S. Miao, B.H. Shanks / Journal of Catalysis 279 (2011) 136–143
137
et al. [3] suggested that the gas-phase reaction of acetic acid with
ethanol on MCM-41 was performed through a dual-site mecha-
nism in which both acetic acid and ethanol must be adsorbed on
the surface for the reaction to occur. However, Chu et al. [11] re-
ported that alcohol structure had a profound effect on the mecha-
nism of the gas-phase esterification of acetic acid changing from a
dual-site mechanism for ethanol to a single-site mechanism for n-
butanol.
min under flowing air. The ion capacities of the sulfonic acid
groups in the functionalized mesoporous silica were quantified
using 2 M NaCl (aq) as the ion-exchange agent. Approximately
0.05 g of the sample was added to 30 ml of the salt solution and al-
lowed to equilibrate for 18 h. Thereafter, it was titrated by drop-
wise addition of 0.005 M NaOH (aq) [15].
2.3. Catalytic reactions
In this study, a heterogeneous acid catalyst, propylsulfonic acid-
functionalized mesoporous silica, was prepared by a co-condensa-
tion method and employed as the catalyst for the kinetics study of
esterification of acetic acid. This reaction system has potentially
important application in the upgrading of fast pyrolysis-derived
bio-oil for fuel applications [12,13]. Mesoporous silica materials–
supported organic acid is an attractive heterogeneous acid catalyst
due to its high surface area, good control of the acid moieties, no
significant Lewis acidic sites and good stability without swelling
in organic solvents. The focus of the present study was to provide
fundamental insight into the similarities and differences existing
between heterogeneous and homogeneous Brønsted acid catalysts
in the esterification reaction. Propane sulfonic acid was chosen as
the homogeneous acid catalyst for comparison since it bears the
same structure as the grafted functional groups in the above heter-
ogeneous catalyst.
The kinetic measurement of esterification of acetic acid with
methanol was carried out in a stirred batch reactor, which was
placed in a thermostatic bath with a magnetic stirrer. The tempera-
ture in the reactor was maintained within ±0.2 K. The reactor was
charged with a measured amount of reagents (acetic acid 3 M, meth-
anol 6 M, as well as the solvent 1,4-dioxane to balance the total vol-
ume to 50 ml). When the mixture was heated to the desired reaction
temperature, the first sample was taken as the zero point for every
run, after which an exact amount of catalyst was added to initiate
the reaction. In all cases, a microscale syringe was used for sampling
at definite time intervals. Stirring speed of 600 rpm was applied to
rule out mass transfer limitation as previously reported [12]. Kinetic
measurement was performed at low acetic acid conversion (<10%).
Pyridine adsorption experiments were carried out by immersing a
known amount of catalyst in 1,4-dioxane containing a known
amount of pyridine overnight. The reaction started by heating the
above mixture to the desired temperature and charging the pre-
heated acetic acid and methanol. The pre-adsorption experiments
were performed by premixing the catalyst with one of the reactants
or both and solvent at ambient temperature overnight followed by
heatingtothedesiredtemperatureandchargingthepreheatedother
one. For the pre-adsorption of catalyst with both acetic acid and
methanol, small amounts of acetic acid and methanol (0.5 g acetic
acid and 0.5 g methanol) were used with 1,4-dioxane to premix with
catalyst overnight and the remaining were charged to start the reac-
tion. An Agilent GC 7890A gas chromatograph equipped with a HP-5
2
. Experimental
2.1. Chemicals and synthesis
SBA-15-functionalized organosulfonic acid materials were syn-
thesized by one-pot co-condensation method as described previ-
ously [14–16]. Tetraethoxysilane (TEOS) (98%, Aldrich) was used
as the silica precursor, and (3-mercaptopropyl)trimethoxysilane
(
MPTMS) (85%, Acros) was used without further purification as
the organosulfonic acid source. Pluronic P123 (BASF Co., USA),
which is a tri-block copolymer of polyethylene oxide–polypropyl-
ene oxide–polyethylene oxide, was used as obtained to tailor the
textural properties of the mesoporous materials. In a typical one-
step synthesis, 4 g of Pluronic P123 was dissolved in 125 g of
column (0.32 mm ꢀ 30 m ꢀ 0.25
lm) and a FID detector was used
for sample analysis. The concentrations of all species except water
were directly quantified. The overall mass balance was more than
8%.
9
1
.9 M HCl at room temperature with continuous stirring. The solu-
tion was subsequently heated to 313 K before adding TEOS. As
usual, TEOS was pre-hydrolyzed for approximately 45 min prior
2.4. Computational method
to the addition of the MPTMS-H
tion of the resulting mixture was 0.0369 TEOS, 0.0062 MPTMS,
and 0.0554 H . It was continuously stirred for 24 h at 313 K
and thereafter aged for 24 h at 373 K under static conditions. The
product was collected and subjected to ethanol refluxing for three
cycles for the extraction of the template. The final product was vac-
uum dried at 373 K for 6 h.
2
O
2
solution. The molar composi-
Computational modeling was used to examine whether the
adsorption behavior observed experimentally was consistent with
molecular energetic. A simple representation of the catalyst sur-
face was used in which the central Si atom was bound to the teth-
ered organosulfonic acid and three O atoms, which were also
bonded to a second Si atom. These terminal Si atoms were satu-
rated with hydrogen atoms at a Si–H distance of 1.46 Å in the cal-
2 2
O
4 3
culations. The (SiH ) groups were held fixed, while the other
atoms were allowed to relax during the structure optimization.
None of the atoms in the adsorbed acetic acid, methanol, and water
molecules were constrained in the optimization of the adsorption
complexes. The calculations, which included structure optimiza-
tion and single-point energies, employed the hybrid density func-
tional B3LYP method with standard DZVP2 basis sets. All of the
calculations were performed using the Gaussian03 program
package.
2
.2. Sample characterization
Nitrogen adsorption–desorption isotherms were measured at li-
quid nitrogen temperature with a Micromeritics ASAP 2020 sys-
tem. Prior to measurement, all samples were degassed at 373 K
for 6 h. The specific surface areas were evaluated using the Bru-
nauer–Emmett–Teller (BET) method and pore size distribution
curves were calculated using the desorption branch of the iso-
therms and the Barrett–Joyner–Halenda (BJH) method. The content
of organic material present in the solids was determined by ele-
mental analysis performed on a Perkin-Elmer Series II 2400 CHNS
analyzer. The decomposition temperature of the organic composi-
tion in the modified mesoporous materials was determined by
thermogravimetric analysis (TGA) with a Perkin-Elmer TGA7
instrument, with heating from 323 to 973 K at a ramp of 10 K/
3. Results and discussion
The propylsulfonic acid-functionalized SBA-15 materials used
in the study were analyzed for their textural properties. The N
2
adsorption–desorption isotherms of the samples had type IV hys-
teresis loops with sharp adsorption and desorption curves, as seen