.
The catalytic performance results reveal that the extraordinary beha-
vior of HTS in acid solution exhibits its great potential to be applied in
our reaction system for long term use.
(88%) is achieved. Owing to a synergistic effect of Lewis acidity of
framework Ti species within HTS and Brønsted acidity of HCl, the in-
termediate product resulted from AC oxidation by H2O2 over HTS is
immediately transferred into DCP in the presence of HCl through pro-
tonation and nucleophilic reaction. Based on the unique mechanism,
the reaction is promoted from AC to DCP rapidly, and AC is almost fully
converted as well as H2O2 with DCP selectivity above 99%. Moreover,
the DCP content reaches 22 wt%, which is 5 times higher than the
traditional Cl2-based process of 4 wt%. It is also found that the HTS
zeolite is very stable and robust towards HCl treatment. Although
leaching and transformation of framework Ti species is observed, the
topology structure is undestroyed and the remaining active center is
sufficient to maintain a high catalytic performance for the DCP synth-
esis. This work provides a mild approach for effective olefins chlor-
ohydrination, which has great potential to be applied on a large scale.
Acknowledgements
We thanks for the financial support of the National Program on Key
Basic Research Project (973 Program) by the Ministry of Science and
Technology of People’s Republic of China (2006CB202508), and
SINOPEC Corporation (S413108). Thanks to Dr. Limei Qiu, Dr. Kang
Zou, Lijun Lu, Dr. Hui Yuan, Dr. Yanjuan Xiang, Jin Zhang, Mudi Xin,
Dr. Yupeng Xu, Zhihua Zhang, Yanqiang Shi and prof. Guangtong Xu for
the analysis support from Analysis Department of Research Institute of
Petroleum Processing.
4. Conclusions
With liquid Cl2 and organic solvent free, a novel AC chlorohy-
drination route via the reaction of AC with H2O2 and HCl solution over
HTS zeolite is proposed, through which a much safer and greener DCP
synthesis method with higher product content and atom economy
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.apcata.2017.06.004.
Fig. 9. AC chlorohydrination evaluation results of HCl treated HTS. (a) The conversion of AC and content of DCP and (b) products selectivity on treatment time. Reaction conditions:
HCl:H2O2:AC = 2:1:1 (mol); concentration of HCl solution, 20 wt%; temperature,30 °C; HTS concentration, 2 wt%; time, 2 h.
24

X. Peng et al.
AppliedCatalysisA,General543(2017)17–25
[15] X. Peng, C. Xia, M. Lin, H. Yuan, B. Zhu, Y. Zhang, B. Wang, X. Shu, Green Chem. 19
(2017) 1221–1225.
References
[16] D.P. Das, K. Parida, Catal. Commun. 7 (2006) 68–72.
[17] R.K. Pandey, R. Kumar, Catal. Commun. 8 (2007) 379–382.
[18] L. Wang, Y. Liu, W. Xie, H. Zhang, H. Wu, Y. Jiang, M. He, P. Wu, J. Catal. 246
(2007) 205–214.
[19] A. Wróblewska, E. Milchert, Org. Process Res. Dev. 10 (2006) 525–533.
[20] R.K. Pandey, R. Kumar, Catal. Commun. 8 (2007) 379–382.
[21] H. Kwart, A.L. Goodman, J. Am. Chem. Soc. 82 (1960) 1947–1949.
[22] N. Wang, X. Wang, L. Yang, L. Chen, W. Li, J. Northeast Normal Univ. (Natural
Science Edition) 44 (2012) 95–100.
[23] Q. Guo, Z. Feng, G. Li, F. Fan, C. Li, J. Phys. Chem. C 117 (2013) 2844–2848.
[24] M. Boccuti, K. Rao, A. Zecchina, G. Leofanti, G. Petrini, Stud. Surf. Sci. Catal. 48
(1989) 133–144.
[25] A. Zecchina, S. Bordiga, C. Lamberti, G. Ricchiardi, D. Scarano, G. Petrini,
G. Leofanti, M. Mantegazza, Catal. Today 32 (1996) 97–106.
[26] P. Ratnasamy, D. Srinivas, Adv. Catal. 48 (2004) 1–169.
[27] A. Stakheev, E. Shpiro, J. Apijok, J. Phys. Chem. 97 (1993) 5668–5672.
[28] S. Vetter, G. Schulz-Ekloff, K. Kulawik, N.I. Jaeger, Chem. Eng. Technol. 17 (1994)
348–353.
[1] R.A. Sheldon, J.A. Van Doorn, C.W.A. Schram, A.J. De Jong, J. Catal. 31 (1973)
438–443.
[2] G.S. Sharifov, Kinet. Catal. 42 (2001) 615–619.
[3] J. Zhang, J. Tan, K. Wang, Ya. Lu, G. Luo, Ind. Eng. Chem. Res. 51 (2012)
14685–14691.
[4] J. Zhang, Y. Lu, Q. Jin, K. Wang, G. Luo, Chem. Eng. J. 203 (2012) 142–147.
[5] R. Tesser, E. Santacesaria, M. Di Serio, G. Di Nuzzi, V. Fiandra, Ind. Eng. Chem. Res.
46 (2007) 6456–6465.
[6] E. Santacesaria, R. Tesser, M. Di Serio, L. Casale, D. Verde, Ind. Eng. Chem. Res. 49
(2010) 964–970.
[7] L. Wang, Y. Liu, W. Xie, H. Zhang, H. Wu, Y. Jiang, M. He, P. Wu, J. Catal. 246
(2007) 205–214.
[8] M.G. Clerici, P. Ingallina, J. Catal. 140 (1993) 71–83.
[9] Y. Sun, Z. Xi, G. Cao, J. Mol. Catal. A: Chem. 166 (2001) 219–224.
[10] J. Lin, F. Xin, L. Yang, Z. Zhuang, Catal. Commun. 45 (2014) 104–108.
[11] A. Wróblewska, J. Mol. Catal. A: Chem. 229 (2005) 207–210.
[12] W. Kim, C. Yun, Y. Kim, J. Park, S. Park, K.T. Jung, Y.H. Lee, S.H. Kim, Ind. Eng.
Chem. Res. 50 (2011) 1187–1195.
[29] A. Tuel, Y.B. Taarit, J. Chem. Soc. Chem. Commun. 21 (1992) 1578–1580.
[30] W. Fan, R. Duan, T. Yokoi, P. Wu, Y. Kubota, T. Tatsumi, J. Am. Chem. Soc. 130
(2008) 10150–10164.
[13] C. Xia, M. Lin, A. Zheng, Y. Xiang, B. Zhu, G. Xu, X. Shu, J. Catal. 338 (2016)
340–348.
[14] M. Lin, C. Xia, B. Zhu, H. Li, X. Shu, Chem. Eng. J. 295 (2016) 370–375.
25