W. Jiao et al. / Applied Catalysis A: General 491 (2015) 78–85
79
Root have demonstrated by theoretical calculations that choice of
solvent has no effect on the reaction activation energy for ethylene
epoxidation and the formation of titanium hydroperoxo species by
chemisorption of H2O2 on TS-1 [17,18].
a 60 m OV-1 capillary column. The internal standard was cyclohep-
tanone. The unconverted H2O2 was determined by titrating with
0.1 M Ce(SO4)2 aqueous solution.
Jacobs and Broadbelt found that the oxidation rate of 1-
hexene over TS-1 in alcoholic solvents decreased in the order:
shows that the reaction kinetic difference in different solvents
is perhaps caused by the difference of substrate concentration
between the intraporous and the extraporous [20]. Clerici and
coworkers [12] found that solvents not only participated in the
reaction (active role), but also promoted the adsorption/desorption
of substrates/products on TS-1 (passive role).
2.3. Adsorption of MAC on Ti-rich TS-1 in different solvents
The adsorption amount and rate of MAC on Ti-rich TS-1 in the
methanol, acetonitrile and TBA solvents free of H2O2 at 20 ◦C were
analyzed by the same gas chromatograph as that used for the anal-
ysis of liquid product. The same amount of doubly distilled water
as that contained in the H2O2 aqueous solution was added in the
solvents in order to consider the role of water. The mesitylene was
used as the internal standard.
Agnieszka and coworkers [1] have studied the epoxidation of
MAC on TS-1 in methanol solvent, showing that the MAC conver-
sion reached only 20% under the optimal conditions. This value is
not economic from the commercial viewpoint. The present work
aims to significantly increase the catalytic activity of titanosilicates
for the epoxidation of MAC by investigating the catalytic reaction
mechanism. It will be shown that Ti-rich TS-1 gave a MAC con-
version of > 85% with epoxide selectivity > 96% in tert-butyl alcohol
(TBA) solvent, and the reaction process is probably dominated by
an Eley–Rideal-type catalytic mechanism.
3. Results and discussion
3.1. Characterization of various titanosilicate catalysts
The XRD measurements show that all the synthesized titanosil-
a strong adsorption band below 220 nm and an intense vibration
band at about 955 cm−1 (TS-1 and Ti-Beta) or 930 cm−1 (Ti-MWW
and Ti-YNU-1), which are characteristic of tetrahedral Ti species in
titanosilicates [8,22]. The specific surface areas of TS-1, Ti-MWW,
Ti-Beta and Ti-YNU-1 are 463, 448, 502 and 660 m2 g−1, as deter-
mined by N2 sorption at −196 ◦C.
2. Experimental
2.1. Catalyst preparation and characterization
TS-1 catalyst was synthesized with tetraethyl silicate (TEOS),
tetrabutyl orthotitanate (TBOT), tetrapropyl ammonium hydroxide
(TPAOH), ammonium carbonate ((NH4)2CO3) and deionized water
according to our previously reported procedures [8]. The chemi-
cal compositions of the resultant synthesis gels were as follows:
0.2(NH4)2CO3:SiO2:(0.01–0.05)TiO2:0.5TPAOH:35H2O. The crys-
samples were treated with 1 M HCl acid at room temperature (RT),
calcined at 560 ◦C and further washed with 1 M HCl acid at RT. Ti-
MWW was synthesized by the postsynthesis method using highly
deboronated B-MWW as silica source [21] and Ti-YNU-1 was syn-
thesized according to the procedures reported in the ref. [22].
Ti-Beta was synthesized by the seeding method following the pro-
cedures reported by Camblor and coworkers [23].
The framework Ti amount was determined by an induc-
tively coupled plasma-atomic emission spectroscopy (ICP-AES,
Autoscan16, TJA). The X-ray diffraction patterns (XRD) were
recorded on a Rigaku Mini Flex II desktop X-ray diffractometer
with CuK␣ radiation (30 kV, 15 mA). N2 sorption isotherms were
measured on a BELSORP-max instrument. Before the measure-
ment, the sample was first evacuated at 300 ◦C for 8 h under high
vacuum conditions (<10−4 mbar). The DR UV–vis spectra were
measured by a Shimadzu UV-3600 spectrophotometer equipped
with an integration sphere. The NH3-TPD measurements were
performed on a Micromeritics Autochem II 2920 chemisorption
analyzer.
Epoxidation of MAC over titanosilicates generates MECH and
diol or ester which forms through hydrolysis or solvolysis of MECH
(Scheme S1). Table 1 summarizes the catalytic results obtained in
the oxidation of MAC over various titanosilicates in different sol-
vents. Although Ti-MWW is intrinsically more active than TS-1 for
the epoxidation of linear alkenes, allyl alcohol, diallyl ether and allyl
chloride [24–27], it gave much lower MAC conversion and H2O2
efficiency than Ti-rich TS-1 regardless of the solvent used (Table 1).
This also held true for Ti-Beta and Ti-YNU-1. It is surprising that
both hydrophobic TS-1 and hydrophilic Ti-MWW, Ti-YNU-1 and
Ti-Beta exhibited higher activity in TBA solvent than in the gener-
ally used methanol or acetonitrile. This reveals the dependence of
the solvent effect of titanosilicates on the substrate molecules.
The MAC conversion nearly linearly increased with the Ti
amount while the MECH selectivity was kept constant, being about
96% (Fig. S1). This indicates that almost all of the Ti species in the
samples were incorporated in the framework, as confirmed by the
DR UV–vis spectroscopy result that all of these samples contain
negligible amounts of extraframework Ti species. As a result, the
similar TON was obtained for all the samples.
Ti-rich TS-1 for epoxidation of MAC
3.3.1. Influence of solvents
Table 2 compared the catalytic results of Ti-rich TS-1 for the
epoxidation of MAC in various solvents. As expected, protic MeOH
solvent gave higher MAC conversion than aprotic MeCN, MeCOMe
molecular size of alcohols such as ethanol (EtOH), 2-propanol (2-
PrOH) and 1-butanol (1-BuOH) resulted in a severe decrease in
the MAC conversion. This seems to support that species I is the
active intermediate [2]. However, it is unexpected that a further
increase in the molecular size of alcohol solvent contrarily greatly
increased the activity. TBA gave a MAC conversion and epox-
ide selectivity as high as 85.4% and 96.7% respectively. It is far
2.2. Catalytic test
The epoxidation of MAC with H2O2 was carried out under stir-
ring conditions in a 20-mL glass flask equipped with a condenser.
The reaction temperature was controlled with a water bath. Unless
specified, the reaction conditions were as follows: 0.05 g catalyst,
5 mmol MAC, 5 mmol H2O2 (35% aqueous solution) and 5 mL sol-
vent. The product was analyzed by a Shimadzu GC-2010 plus gas
chromatograph equipped with a flame ionization detector (FID) and