High-temperature reaction of SiO2 with methanol: Nucleophilic assistance of some N-unsubstituted benzazoles

Article abstract of DOI:10.1016/j.apcata.2013.03.001

SiO₂-containing materials (quartz, Pyrex glass, silica gel, and H-Y zeolite) react slowly with methanol at 350 C under both supercritical and gas-phase conditions. The amount of SiO₂ reacted with supercritical methanol depends on the kind of the material and is varied from 0.05 wt% for quartz sand to 4.0 wt% for wide-porous silica gel for 5 h of the reaction. The main products of the reaction are methyl orthosilicates, mainly tetramethyl orthosilicate. N-Unsubstituted 1H-indole, 1H-benzimidazole, and 1H-indazole additives considerably increase the amount of reacted SiO₂-containing material. Noteworthy, quartz sand solubility is increased ca 14-fold when indole is used at the same reaction conditions. These benzazoles provide a nucleophilic assistance to the reaction between SiO₂ and methanol. During the assistance, the azole ring is methylated by methanol, and the participation of SiO₂-surface or tetramethyl orthosilicate molecule facilitates the alkylation reaction of benzazole used.

Full text of DOI:10.1016/j.apcata.2013.03.001

Applied Catalysis A: General 456 (2013) 159–167
Contents lists available at SciVerse ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
High-temperature reaction of SiO2 with methanol: Nucleophilic
assistance of some N-unsubstituted benzazoles
Andrey M. Chibiryaeva,b,∗, Ivan V. Kozhevnikovb,c,∗∗, Oleg N. Martyanovb,c
a Vorozhtsov Institute of Organic Chemistry, Siberian Branch of the Russian Academy of Sciences, Acad. Lavrent’ev ave. 9, 630090 Novosibirsk, Russia
b Novosibirsk State University, Pirogov st. 2, 630090 Novosibirsk, Russia
c Boreskov Institute of Catalysis, Siberian Branch of the Russian Academy of Sciences, Acad. Lavrent’ev ave. 5, 630090 Novosibirsk, Russia
a r t i c l e i n f o
a b s t r a c t
Article history:
SiO2-containing materials (quartz, Pyrex glass, silica gel, and H-Y zeolite) react slowly with methanol at
350 ◦C under both supercritical and gas-phase conditions. The amount of SiO2 reacted with supercritical
methanol depends on the kind of the material and is varied from 0.05 wt% for quartz sand to 4.0 wt% for
wide-porous silica gel for 5 h of the reaction. The main products of the reaction are methyl orthosilicates,
mainly tetramethyl orthosilicate. N-Unsubstituted 1H-indole, 1H-benzimidazole, and 1H-indazole addi-
tives considerably increase the amount of reacted SiO2-containing material. Noteworthy, quartz sand
“solubility” is increased ca 14-fold when indole is used at the same reaction conditions. These benzazoles
provide a nucleophilic assistance to the reaction between SiO2 and methanol. During the assistance, the
azole ring is methylated by methanol, and the participation of SiO2-surface or tetramethyl orthosilicate
molecule facilitates the alkylation reaction of benzazole used.
Received 24 December 2012
Received in revised form 26 February 2013
Accepted 1 March 2013
Available online xxx
Keywords:
SiO2-containing materials
Supercritical methanol
Tetramethyl orthosilicate
Benzazole methylation catalyst
Nucleophilic assistance
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
process, MTO) [8–12], or dimethyl ether [13]. All of these catalytic
processes have common feature since they occur in gas phase at
Traditionally, silicas (SiO2) are a convenient and widely used
supports for different catalysts in organic synthesis. The reasons for
such popularity are their chemical inertness, high thermal stability,
enhanced mechanical strength and capability to use it as granules
or powder, easy regulation of pore sizes and specific surface areas,
ability to absorb effectively water and/or different organic com-
pounds, multiple regenerations without considerable loss of initial
activity. Due to these physicochemical properties of SiO2, quartz
or Pyrex glassware is traditionally used for chemical reactions, in
which chemists try to avoid the effects of reactor’s wall material.
Silica-supported catalysts are commonly used in many impor-
tant industrial processes such as oxidative coupling of methane
(OCM) [1] or partial oxidation of the last into methanol and
formaldehyde [2], syntheses from syngas to dimethyl ether [3], oxi-
dized C2-compounds [4], or higher alcohols [5], transformations
of methanol into hydrocarbons [6,7], olefins (methanol-to-olefin
temperature from 300 up to 800 ◦C and one of the chemical com-
ponents of the reaction mixture is methanol (together with SiO2).
These temperatures are above the critical temperature of methanol.
In the supercritical state, methanol exhibits high reactivity and can
be used as an alkylating agent for catalytic N-/O-methylation of
amines [14], diols [15], and phenols [16]. It is, however, not known
if is SiO2 indeed inert under supercritical and gas-phase conditions
at high temperature in the presence of methanol?
In earlier studies [17,18], significant differences were found
for oxidative transformation of methane to methanol when oxi-
dation was conducted in stainless-steel or Pyrex-lined reactors
at temperatures between 375 and 500 ◦C. Higher conversion
and selectivity were obtained in the reaction carried out in
Pyrex-lined (and quartz-lined too) reactor. The real origin of
the differences remained unknown. Other studies have demon-
strated that methoxide species CH3O− can be detected in
methanol oxidation reaction over silica-supported catalyst, for
example, over molybdenum oxide [19,20]. However, methoxide
anion could not be formed by oxidation of methanol molecule.
Only reduction (CH3OH → CH3O− + [H•]) or thermal dissociation
(CH3OH → CH3O− + H+) can lead to methoxide species forma-
tion. At the same time, these species can be formed easily,
for instance, by thermal dissociation of tetramethyl orthosilicate
(Si(OCH3)4 → CH3O− + (CH3O)3Si+). It is well known that chemi-
cal modification of silica by some alcohols at high temperature
∗
Corresponding author at: Vorozhtsov Institute of Organic Chemistry, Siberian
Branch of the Russian Academy of Sciences, Acad. Lavrent’ev ave. 9, 630090 Novosi-
birsk, Russia. Tel.: +7 383 326 96 89; fax: +7 383 330 97 52.
∗∗
Corresponding author at: Boreskov Institute of Catalysis, Siberian Branch of the
Russian Academy of Sciences, Acad. Lavrent’ev ave. 5, 630090 Novosibirsk, Russia.
Tel.: +7 383 326 96 89; fax: +7 383 330 96 87.
E-mail addresses: chibirv@nioch.nsc.ru, chibirv@catalysis.ru (A.M. Chibiryaev),
kiv@catalysis.ru (I.V. Kozhevnikov).
0926-860X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2013.03.001

160
A.M. Chibiryaev et al. / Applied Catalysis A: General 456 (2013) 159–167
(Autoclave Engineers), and Hastelloy C-276 alloy is the reactor’s
wall material (A). The autoclave is equipped with mechanical agi-
tator MagneDrive® (Autoclave Engineers) and spinning catalytic
basket Harshaw® (Autoclave Engineers) of volume ca 15 mL, two
thermocouples (T1) and (T2), and sampling system with 6-port
injector (S) Valco C1-2006 (VICI Valco Instruments Co. Inc.). The
starting alcoholic solutions of benzazole were loaded by using a
high-pressure syringe pump Teledyne ISCO 260D (C). Mixer speed
and temperatures (T1) and (T2) are regulated by Sentinel controller
(B), which additionally registers reaction pressure by electronic
pressure sensor (P).
Fig. 1. Experimental set-up.
2.3. General experimental procedure
results in its surface alkoxylation and formation of isolated
Si–OR functionality on the surface [21]. In this case, the inter-
nal lattice structure of silica remains unchanged. But there is
no data on total chemical destruction of silica material by alco-
hols, when SiO2 is “dissolved”. We hypothesize that SiO2 is able
to react with methanol at high temperature without any cata-
lyst/additives to give tetramethyl orthosilicate (TMOS), in spite
of reaction of silica and lower alcohols in the presence of an
alkali metal hydroxide. The last conversion results in the formation
of tetraalkyl orthosilicates too and occurs at temperature above
200 ◦C [22–25]. The further studies to confirm this hypothesis are
required.
In this paper, we conducted a detailed study of the high-
temperature reaction between SiO2 and methanol. In particular,
we studied the reactivity of silica gel, quartz, Pyrex glass, and
H-Y zeolite CBV-720 in supercritical methanol at 350 ◦C and
determined the structures of the resulting products. Additionally,
we have investigated the effect of some N-unsubstituted ben-
zazoles, such as 1H-indole, 1H-benzimidazole, and 1H-indazole,
on reactivity of SiO2 in the high-temperature reaction with
methanol.
Initially, the removable catalytic basket was charged by one of
selected solid SiO2-containing materials: quartz sand – ∼17.0 g, sil-
ica gel – ∼6.7 g, Pyrex glass – ∼17.0 g, zeolite – ∼1.5 g. The amounts
of solids loaded into the catalyst basket were different due to the
different bulk density of these materials. Then the SiO2-containing
material was additionally dried directly in the basket at 180 ◦C for
0.5 h. After that, the weight of filled catalytic basket was measured
with accuracy of 0.5 mg ( 1.0 for silica gel only).
Before heating, the batch reactor (285 mL) with filled catalytic
basket was charged by the required alcohol (methanol, ethanol,
or isopropanol) that was pumped up from the vessel (5) (see Fig.
1). Finally, the reactor was purged with CO2. The initial volume of
methanol loaded was ∼70 mL for runs with benzazoles or ∼117 mL
of the alcohol for all runs without using heterocycles. The heating
time was 50–52 min to reach the reaction temperature from 25 ◦C.
After heating the reactor up to 350 ◦C, benzazole solution (5.8 mmol
in ∼47 mL of methanol) was pumped into the reactor for 10 min.
The starting point of the reaction time was at the end of pumping
the benzazole solution or the reaching of reaction temperature (for
the reaction without benzazole additives).
The reaction occurred at constant temperature of 350 1 ◦C,
mixer speed was 600 rpm in all experiments. During the reac-
tion, samples of the reaction mixture were collected hourly
to the sampler (7) by using a sampling system (S). An alco-
hol (methanol, ethanol, or isopropanol, 2.5–3 mL) was pumped
into the reactor after each sampling to keep a constant reac-
tion pressure. After cooling the reactor to room temperature
(30 min), the final reaction mixture was poured into the vessel
(6) through the port (3) (Fig. 1) and, along with other samples,
analyzed by GC–MS (gas chromatography–mass spectrometry) and
GC–AED (gas chromatography with an atomic emission detector)
methods.
The catalytic basket with SiO2-containing material was washed
out with the same alcohol (50 mL) at 80 ◦C for 0.5 h. After opening
the reactor, the catalytic basket was taken out, dried in the air for
10 h, and finished drying at 180 ◦C for 0.5 h. Then the final weight
of filled catalytic basket was measured. The amount of reacted SiO2
was determined as the difference between the initial weight and
the final one of the filled basket.
2. Experimental
2.1. Reagents and solvents
The following solvents and reagents were used without addi-
tional purification: indole (≥99%, Aldrich), benzimidazole (≥98%,
Aldrich), indazole (≥98%, Aldrich), methanol (≥99.8% with a water
content ≤0.05%, J. T. Baker®), ethanol (≥99.8% with a water content
≤0.16%), and 2-propanol (≥99.7%, FCC, Aldrich).
Wide-porous silica gel (ChromAnalyt, Ltd., Russia) with average
2
pore diameter 141 A, specific surface area 270–280 m g−1, pore
˚
volume 0.35 cm3 g−1 was preliminary calcined at 250 ◦C for 2 h and
then milled. The 0.8–1.0 mm fraction was used in experiments.
Crashed quartz and Pyrex glass (1–1.3 mm fractions) were prelimi-
nary washed out with warm 10% aq. HCl (60 ◦C) for 2 h followed by
multiple washing with distilled water to neutral pH. These solids
were finally calcined at 500 ◦C for 3 h. H-Y Zeolite CBV-720 (mole
˚
ratio of SiO2/Al2O3 is 30; average pore diameter is 24.28 A; specific
surface area is 780 m2 g−1) as a catalyst was commercial (Zeolyst
International) and used without any conditioning.
2.4. Reaction of quartz sand and supercritical alcohol both with
and without the addition of benzazole additives
Critical parameters of used alcohols are the following: for
methanol – Tcr = 240 ◦C, Pcr = 80 atm, ꢀcr = 0.27 g cm−3; for ethanol
– Tcr = 244 ◦C, Pcr = 63 atm, ꢀcr = 0.28 g cm−3; for isopropanol –
The reaction was carried out accordingly to General experimental
procedure. The calculated initial reaction density of supercritical
phase was ꢀ = 0.33 g cm−3. The reaction pressure at supercritical
conditions depended on used alcohol: for methanol – 215 1, for
ethanol – 153 1, for isopropanol – 120 1 atm. Benzazole (1H-
indole, 1H-benzimidazole, or 1H-indazole; 0.68 g) was only used
for the reaction in supercritical methanol (sc-MeOH) and pumped
Tcr = 235 ◦C, Pcr = 47 atm, ꢀcr = 0.27 g cm−3
.
2.2. Experimental set-up
The scheme of experimental set-up is shown in Fig. 1.
The set-up comprises a 285 mL high-temperature batch reactor

A.M. Chibiryaev et al. / Applied Catalysis A: General 456 (2013) 159–167
161
Table 1
up into the reactor as methanol solution immediately after heating
Temperature program and specific parameters for the GC–AED method.
the reactor up to 350 ◦C.
GC experimental conditions
2.5. Reaction of silica gel, crashed Pyrex glass, and H-Y zeolite
with sc-MeOH both with and without the indole additive
Injection port
Split
280 ◦
Injector temperature
Oven temperature program
Column
C
40–280 ◦C (7 ◦C min−1), 280 ◦C (5 min)
Part no. 19091S-916, HP–1MS,
60 m × 0.32 mm i.d. × 0.25 ␮m (100%
dimethylpolysiloxane)
The reaction was carried out accordingly to General experimen-
tal procedure. The calculated initial reaction density of supercritical
phase was ꢀ = 0.33 g cm−3. The reaction pressure at supercriti-
cal conditions was 215 1 atm. Methanol solution of 1H-indole
(0.68 g) was introduced into the reactor immediately after heating
up to 350 ◦C.
Inlet conditions
Column pressure – 13 psi, split flow –
100 mL min−1
Split ratio
Carrier gas
Injection volume
10:1
Helium, make-up flow – 230 mL min−1
2 ␮L
2.6. Gas-phase reaction of methanol and indole both with and
without the quartz sand
AED parameters
Flow of spectrometer purge gas
Transfer line/cavity temperatures
Data rate
500 mL min−1
320 ◦
5 Hz
C
Before heating, the reactor was provided with the removable
catalytic basket that was empty for non-catalytic reaction or filled
with ∼17.0 g of quartz sand for catalytic one. Then the methanol
solution of indole (0.22 g, 1.9 mmol, in ∼37–38 mL) was loaded into
the reactor. The calculated initial reaction density for gas phase
was ꢀ = 0.11 g cm−3. The indole concentration was c = 0.007 mol L−1
in gas phase. The reaction was carried out accordingly to General
experimental procedure but without hourly collection of samples
of the reaction mixture. The reaction pressure was 52–56 atm at
350 ◦C. The reaction time was 4 h. The starting point of the reac-
tion time was at the end of heating the reactor up to the operating
temperature.
After cooling the reactor to room temperature (30 min), the final
reaction mixture was poured into to the vessel (6) through the port
(3) (Fig. 1) and analyzed by GC–MS and GC–AED methods.
Quartz sand in the catalytic basket was washed out with
methanol (50 mL) at 80 ◦C for 0.5 h. After opening the reactor, the
catalytic basket was taken out, dried in the air for 10 h, and finished
drying at 180 ◦C for 0.5 h. Then the final weight of filled catalytic
basket was measured. The amount of quartz sand that reacted was
negligible.
Nitrogen detection
At 388 nm using the CN molecular
emission spectral band
O2/H2/CH4–80:40:50 (psi)
Reagent gases
different forms of SiO2 could react with MeOH at different reaction
rates.
This assumption seems to be reasonable, since SiO2-containing
materials have different density, specific surface area, chemical
impurities, and some other physical properties, which can have
influence on chemical ones. Therefore, SiO2-containing materials
selected for the present study were wide-porous silica gel, quartz
sand, crashed Pyrex glass, and commercial H-Y zeolite CBV-720.
Nevertheless, the main comparative study was performed using
silica gel and quartz sand. All experiments were conducted three
or more times.
3.1. “Dissolution” of quartz and silica gel in sc-MeOH without
additives
2.7. Analysis of reaction mixtures
It was found out the reaction of quartz sand with supercritical
methanol is slow. The conversion of the starting quartz is very small
for 5 h of the reaction - less than 0.05 wt% (line “quartz only”, Fig. 2).
But silica gel is “dissolved” faster at the same reaction conditions
(4 wt%, line “silica gel only”). Thus, 275 mg of silica gel from ∼6.8 g
of its initial amount is “dissolved” for 5 h of the reaction despite
The identification of products was performed by GC–MS method
on a gas chromatograph Varian Saturn 2000 with ion trap as
detector using a 30 m-long quartz capillary column НР–5 with
0.25 mm i.d.; the film thickness of the stationary phase (5%
diphenylsiloxane–95% dimethylsiloxane copolymer) was 0.25 ␮m;
helium with a constant flow rate of 1 mL min−1 was used as the
carrier gas; the injector temperature was 300 ◦C; the temperature
program: 50–200 ◦C (7 ◦C min−1), 200–300 ◦C (20 ◦C min−1), and
300 ◦C (20 min).
The qualitative analysis was carried out by comparing the reten-
tion time of the components and their complete mass spectra with
corresponding data for pure compounds and with data from NIST
and Wiley7 mass spectral libraries. The quantitative analysis of
nitrogen-containing products was made by the GC–AED method
on an Agilent 6890 N gas chromatograph equipped with a jas
2370AA Atomic Emission Detector using a capillary low-bleed col-
umn HP–1MS (19091S-916), 60 m × 0.32 mm i.d. × 0.25 ␮m (100%
dimethylpolysiloxane). Table 1 shows the experimental conditions
for the GC–AED analysis. Calculations of conversion and selectivity
were based on mol nitrogen basis as a very sensitive and accurate
GC method of the quantitative analysis [26].
3. Results and discussion
We have studied the reaction of SiO2 with supercritical
Fig. 2. Reaction of silica gel and quartz with sc-MeOH for 5 h at 350 ◦C. Effect of
benzazole additives on amounts of reacted SiO2.
methanol at 350 ◦C and ∼215 atm. It was a priori assumed that

162
A.M. Chibiryaev et al. / Applied Catalysis A: General 456 (2013) 159–167
Table 2
Main orthosilicates detected by GC–MS analysis in the reaction mixtures.
RT (min)a
MS, m/z M+
Formula
Orthosilicates
3.07
3.92
8.69
152
138
258
Si(OMe)4
Si(OH)(OMe)3
Si2O(OMe)6
Tetramethyl orthosilicateb
Hydroxytrimethoxysilanec
Hexamethyl diorthosilicate
(hexamethoxydisiloxane)d
a
Retention time.
b
c
The relative concentration of TMOS is typically 85–98 wt% of all orthosilicates.
Si(OH)(OMe)3 has usually a very low relative concentration in the reaction mix-
tures.
d
The relative concentration of hexamethoxydisiloxane is 1–8 wt% of all orthosili-
cates.
the fact that the initial total amount of silica gel was 2.5 times less
than the starting weight of quartz sand. (Different starting amounts
of silica gel and quartz sand loaded into the catalyst basket were
limited due to different bulk density of these materials.) In addition,
the crashed Pyrex glass manifested the “solubility”, which is similar
to quartz sand and strongly different from silica gel.
Fig. 3. Orthosilicates content (recalculated to weight of Si per litre) vs time of keep-
ing the reaction mixture. The lower curve – “quartz + methanol”, the upper curve –
“quartz + methanol + indole”.
It is known that an alcohol chemisorbed on silica can react with
its surface alkoxylating at elevated temperature. This esterification
reaction occurs on both isolated silanol groups ( Si OH) and sil-
during the reaction SiO2 + 4CH3OH → Si(OMe)4 + 2H2O, where
methoxysilanols (for example, hydroxytrimethoxysilane) seem to
be intermediates. The condensation of Si(OH)(MeO)3 is known to
result in dimerization and formation of hexamethoxydisiloxane:
oxane bridges ( Si OSi ) and proceeds via cleavage of its Si
O
bonds [27,28]. In our opinion, the significant difference between
the reactivity of silica gel and quartz is due to the enhanced porosity
and “friable” molecular packing of the first compared with quartz.
Consequently, internal and external surfaces of silica gel have more
amounts of free isolated Si OH groups that actively interact with
methanol and facilitate the following “dissolution” of SiO2. An
additional argument is that the reaction involving the opening of
2(MeO)3Si OH → (MeO)3Si
O Si(OMe)3 + H2O [30]. Mono- and
di-orthosilicates formed in these reactions may undergo further
oligomerization. At room temperature, these processes are very
slow and require prolonged period of time to form detectible
amounts of these products.
It should be noted that the quantitative analysis of alkyl ortho-
silicates by the GC–MS method is not a simple routine procedure.
The difficulty is that the amounts of detected orthosilicates in the
reaction mixture for binary system “methanol + water” strongly
depend on the time interval elapsed after the sampling prior to
the GC–MS analysis. Recently, we demonstrated this for reaction
mixtures “TMOS/water/indole” [31]. When the reaction mixture
containing methanol (solvent), TMOS, water, and indole is kept at
room temperature, TMOS starts to undergo hydrolysis and poly-
condensation reactions. A decrease of the amount of the detected
orthosilicates is caused by oligomerization and gelification. The
oligomeric orthosilicates formed from TMOS or Si2O(OMe)6 can-
not be detected by the GC–MS method due to their large molecular
weights and high boiling points (higher than injector temperature
of chromatograph).
Fig. 3 shows the dependence of detected amounts of alkyl
orthosilicates on the time of keeping the reaction mixture at
room temperature. The lower curve refers to the reaction mix-
ture obtained during the conversion of quartz sand in sc-MeOH,
the upper curve – during the conversion of quartz sand in sc-
MeOH with 1H-indole. Molar ratios of orthosilicates/water and
orthosilicates/water/indole compounds were 1:25 and 1:3.6:3.7,
respectively (taken into account the residual water content in
methanol). Estimated concentrations of orthosilicates and water
at starting point of curves were ∼0.2/0.4 and ∼2.0/0.9 g L−1, cor-
respondingly. The starting point of curves was at the end of the
reaction after cooling the reactor to room temperature.
siloxane Si
O Si bridges is favoured at higher temperature [29].
Due to the more fragile skeletal silica network, the contribution of
this esterification pathway to the total “dissolution” process could
be higher for silica gel too.
Our attempts to carry out the reaction of quartz both with super-
critical ethanol and isopropanol have failed at the same reaction
conditions. It was shown that the starting amount of quartz sand
does not change within weighing accuracy ( 0.5 mg) after 5 h of
the reaction in these alcohols. It is quite unexpected result since
the esterification reaction of silica surface by ethanol or other lower
alcohols is well-studied process to produce surface-alkoxylated sil-
ica [27–29]. Presumably, the ability of an alcohol to react with
siloxane bridge disrupting the Si O bond is crucial for the “disso-
lution” of SiO2. Therefore, exclusively methanol demonstrated the
considerable ability “to dissolve” the quartz sand.
3.2. Main reaction products obtained from quartz sand or silica
gel in sc-MeOH
We determined the structure of main products in the reaction
and quantified them. The GC–MS method was chosen for this pur-
pose. NIST-2008 and Wiley 7N mass spectral libraries were used
for the identification. It was shown that the reaction results in the
formation of alkyl orthosilicates, mostly tetramethyl orthosilicate.
Table 2 shows the combined data on the GC–MS analy-
sis of main orthosilicates identified in all reaction mixtures.
Two orthosilicates were the main dissolved Si-containing com-
pounds (about 99 wt% of all orthosilicates), which, in fact, were
formed from SiO2. These were TMOS (RT 3.07 min) and hex-
amethyl diorthosilicate (hexamethoxydisiloxane, RT 8.69 min).
In addition, there was a very weak peak on chromatograms,
which belonged to hydroxytrimethoxysilane (RT 3.92 min). The
formation of hydroxytrimethoxysilane and hexamethyl diorthosil-
icate can be explained by the presence of water formed
The dependence obtained shows that the initial amount of
detected orthosilicates is reduced by three times for the first
day of keeping (Fig. 3). Interesting to note, when molar ratio
of orthosilicates/water/indole compounds was 1:2:1.4 at start-
ing concentrations of orthosilicates and water being equal to
∼9/2 g L−1, the amount of detected orthosilicates was reduced by
half for first two days of keeping [31]. All these data mean that the
initial molar ratio and concentration of orthosilicates, water, and

A.M. Chibiryaev et al. / Applied Catalysis A: General 456 (2013) 159–167
163
Table 3
N
Conversion of benzazoles and ratio of main alkylation-type products in sc-MeOH
with using quartz.a
N
N
H
N
H
N
H
Benzazole
Benzazole
Ratio of C-/N-alkylation
conversion, mol.%
products, mol/mol
1
2
3
Indazole
Benzimidazole
Indole
10
48
54
∼1/4
∼1/2
Fig. 4. Structures of benzazoles used as additives.
∼10/1
a
Detected by the GC–MS method.
indole compounds strongly influence the oligomerization rate of
orthosilicates.
We have found that selected benzazoles react with sc-MeOH
in the presence of quartz at our reaction conditions to form CH3-
substituted benzazoles, mainly – monoalkylated products. Table 3
shows qualitative results of the benzazoles reaction for 5 h. It is seen
that conversions both of indole and benzimidazole are compara-
ble and equal to one-half of their amounts loaded initially into the
reactor. At the same time, the conversion of indazole is five times
less than for other benzazoles. It is much more important to focus
on the ratio of C-/N-alkylation products, which describes in some
way the relationship of C-/N-nucleophilicity of used benzazole.
N-Methylated products are principally formed from both indazole
and benzimidazole. However, the contribution of C-alkylated prod-
ucts increases from indazole to benzimidazole (from 1/4 to 1/2,
Table 3). Dramatically abrupt change of the ratio occurs in the
reaction of indole (up to 10/1), when the C-methylated indoles as
products dominate overwhelmingly. Thus, “the solubility” of quartz
in sc-MeOH was directly proportional to the conversion of benza-
zoles into C-alkylated products, in other words - the reactivity of
quartz sand in sc-MeOH correlates to the C-nucleophilicity of ben-
zazoles. The auxiliary role of benzazoles is a nucleophilic assistance
to the reaction of MeOH with solid SiO2 surface, when SiO2 is con-
verted to TMOS together with methylation reaction of benzazoles
by sc-MeOH.
3.3. Effect of indazole, benzimidazole, and indole on the reaction
of quartz and silica gel with sc-MeOH
To convert SiO2 (acidic oxide) into TMOS (ester of orthosili-
cic acid) at high temperature, it is usually needed an alkali
hydroxides [22–25]. We assumed that organic bases, for example,
N-unsubstituted benzazoles (1H-benzimidazole 1, 1H-indazole 2,
and 1H-indole 3, Fig. 4), will also promote the “dissolution” of SiO2-
containing materials in sc-MeOH at high temperature. The basicity
of azoles used is as follows: benzimidazole > indazole ꢀ indole [32].
If the idea is correct, the rapidity of SiO2 “dissolution” will correlate
to the basicity of the azoles.
It was found out that the amount of “dissolved” quartz increases
by 1.4 times with using the indazole (from 7.0 to 9.5 mg, line
“quartz + indazole”, Fig. 2). More significant effect is achieved when
benzimidazole was used – the quartz “solubility” increases by ∼5
times (line “quartz + benzimidazole”). It seems obvious that these
two results of the reaction correlate with the basicity of selected
benzazoles. However, the result of the reaction “quartz + indole” is
quite unexpected. Contrary to our expectations, indole additives
have a dramatic effect on the reactivity of quartz increasing its
“solubility” in sc-MeOH by 14 times (line “quartz + indole”). Nev-
ertheless, this result (0.6 wt% of starting amount of quartz sand)
is not the best if we compare it with the “solubility” of silica gel
without additives of benzazoles (4 wt%, line “silica gel only”).
Interestingly, when indole was introduced in the reaction with
sc-MeOH, the “solubility” of silica gel increases by 1.4 times only
(line “silica gel + indole”, Fig. 2) compared with the reaction of sil-
ica gel without additives. The possible reason is the following. Our
preliminary data show that the amount of silica gel, which was “dis-
solved” in equal volume of sc-MeOH with considerable increasing
the reaction time (up to 24 h), has the upper limit which is prefer-
ably determined by the water content in the reaction mixture. It
is reasonable to assume that the reaction of SiO2 with methanol
resulting in formation of water together with TMOS is reversible.
Therefore, the SiO2 “dissolution” will be stopped when a certain
concentration of water is reached in the reaction mixture. We
observed this effect in several preliminary runs with silica gel, but
this phenomenon requires further detailed study.
3.5. Influence of SiO2 ability “to be dissolved” on indole
conversion in supercritical and gas-phase MeOH
Since the glassware and other SiO2-containing materials can
slowly react with methanol at high temperature to give orthosili-
cate (TMOS), we believe that these materials should be used with
caution for several chemical processes. For example, it is necessary
to consider when some reactions carry out at similar conditions
and both SiO2 and methanol are starting reaction components. The
problem is that tetramethyl orthosilicate, which is formed in the
reaction of SiO2 with methanol, could catalyze undesirable side
reactions. Presently, the catalytic activity of TMOS in organic reac-
tions is poorly studied.
Recently, the effective indole C3-methylation by sc-MeOH
was described by Kishida et al. [33]. The reaction carried
out in Pyrex glass sealed tubes was determined by the
authors as “without catalyst” conversion. Our attempts to
repeat Kishida’s results in batch reactor, in which corrosion-
resistant Hastelloy® C-276 alloy was the reactor’s wall
material, were failed. The indole conversion and selectivity
of C3-methylation achieved by us were about three times lower.
Then, it has been suggested that TMOS, which is formed at high
temperature from Pyrex glass in sc-MeOH, has a catalytic effect
on the indole C3-methylation. More recently, we have proved
the idea by demonstrating the use of TMOS or TEOS (tetraethyl
orthosilicate) significantly increases the selectivity of indole
C3-methylation by sc-MeOH (up to 95–96 mol%) [31]. The catalytic
role of tetraalkyl orthosilicates was unequivocally established in
this reaction.
3.4. Correlation of quartz conversion in sc-MeOH with
C-nucleophilicity of indole, benzimidazole, and indazole
Bearing in mind that the basicity of benzazoles did not have a
crucial importance for the “dissolution” of SiO2-containing mate-
rials, we tried to determine, which property of benzazoles has
the key influence on the reactivity of SiO2 in sc-MeOH. We
believe that the property is the nucleophilicity of benzazoles,
and not N-, but mainly C-nucleophilicity. N-Nucleophilicity of
selected benzazoles correlates accurately with their basicity (see
above), but C-nucleophilicity of these benzazoles is as follows:
indole ꢀ benzimidazole > indazole. It means that the reactivity of
indole is many times higher than that of benzimidazole (and much
more – of indazole) in aromatic electrophilic substitution reactions,
when these benzazoles act as a C-nucleophile.
To demonstrate the dependence of indole methylation reaction
on the use of different SiO2-containing materials, we performed

164
A.M. Chibiryaev et al. / Applied Catalysis A: General 456 (2013) 159–167
Table 4
Comparative indole conversion and regioselectivity of C3-methylation catalyzed by different forms of SiO2.
SiO2-containing catalyst
Indole conversion,
mol.%d
3-Methylindole
yield, mol.%d
3-Methylindole
selectivity, %
Without catalyst in sc-MeOHa
Without catalyst in gas phaseb
Quartz sand in sc-MeOHa
29
17
47
19
34
40
23
45
77
18
11
41
17
33
35
19
22
30
62
65
87
88
96
87
83
50
38
Quartz sand in gas phase b
After removal of quartz sand in sc-MeOHa
Crashed Pyrex glass in sc-MeOHa
Silica gel in sc-MeOHa
HY zeolite CBV-720 in sc-MeOHa
CeHY zeolite in gas phase (Ref. [34])c
a
In supercritical methanol, molar ratio of methanol/indole is 500, 350 ◦C, 213–215 atm, 4 h.
In gas-phase methanol, molar ratio of methanol/indole is 500, 350 ◦C, 52–56 atm, 4 h.
b
c
In gas-phase methanol, molar ratio of methanol/indole is 6, HY zeolite impregnated by 3 wt% Ce+ salt, 350 ◦C, 4 h.
Detected by the GC–MS method.
d
CH3
is higher than that silica gel is used as catalyst (line 7). The
C3-methylation selectivity achieves 62 mol% (4 h) at non-catalytic
conditions. It is seen that the graphical dependence is reaching a
plateau after 5 h of the reaction.
sc MeOH
3
+
1
SiO2
350 oC, 213 atm
N
H
N
H
N
CH3
Poor catalytic efficiency is demonstrated by H-Y zeolite (this
result is not shown on Figs. 5–6, see line 8 in Table 4), those use
leads to ∼45% of indole conversion (4 h). Selectivity on 3-methyl-
1H-indole 4 (3-MI, skatole) is much lower (about 50 mol%, 4 h) due
to increasing the amount of 1-methyl-1H-indole 5 (∼7–9 mol%) as
well as di-, tri- and poly-methylated indoles (cf. Ref. [33]).
Apparently, the acidity and other properties of zeolite CBV-720
proved to be too strong to provide acceptable selectivity of alkyl-
ation by methanol. It appears that HY zeolite is similar to usual
Friedel–Crafts catalyst (a strong Lewis acid) at our reaction con-
ditions. The catalytic activity of TMOS formed from this zeolite
during the reaction is insufficient to that the product of indole C3-
methylation would be the main one. Earlier, it has been shown
under gas-phase conditions that the use of modified HY zeolites
in the reaction of indole with methanol can lead to high selectivity
of the C3-alkylation (96% for 4 h at 300 ◦C), but only at low conver-
sion of starting indole (≤35%) [34]. The highest catalytic efficiency
at 350 ◦C has been demonstrated by HY zeolite modified with Ce+
ions (line 9 in Table 4). However, the yield of 3-MI 4 was only 30% at
high conversion of indole (77%). This confirms our conclusion that
zeolites are not quite effective catalysts for the indole alkylation by
alcohols at high temperature (above 300 ◦C).
3
4
5
Scheme 1. The main products of indole monomethylation by sc-MeOH.
Fig. 5. Indole conversion in sc-MeOH with using different SiO2-containing materials
(mol%).
the reaction of sc-MeOH at 350 ◦C with using the crushed Pyrex
glass, quartz sand, silica gel, and commercial H-Y zeolite CBV-
720. Scheme 1 shows main products of the indole alkylation by
sc-MeOH.
Figs. 5–6 and Table 4 show the indole conversion that was only
29% in non-catalytic reaction (line 1 in Table 4). This conversion
Catalysis by wide-porous silica gel gives 83 mol% of the selec-
tivity on C3-alkylation (4 h), although the indole conversion is only
23% for the same reaction time (line 7 in Table 4). Dependences on
Figs. 5–6 demonstrate that the use of SiO2 slows down the indole
conversion during first 5 h compared with non-catalytic reaction,
but then the conversion increases more rapidly. In addition, the
selectivity of the reaction with SiO2 is higher by 20% (after 4 h)
than that without catalyst.
Interestingly, graphical dependences of the indole conversion
and the yield of skatole 4 vs the reaction time, which were obtained
for catalysis by silica gel, are nearly straight within chosen range of
reaction times.
The crashed Pyrex glass as a catalyst shows linear dependences
too. The indole conversion grows up to 40% (4 h), and the selectivity
on 3-MI 4 (87 mol%, line 6 in Table 4) is noticeably higher compared
to other reactions listed above (Figs. 5–6). Such selectivity of the
indole C3-alkylation was comparable with Kishida’s one achieved
for the reaction in Pyrex glass sealed tube [33].
Quartz sand demonstrates the best catalytic result in the alkyl-
ation reaction (Figs. 5–7). The conversion of indole reaches 47%
at C3-selectivity of 87 mol% for 4 h of reaction (line 3 in Table 4).
Increasing of the reaction time up to 19 h (Fig. 7) leads to the com-
plete indole conversion (99% and higher) accompanied by slight
Fig. 6. Total yield of 3-methylindole 4 in sc-MeOH (mol%).

A.M. Chibiryaev et al. / Applied Catalysis A: General 456 (2013) 159–167
165
observed is the main evidence of SiO2 catalytic activity in gas
phase. In addition, the emergence of TMOS in the reaction mixture
(GC–MS data) shows that SiO2 is also reacting with methanol in
the gas-phase conditions at high temperature.
The comparison of experimental data obtained at gas-phase and
supercritical conditions suggests that the reaction of SiO2 with
methanol at 350 ◦C is faster at high pressures, i.e. the higher the
density of the reaction medium (and partial pressure of methanol)
is, the quicker and more actively the desilication of zeolites, silicas
or glasses occurs. The effect of the reaction temperature assumes to
be the same: the higher the temperature is, the faster the reaction
of SiO2 with methanol is. Interestingly, such desilication may be yet
another reason for the catalyst deactivation in the MTO process or
the one-step DME synthesis. It is well known that catalysts or cat-
alyst systems in aforementioned processes are usually deactivated
due to sintering of active copper sites at high reaction tempera-
ture, forming of coke on the surface of the catalyst, and poisoning
by some contaminants [35–38]. At the same time, if the desilica-
tion of SiO2-containing catalysts takes place under the action of
methanol at high temperature and pressure, the partial destruc-
tion of active sites of catalysts should occur by crushing the catalyst
matrix.
It was interesting to establish the origin of catalytic effect on the
indole alkylation. Is it the surface area of SiO2-containing mate-
rial (heterogeneous catalysis) or the action of TMOS–methanol
binary system (homogeneous catalysis) as we demonstrated earlier
[31]?
At first, the reaction was carried out between quartz sand and
sc-MeOH at 350 ◦C for 2 h, after that the catalytic basket with
quartz sand was removed from the reactor. When the reactor was
bolted and heated up to 350 ◦C, the concentrated indole solution
in methanol was loaded by using a high-pressure syringe pump.
Finally, the reaction was carried out for 6 h under supercritical con-
ditions in the presence of compounds formed preliminary after
contact of quartz sand with sc-MeOH (TMOS and some other methyl
orthosilicates). A comparison of the reactions in presence of quartz
and after its removal from the reactor is shown in Fig. 7. One can see
that the removal of quartz from the reactor resulted in decreasing
the final indole conversion for 4 h of the reaction (from 47 down
Fig. 7. Comparison of catalytic methylation by using sc-MeOH with quartz in reactor
volume and after preliminary removal of quartz from reactor.
decrease of the selectivity from 88 (2 h) to 82 mol% (19 h). The
higher yield of 3-MI 4 is achieved for 15–17 h of the reaction (about
81 mol%), and then concentration of the C3-methylated indole 4 in
the reaction mixture begins to decrease due to its further alkylation.
Thus, we have demonstrated for the reaction of indole alkylation
that the catalytic efficiency of SiO2 depends strongly on the type of
SiO2-containing material.
It seemed to be important to determine whether quartz sand has
catalytic effect in gas-phase reaction of indole with methanol. We
scaled down for three times initial concentrations of indole and
methanol that were earlier used for the reaction in supercritical
MeOH. It was necessary to decrease the reaction density inside the
reactor lower than critical value. At the same time, the amount of
quartz sand as a catalyst was no change.
Gas-phase reaction that occurred at the same 350 ◦C showed
that the products distribution does not differ from one obtained in
sc-MeOH. However, both non-catalytic and catalytic conversions
of indole are significantly decreased in gas-phase reaction. The
reaction without using the quartz sand gave 17% of conversion
for 4 h with 65 mol% of selectivity to 3-methylindole 4 (line 2 in
Table 4). The selectivity was reaching up to 88 mol% when quartz
sand was used (at 19% of indole conversion, line 4). The selectivity
O
O
O
Methano l
release
Si
H
O
H3C
O
H
O
H
O
O
Si
CH3
Free silanol
group
Fr ee silanol
group
Free me thoxide
functionality
N
OH
A
H
OCH3
OH
Si
CH3
CH3
Si
CH3OH
- H2O
Me OH as alkylating
agent
O
O
O
O
N
N
H
O
O
3-methyl-3 -indole
H
B
4
Initial sur face of
SiO2-containing
material
Surface me thylated
by MeO H
H
OH
OH
Si
H3CO
H
H3C
Si
H3CO
N
O
O
O
O
H
O
Opening of
siloxane bridge
O
Si
O
O
O
Mo dified (methyl-
hydroxylated) sur face
Methylated surface as
alkylating agent
Scheme 2. The proposed nucleophilic assistance of indole for participation of SiO2-surface in catalysis.

166
A.M. Chibiryaev et al. / Applied Catalysis A: General 456 (2013) 159–167
Free silanol
Isolated
group
OH
Si
methoxide group
OCH3
OH
Si
OH
Si
H3CO
Si
Methanol
release
Water
release
O
O
O
O
O
O
O
O
O
O
O
Methylated surface
Recovered surface of
SiO2-containing material
Modified (methyl-
hydroxylated) su rface
CH3OH
Si(OCH3)4 + H2O
("dissolution" of SiO2)
Scheme 3. The further plausible transformations of methyl-hydroxylated SiO2-surface.
to 34 mol%, line 5 in Table 4). Nevertheless, the selectivity of C3-
methylation became extremely high (96%) and the same as if the
reaction would be catalyzed homogeneously by TMOS [31].
as an acidic oxide to give mainly tetramethoxysilane–tetramethyl
ester of orthosilicic acid. Surprisingly, the additive of aromatic aza-
heterocycles, such as N-unsubstituted indole, benzimidazole, and
indazole, is able to increase considerably the amount of reacted
SiO2-containing material. The key role of tested benzazoles is the
nucleophilic assistance for the reaction of SiO2 and methanol. At
the same time, tetramethyl orthosilicate that is formed from SiO2
under these reaction conditions has appreciable catalytic activity
on methylation reaction of benzazole, especially on indole alkyl-
ation, and promotes (facilitates) the conversion of benzazole to C-
and/or N-methylated derivatives. We hypothesize that the methyl-
ation process of benzazole is the aromatic electrophilic substitution
by methanol or methoxylated solid SiO2-surface as a methylat-
ing agent and proceeds via cyclic, multi-centred transition state.
Importantly, both ethanol and isopropanol do not undergo this
reaction with SiO2 at the same experimental conditions.
Thus, it could be concluded that the high selectivity of the reac-
tion is the result of homogeneous catalysis by TMOS. The last is
formed from quartz (in general – from any SiO2-containing mate-
rial) at high temperature during the “dissolution” in sc-MeOH. At
the same time, the solid quartz surface is also involved in the
overall catalytic conversion. This participation could be as fol-
lows (Scheme 2). At first stage, methanol reacts with isolated
Si OH groups of SiO2-containing material to methylate partially
its surface. At second step, the chemically modified surface par-
ticipates for indole alkylation as a part of catalytic binary system
(
SiOCH3 + MeOH, cf. Ref. [31]).
At least, two pathways of the alkylation could be realized
(Scheme 2). The pathway (A): an alkylating agent is МeOН; 1H-
indole is coordinated to the solid surface by hydrogen bonding of
indole N H with oxygen atom of SiOCH3 group of modified sur-
face (as shown in Scheme 2) or with oxygen atom of silanol Si OH
group of initial unmodified SiO2-surface. The role of methoxy-
lated SiO2-surface is to adopt H atom of indole H N bond and to
facilitate the HO CH3 bond rupture due to Lewis acid properties.
The pathway (B): an alkylating agent is methoxylated SiO2-surface
(isolated Si OCH3 group); 1H-indole is coordinated to the solid
surface by hydrogen bonding of indole N H with oxygen atom
These data suggest that all SiO2-containing materials (includ-
ing silica-supported catalysts too) are chemically destroyed by
methanol at 350 ◦C and above, especially in the presence of differ-
ent nucleophilic additives. Additionally, methyl orthosilicates can
catalyze some reactions (for example, undesirable ones) between
different chemical components of the reaction mixture.
Acknowledgment
of siloxane Si
O Si bridge (as shown in Scheme 2). The role
This work was supported by the Russian Foundation for Basic
of MeOH molecule is nucleophilic assistance for indole methyl-
ation by SiO CH3 moiety. The result of such auxiliary is the
facilitation of SiO CH3 and SiO Si bonds cleavage and the
additional methoxylation of SiO2-surface. The consolidated result
on the participation of both indole and methanol molecules is a
“loosening” of initial SiO2-surface structure and formation of new
well-developed “methyl-hydroxylated” one. The last is more friable
and much less chemically robust than the unmodified SiO2-surface
and could reversible reacts with MeOH (easier and faster) to give
TMOS or can eliminate (release) H2O and MeOH molecules to form
a poor-developed SiO2-surface which can also be converted into
Si(OCH3)4 under the reaction conditions, but slower than “methyl-
hydroxylated” surface (Scheme 3).
Research (No. 11-03-12049-OФИ-м, 2011).
References
[1] H. Zhang, J. Wu, S. Qin, C. Hu, Ind. Eng. Chem. Res. 45 (2006) 7090–7095.
[2] K. Aoki, M. Ohmae, T. Nanba, K. Takeishi, N. Azuma, A. Ueno, H. Ohfune, H.
Hayashi, Y. Udagawa, Catal. Today 45 (1998) 29–33.
[3] S.P. Naik, H. Du, H. Wan, V. Bui, J.D. Miller, W.W. Zmierczak, Ind. Eng. Chem.
Res. 47 (2008) 9791–9794.
[4] H. Yin, Y. Ding, H. Luo, L. Yan, T. Wang, L. Lin, Energy Fuels 17 (2003) 1401–1406.
[5] W. Feng, Q. Wang, B. Jiang, P. Ji, Ind. Eng. Chem. Res. 50 (2011) 11067–11072.
[6] A.T. Aguayo, D. Mier, A.G. Gayubo, M. Gamero, J. Bilbao, Ind. Eng. Chem. Res. 49
(2010) 12371–12378.
[7] S.M. Abubakar, D.M. Marcus, J.C. Lee, J.O. Ehresmann, C.-Y. Chen, P.W. Kletnieks,
D.R. Guenther, M.J. Hayman, M. Pavlova, J.B. Nicholas, J.F. Haw, Langmuir 22
(2006) 4846–4852.
[8] A. Sassi, M.A. Wildman, J.F. Haw, J. Phys. Chem. B 106 (2002) 8768–8773.
[9] Zh. Zhu, M. Hartmann, L. Kevan, Chem. Mater. 12 (2000) 2781–2787.
[10] S. Ivanova, B. Louis, B. Madani, J.P. Tessonnier, M.J. Ledoux, C. Pham-Huu, J. Phys.
Chem. C 111 (2007) 4368–4374.
[11] M. Castro, S.J. Warrender, P.A. Wright, D.C. Apperley, Y. Belmabkhout, G.
Pirngruber, H.-K. Min, M.B. Park, S.B. Hong, J. Phys. Chem. C 113 (2009)
15731–15741.
[12] B.P.C. Hereijgers, F. Bleken, M.H. Nilsen, S. Svelle, K.-P. Lillerud, M. Bjørgen, B.M.
Weckhuysen, U. Olsbye, J. Catal. 254 (2009) 77–87.
[13] A. Ciftci, N.A. Sezgi, T. Dogu, Ind. Eng. Chem. Res. 49 (2010) 6753–6762.
4. Conclusions
We have demonstrated that different SiO2-containing materi-
als react with methanol at 350 ◦C under both supercritical and
gas-phase conditions. The amount of the material reacted with
supercritical methanol can be varied from 0.05 wt% for quartz sand
to 4.0 wt% for silica gel for 5 h of the reaction. SiO2 reacts in this case

A.M. Chibiryaev et al. / Applied Catalysis A: General 456 (2013) 159–167
167
[14] T. Oku, T. Ikariya, Angew. Chem. Int. Ed. 41 (2002) 3476–3479.
[15] T. Oku, Y. Arita, H. Tsuneki, T. Ikariya, J. Am. Chem. Soc. 126 (2004) 7368–7377.
[16] T. Oku, Y. Arita, T. Ikariya, Adv. Synth. Catal. 347 (2005) 1553–1557.
[17] R. Burch, G.D. Squire, S.C. Tsang, J. Chem. Soc., Faraday Trans. 1 (85) (1989)
3561–3568.
[27] M.A. Natal-Santiago, J.A. Dumesic, J. Catal. 175 (1998) 252–268.
[28] G.C. Ossenkamp, T. Kemmitt, J.H. Johnston, Chem. Mater. 13 (2001) 3975–3980.
[29] T. Luts, A. Katz, Top. Catal. 55 (2012) 84–92.
[30] P.J. Skrdla, M. Shnayderman, L. Wright, T.P. O’Brien, J. Non-Cryst. Solids 352
(2006) 3302–3309.
[18] P.S. Yarlagadda, L.A. Morton, N.R. Hunter, H.D. Gesser, Ind. Eng. Chem. Res. 27
(1988) 252–256.
[31] I.V. Kozhevnikov, A.L. Nuzhdin, G.A. Bukhtiyarova, O.N. Martyanov, A.M.
Chibiryaev, J. Supercrit. Fluids 69 (2012) 82–90.
[19] M. Seman, J.N. Kondo, K. Domen, R. Radhakrishnan, S.T. Oyama, J. Phys. Chem.
B 106 (2002) 12965–12977.
[20] M. Seman, J.N. Kondo, K. Domen, C. Reed, S.T. Oyama, J. Phys. Chem. B 108 (2004)
3231–3239.
[32] L.-Z. Chen, R. Flammang, A. Maquestiau, R.W. Taft, J. Catalán, P. Cabildo, R.M.
Claramunt, J. Elguero, J. Org. Chem. 56 (1991) 179–183.
[33] N. Kishida, T. Kamitanaka, M. Fusayasu, T. Sunamura, T. Matsuda, T. Osawa, T.
Harada, Tetrahedron 66 (2010) 5059–5064.
[21] H. Utsugi, H. Horikoshi, T. Matsuzawa, J. Colloid Interface Sci. 50 (1975)
154–161.
[34] D.V. Gopal, B. Srinivas, V. Durgakumari, M. Subrahmanyam, Appl. Catal. A 224
(2002) 121–128.
[22] E. Suzuki, M. Akiyama, Y. Ono, J. Chem. Soc., Chem. Commun. 2 (1992) 136–137.
[23] Y. Ono, M. Akiyama, E. Suzuki, Chem. Mater. 5 (1993) 442–447.
[24] T. Kemmitt, W. Henderson, Aust. J. Chem. 51 (1998) 1031–1036.
[25] L.N. Lewis, F.J. Schattenmann, T.M. Jordan, J.C. Carnahan, W.P. Flanagan, R.J.
Wroczynski, J.P. Lemmon, J.M. Anostario, M.A. Othon, Inorg. Chem. 41 (2002)
2608–2615.
[35] W. Wu, W. Guo, W. Xiao, M. Luo, Fuel Process. Technol. 108 (2013) 19–24.
[36] F.S.R. Barbosa, V.S.O. Ruiz, J.L.F. Monteiro, R.R. de Avillez, L.E.P. Borges, L.G.
Appel, Catal. Lett. 126 (2008) 173–178.
[37] A.T. Aguayo, J. Eren˜a, I. Sierra, M. Olazar, J. Bilbao, Catal. Today 106 (2005)
265–270.
[38] J. Eren˜a, R. Garon˜a, J.M. Arandes, A.T. Aguayo, J. Bilbao, Catal. Today 107–108
(2005) 467–473.
[26] A.M. González, P.C. Uden, J. Chromatogr. A 898 (2000) 201–210.