86
T.Y. Kim et al. / Journal of Catalysis 323 (2015) 85–99
studies on the catalytic conversion of vicinal diols can provide a
basis for the conversion of biomass-derived feedstocks.
to express the SiO2-supported Cs catalysts: xCs/SiO2-y, where x is
the wt.% of Cs2O loading and y is the calcination temperature
(723–1023 K). The catalysts prepared using other supports are
expressed in the same way.
However, despite the importance of dehydration and vicinal
diols, researches on the catalytic dehydration of vicinal diols have
received little attention, and only few studies have been reported
on this issue. Several studies on dehydration of 2,3-butanediol to
3-buten-2-ol were reported [25], but most of the reports related
to this subject have been limited to acid-catalyzed dehydration,
yielding aldehydes and ketones [26–32]. Such restricted uses and
studies will not be sufficient to identify efficient reaction pathways
for biomass conversion for replacing conventional petrochemical
products. In this regard, the discovery of novel dehydration reac-
tions of vicinal diols has the potential to open up new possibilities
for the utilization of biomass, and thereby, has the potential to
expand the scope of chemicals that can be produced from biomass.
Herein, we report on a novel dehydration reaction of vicinal
diols, leading to the formation of epoxides. Since the reaction
involves the formation of an epoxide ring with the elimination of
H2O, we refer to the reaction as ‘‘dehydrative epoxidation.’’ We
first compared the Cs/SiO2 catalyzed reaction of vicinal diols,
including ethylene glycol (1), 1,2-propanediol (2), and 2,3-butane-
diol (3), with previously reported dehydration reactions. The diol 3
was then used as a representative model compound to explore the
dehydrative epoxidation in more detail. A series of active catalysts
were screened, and the optimum catalyst for the dehydrative epox-
idation was identified. The acidic–basic and physicochemical prop-
erties of the catalysts were characterized, in order to identify the
cause of the difference in catalytic activity with respect to dehyd-
rative epoxidation. In addition, the stereochemistry involved in
dehydrative epoxidation was explored. DFT calculations were car-
ried out to determine the energetics of the reaction, including the
transition states as well as the configurations and adsorption of the
molecules. On the basis of the experimental results and the DFT
calculations, the active site of the catalyst and a reaction mecha-
nism for dehydrative epoxidation are proposed.
2.2. Reactivity tests
A fixed-bed quartz reactor was used to evaluate the catalytic
performances of the catalysts for the dehydration of 2,3-butanediol,
1,2-propanediol, and ethylene glycol. Temperature was measured
by a K-type thermocouple and was controlled with external electri-
cal furnace. Reactions were performed using 0.1 g of catalyst sam-
ples held on a porous quartz bed with aqueous solutions of 2,3-
butanediol (Acros Organics, 98%, mixture of racemic and meso
forms), meso-2,3-butanediol (Sigma–Aldrich), 1,2-propanediol, eth-
ylene glycol (Alfa Aesar, 97%). The compositions of reactants in the
aqueous solutions were 9.9 wt.% (2.1 mol%) for 2,3-butanediol,
8.3 wt.% (2.1 mol%) for 1,2-propanediol, and 6.9 wt.% (2.1 mol%)
for ethylene glycol in order to ensure a same space velocity of reac-
tants (Space velocity (SV) = total flow rate/amounts of cata-
lyst = 1.38 L/(min(g cat)). The reactor was heated to the desired
temperatures (673 K) with ramping rate of 10 K/min and main-
tained for 30 min with a flow of dry N2 (99.999%, 30 cm3/min).
The reactant solutions were then injected with a rate of 1.5 cm3/h
(1.64 mmol 2,3-butanediol/h) into a pre-heating zone which was
maintained at 473 K and was connected to a top of the reactor.
Effluent gases were passed through a helix-type condenser and
were collected in a sample tube containing 20 ml of deionized
water. Acetonitrile was used as an external standard for quantifica-
tion and added to products collected hourly. The products were
analyzed using gas chromatography (Younglin ACME 6100 instru-
ment) equipped with a FID detector and RtxÒ-VRX capillary column
(Restek, cat. # 19316). The data acquired at 2 h of the reaction were
used to compare the catalytic activity of the catalysts. The turnover
frequency (TOF) and specific formation rates of products for sup-
ported catalysts were calculated by the following equations:
ꢀ
ꢁ
2. Experimental and theoretical methods
Converted amounts of reactant ðmol=hÞ
Loaded amounts of metal ðmolÞ
Turnover frequency hꢀ1
¼
2.1. Preparation of catalysts
ꢀ
ꢁ
Formed amounts of productðmol=hÞ
Loaded amounts of metalðmolÞ
Specific formation rate hꢀ1
¼
Sodium acetate trihydrate (TCI), potassium oxalate monohy-
drate (Samchun Chemical), cesium acetate (Samchun Chemical),
magnesium nitrate hexahydrate (Fluka), calcium nitrate tetrahy-
drate (Sigma–Aldrich), strontium nitrate (Sigma–Aldrich), lantha-
num nitrate hexahydrate (Junsei Chemical), and cerium nitrate
hexahydrate (Kanto chemical) were used as precursors for
supported basic metal oxide catalysts and were used without fur-
ther purification. Al-MCM-41 (Aluminosilicate, mesostructured,
Sigma–Aldrich) was used as received. The catalysts were prepared
by the incipient wetness impregnation method. In typical proce-
dures, at first, a predetermined amount of precursor was dissolved
in 3 ml of deionized water. The aqueous solution of precursor was
then added dropwise to 2.0 g of a support oxide (SiO2 (AerosilÒ
200), Al2O3 (AeroxideÒ Alu C), and TiO2 (AeroxideÒ P25)) with vig-
orous mixing by hand for at least 20 min. To prevent a complete
wetness, impregnation was conducted repeatedly by dosing a
small quantity of precursor solution and followed by drying at
room temperature. The impregnated powders were dried at
343 K overnight, and then be grounded and calcined at desired
temperature (5 K/min) for 4 h.
2.3. Characterization
Temperature-programmed desorption (TPD) of NH3 was carried
out using Micromeritics Autochem II chemisorption analyzer. Prior
to the analysis, 0.1 g of sample was heated at 673 K for 1 h under a
He flow to remove adsorbed impurities. After cooling to 323 K, the
sample was saturated with probe by a flow of 10.2% NH3/He. The
physisorbed probe was removed by flushing of He flow at 373 K.
After the sample was cooled down to 323 K and the TCD signal
was stabilized, the signal was recorded with increasing the tem-
perature from 323 K to 673 K at a rate of 10 K/min under a flow
of He.
The Hammett indicator titration method was utilized to com-
pare the base strength of the prepared catalysts. In typical proce-
dure, the Hammett indicators were dissolved in methanol to
obtain 0.05 M solutions. Prior to the titration, a determined
amounts (10 mg) of a catalyst were heated at 423 K in vacuum
For the catalysts used in the screening of active materials (Sec-
tion 3.2), 10 wt.% (Cs2O, CaO, SrO, La2O3, CeO2), 2.4 wt.% (Na2O),
3.6 wt.% (K2O), and 3.1 wt.% (MgO) were loaded on a SiO2 support,
and the notation of M/SiO2 is used. Calcination was carried out at
823 K for 4 h. In Sections 3.3 and 3.4, the following notation is used
oven to remove adsorbed species. Then, it was titrated by 400 ll
of each indicator solution, and color change was recorded.
Following four indicators were used: neutral red (pKa = 6.8),
phenolphthalein (pKa = 8.2), 2,4-dinitroaniline (pKa = 15.0), and