Deoxygenation of m-cresol on Pt/γ-Al2O3 catalysts

Article abstract of DOI:10.1016/j.cattod.2013.04.011

In this work the deoxygenation of m-cresol, used as a bio-oil model compound, was studied at atmospheric pressure, using Pt/γ-Al ₂O₃ catalysts. The main products obtained in this system were toluene, benzene and methylcyclohexane. Possible reaction routes for cresol deoxygenation were proposed. The influence of the metal loading, H ₂/cresol ratio and temperature on products distribution was analyzed. Catalyst deactivation was studied by TPO. The catalysts were characterized by XPS, XRD, BET, TPR, and TEM. Product distribution is explained based on the relative rates of cresol dehydroxylation and demethylation, which leads to toluene and phenol formation, respectively. High selectivity to toluene was obtained with the catalyst with the higher metal loading (1.7 wt.%), while in the case of low metal loading (0.05 wt.%) high yields of light hydrocarbons were obtained. Coke deposition took place preferentially on the metal particles and could be removed by treatment at low temperatures, e.g. 350 C. This catalyst displays a very good activity and selectivity towards toluene formation at mild temperature (300 C) and atmospheric pressure, which is a better performance than that already reported for other catalysts.

Full text of DOI:10.1016/j.cattod.2013.04.011

Catalysis Today 213 (2013) 9–17
Contents lists available at SciVerse ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Deoxygenation of m-cresol on Pt/␥-Al O catalysts
2
3
∗
M.S. Zanuttini, C.D. Lago, C.A. Querini , M.A. Peralta
Instituto de Investigaciones en Catálisis y Petroquímica (INCAPE) (FIQ, UNL, CONICET), Santiago del Estero 2654, 3000 Santa Fe, Argentina
a r t i c l e i n f o
a b s t r a c t
Article history:
In this work the deoxygenation of m-cresol, used as a bio-oil model compound, was studied at atmo-
spheric pressure, using Pt/␥-Al2O3 catalysts. The main products obtained in this system were toluene,
benzene and methylcyclohexane. Possible reaction routes for cresol deoxygenation were proposed. The
influence of the metal loading, H2/cresol ratio and temperature on products distribution was analyzed.
Catalyst deactivation was studied by TPO. The catalysts were characterized by XPS, XRD, BET, TPR,
and TEM. Product distribution is explained based on the relative rates of cresol dehydroxylation and
demethylation, which leads to toluene and phenol formation, respectively. High selectivity to toluene
was obtained with the catalyst with the higher metal loading (1.7 wt.%), while in the case of low metal
loading (0.05 wt.%) high yields of light hydrocarbons were obtained. Coke deposition took place pref-
Received 17 December 2012
Received in revised form 10 April 2013
Accepted 12 April 2013
Available online 15 May 2013
Keywords:
Deoxygenation
Bio-oil
Cresol
Platinum
Alumina
◦
erentially on the metal particles and could be removed by treatment at low temperatures, e.g. 350 C.
This catalyst displays a very good activity and selectivity towards toluene formation at mild temperature
◦
(
300 C) and atmospheric pressure, which is a better performance than that already reported for other
catalysts.
©
2013 Elsevier B.V. All rights reserved.
1
. Introduction
The lignocellulosic biomass fast pyrolysis is one of the possi-
were used to deoxygenate phenolic compounds at high hydrogen
pressures. Table 1 presents the catalysts used at high H2 pressure,
and relevant information recently reported in this field. Phenolic
compounds represent an important fraction of the bio-oil. There
are few reports in which good results are presented, related to the
deoxygenation of these compounds with low hydrogen consump-
tion (atmospheric pressure) and high selectivities to high octane
deoxygenated aromatics. For example, phenol was deoxygenated
with high selectivity to benzene, at atmospheric pressure with the
Ni/SiO2 catalyst [26]. On the other hand, results obtained in the
deoxygenation of phenol, 2-methoxy-4-(2-propanyl) phenol and
other phenolic compounds using HZSM-5 catalysts were not satis-
ble alternatives to generate biofuels. The liquid product obtained
in this process is called bio-oil. Although it is a potential fuel,
it cannot be used as such without prior refining due to its high
viscosity, low heating value, corrosiveness and instability [1,2].
Moreover, the high reactivity complicates the use of bio-oil as a
fuel and its long term storage. The refining essentially involves
the removal of oxygen, since it imparts to this biofuel the unde-
sirable properties mentioned above [3]. Bio-oil deoxygenation has
been studied using hydrotreating catalysts at high pressures of
hydrogen or a mixture of hydrogen and carbon monoxide, and/or
in the presence of hydrogen donor solvents [4–9]. Acid catalysts
such as zeolites or silica-alumina, working at low pressures were
also used [10–15]. However, when studying deoxygenation reac-
tions using model compounds with different types of catalysts it
was generally observed that the phenolic fraction was not deoxy-
genated or it required a high hydrogen consumption and high
pressure. Besides, when using high hydrogen pressures, the forma-
tion of fully hydrogenated products of lower octane number and
therefore, less important in order to feed the fuel pool, was gen-
erally observed. Several catalysts, mostly hydrotreating catalysts,
factory [27–30]. The Fe/SiO catalyst [31] was used to deoxygenate
2
guaiacol but the selectivity to benzene–toluene–xylenes fraction
(BTX) was low. Guaiacol was also deoxygenated with Pt–Sn [32]
and V O5/Al O catalysts [33], but phenol was the main product in
2
2
3
both cases. Recently, Resasco et al. [34] reported the used of gallium
modified beta-zeolite for hydrodeoxygenation of m-cresol, work-
ing at low pressures (1 atm), but at rather high temperatures, in the
◦
range of 400–550 C. Total conversion of m-cresol was obtained at
W/F = 24, with a toluene yield of 40%.
In sum, previous reported research related to bio-oil deoxygen-
ation was not completely successful mainly due to two reasons.
On one hand, when working under conditions of low hydrogen
pressure with acid catalysts, a high bio-oil fraction was not deoxy-
genated. This fraction was formed mainly by phenolic compounds.
On the other hand, in order to deoxygenate the phenolic frac-
tion, high hydrogen consumption was required and completely
∗ Corresponding author at: Santiago del Estero 2654, (3000) Santa Fe, Argentina.
Tel.: +54 342 4533858; fax: +54 342 4531068.
E-mail address: querini@fiq.unl.edu.ar (C.A. Querini).
0
920-5861/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2013.04.011

1
0
M.S. Zanuttini et al. / Catalysis Today 213 (2013) 9–17
Table 1
Deoxygenation of phenolic compounds at high H2 pressure.
Catalyst
Reagent
Main products
P (MPa)
Ref.
MoS2, CoMoS
Co–Mo–B
NiMoP/Al2O3
CoO–MoO3/Al2O3 sulfided
MoO3–NiO/Al2O3
Rh/SiO2–Al2O3, Ru/SiO2–Al2O3
NiMo sulfide
MoS2, CoMoS2
Ni,Mo/Al2O3, Mo/Al2O3, Co,Mo/Al2O3,
CoMoS
Guaiacol
Phenol
2-ethylphenol
Cresol
Phenol
Guaiacol
Phenol
Phenol
2-ethylphenol
Guaiacol
Cyclohexane, cyclohexene and benzene
Cyclohexane
Ethylcyclohexane
Toluene and methylcyclohexane
Benzene, cyclohexane and methylcyclopentane
Toluene and methylcyclohexane
Benzene, cyclohexane, cyclohexene
Benzene, cyclohexane, cyclohexene
Ethylcyclohexane, ethylbenzene
4
4
7
6.9
2.8
4
2.8
2.8
7
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
Benzene, toluene, methylcyclohexane, methylcyclohexene
4
hydrogenated products were usually obtained. Lobo et al. [48,49]
studied the deoxygenation of m-cresol using platinum supported
on alumina, at 260 C and a W/F = 0.58 gcath/gcresol, and proposed
a reaction sequence in which the m-cresol is hydrogenated on
the metal, dehydrated on the acid sites (alumina) and finally
dehydrogenated on the metal. The catalyst used by Lobo et al.
Particle size distributions were determined by Transmission
Electron Microscopy (TEM). The reduced samples were placed in
distilled water. One drop of this suspension was then placed on
holey carbon supported on a copper grid. The micrograph images of
the samples were acquired with a JEOL 100 CX model microscope at
100 kV, and a magnification of 450,000 ×. The metallic dispersions
were calculated by the following equations [35,36]:
◦
[
48] was bifunctional, with the metal being active in hydrogena-
tion–dehydrogenation and hydrogenolysis reactions, and the acid
mainly in dehydration and hydrocraking reactions. B. Subramaniam
0
.821
DPt
dPt =
(1)
et al. [50] compared Pt/Al O3 with other supported metal catalysts
2
in m-cresol conversion in liquid phase, and found that Pt was more
active than the other catalysts.
where DPt is the metallic dispersion and dPt is the particle size (Dva)
in nm. Dva is a volume-area average size defined as follows:
In this work, the deoxygenation of cresol (methyl·phenol) was
ꢀ
nidi3
ꢀ
studied in conditions of low hydrogen pressure using Pt/␥-Al O
2
3
Dva =
(2)
nidi2
catalysts. These catalysts were chosen after a screening of different
types of catalysts comprising acid zeolites, different oxides, tran-
sition metals and noble metals. m-Cresol was chosen as a bio-oil
model compound corresponding to the phenolic fraction. The effect
of operative variables on the catalytic activity, selectivity to differ-
ent deoxygenated products and catalyst stability was analyzed, and
possible reaction routes were proposed.
The amount of carbonaceous materials deposited on the spent
catalysts were determined by Temperature-Programmed Oxida-
tion (TPO), using a stream of 5% (v/v) O in N2 and a heating rate
of 12 C.min . The oxidation products were detected with a flame
ionization detector (FID) after methanation. Further details can be
found elsewhere [37].
2
◦
−1
The Pt dispersion was also determined by CO pulse chemisorp-
tion. Dynamic CO pulse chemisorption measurements were carried
out by sending 250 ␮L pulses of 1% CO/He on reduced samples.
CO was detected with a FID, after quantitative methanation in a
Ni/kieselghur catalyst at 400 C, which largely improves the sensi-
tivity.
The chemical state of Pt was determined by X-ray Photoelectron
Spectroscopy (XPS). The XPS analyses were performed in a multi-
technique system (SPECS) equipped with a dual Mg/Al X-ray source
and a hemispherical PHOIBOS 150 analyzer operating in the fixed
analyzer transmission (FAT) mode. The samples were mounted on
a sample rod, placed in the pretreatment chamber of the spec-
2
. Experimental
2.1. Catalysts preparation
◦
The catalysts were prepared by wet impregnation of plat-
inum precursor on ␥-Al O support. Tetraammonium platinum (II)
nitrate (metal content 50%) was supplied by Alfa Aesar. An aqueous
solution of 1% of Pt(NH ) (NO ) was used to prepare the catalyst. A
suspension of ␥-Al O3 in the metal precursor solution was stirred
on a hot plate at 110 C until complete evaporation. The impreg-
nated catalyst was dried in an oven at 110 C for 12 h. The dried
sample was calcined in an electric furnace at 350 C for 2 h. Platinum
content was varied between 0.05 and 1.7 wt.%.
2
3
3
4
2 3
2
◦
◦
◦
trometer, submitted to a reduction in H /Ar during 10 min, and
2
then evacuated at room temperature. The spectra were obtained
with pass energy of 30 eV. The spectra were processed using the
software Casa XPS (Casa Software Ltd., UK). The intensities were
estimated by calculating the integral of each peak after subtracting
a Shirley-type background and fitting the experimental curve to a
combination of Lorentzian and Gaussian lines. Bending energy val-
ues were referred to the C1 peak (284.6 eV). For the quantification
of the elements, sensitivity factors provided by the manufacturers
were used.
2.2. Catalysts characterization
BET surface areas were obtained using
a Quantachrome
Autosorb 1 analyzer. Pore volumes were estimated by means of
the t-plot.
Catalysts crystalline structures were characterized by X-ray
diffraction (XRD). The X-ray diffractograms were obtained with
a Shimadzu XD-D1 instrument with monochromator using CuK␣
radiation at a scanning rate of 4 min , from 2ꢀ = 5 to 100 .
Reducibility of metallic catalysts was studied by temperature-
programmed reduction (TPR). Temperature-programmed reduc-
tion (TPR) experiments were performed employing an OKHURA
TP-2002 S system, equipped with a thermal conductivity detec-
tor (TCD) detector. TPR runs were carried out with a heating rate
◦
−1
◦
◦
2.3. Catalytic activity
The catalytic activity was measured at atmospheric pressure
in a continuous-flow fixed-bed reactor, made of 5 mm internal
diameter quartz tube. The catalyst bed was supported with quartz
wool. Above the catalytic bed, quartz beads were loaded in order to
improve the heat transfer. The catalyst was pretreated under flow of
◦
−1
−1
in 5% H /Ar (30 mL min ). The temperature was
2
of 10 C min
increased from 20 C to 900 C.
◦
◦
−1
◦
−1
H (30 ml min ) by heating at 10 C min from room temperature
2

M.S. Zanuttini et al. / Catalysis Today 213 (2013) 9–17
11
Fig. 1. (A) XRD patterns of fresh Pt (1.7%)/␥-Al2O3 catalyst (grey) and ␥-Al2O3 (black). JCPDS-ICDD characteristic signals for Pt (dash line) and ␥-Al2O3 (solid line). (B) TEM
results, histograms of Pt (1.7%)/␥-Al2O3 and Pt(0.05)/␥-Al2O3.
◦
◦
to 300 C or 500 C and maintained at this temperature for 1 h.
Two
sets
of
)
experiments
and W/F = 38 gcath (gcresol
corresponding
to
were
−
1
−1
The catalyst was then cooled down to the reaction temperature
200–450 C). The carrier was bubbled in liquid cresol maintained
at a preselected temperature in the range 55–90 C, in order to
saturate the gas. The carrier flow rate was 5–50 ml min . Under
these conditions the cresol partial pressure in the gas stream fed
to the reactor was between 0.51 and 107.2 Torr, and the contact
W/F = 0.48 gcath (gcresol
repeated 5 times. A good reproducibility was achieved. Relative
standard deviation of product yields was 8%.
)
◦
(
◦
−
1
3
. Results and discussion
−
1
time (W/F) was between 0.1 and 38 gcath (gcresol
)
. After each run,
3
.1. Catalysts characterization
the catalyst bed was purged with H2 at the reaction temperature
during 30 min. Experiments feeding a mixture of cresol and phenol,
were also carried out. In this case, the mixture was fed by a syringe
pump (Apema). The gas flow rates were controlled with mass flow
controllers (Aalborg Instruments and Controls, Inc.,). The reactor
outlet stream was analyzed in a GC (SRI 8610) connected online and
equipped with a ZB-5 capillary column (15 m) and FID detector. A
split ratio of 100 was used. Standard samples were used in order to
identify the reaction products. In addition, a GC–MS (Varian Saturn
Fig. 1A shows the diffractograms corresponding to pure alumina,
and to the Pt(1.7%)/␥-Al O catalyst. JCPDS-ICDD characteristic sig-
2
3
nals for Pt (dash lines) and ␥-Al O3 (solid lines) are included. The
2
X-ray diffractogram presents the typical signals of the ␥-Al O at
2
3
◦
◦
◦
◦
◦
2
1
ꢀ = 45.901 , 67.093 , 37.635 , 39.524 , and 19.466 (JCPDS-ICDD
0-425). No signals corresponding to Pt crystals were observed.
The diffractograms obtained with the catalysts are equal to those
obtained with the support, what indicates that the metal parti-
cles are small. In the equipment used to obtain these results, the
minimum Pt size that displays signals in the XRD spectrum is
approximately 50 A˚ .
2
000) equipped with a HP-5 capillary column was used to iden-
◦
tify the reaction products collected in a condenser cooled at 0 C.
The non-condensed gases were analyzed by GC using a Petrocol
(
Supelco) capillary column (100 m).
Fig. 1B shows the histograms that correspond to the Pt(1.7%)/␥-
In order to evaluate catalytic activities at different contact times,
Al O and Pt(0.05%)/␥-Al O . It can be observed that the Pt is highly
−
1
2
3
2
3
the W/F was varied between 37.7 and 11.3 gcath (gcresol
)
chang-
dispersed in all samples, with particles diameter in the order of 14 A˚ .
ing the catalyst mass loaded in the reactor and maintaining the
hydrogen flow rate and the saturator temperature constant. Values
However, even though the mean particle size is not very different
between the Pt(1.7%) and Pt(0.05%) catalysts, there is an evident
difference in the amount of smaller particles. The catalyst with
.05% Pt has an important fraction of particles with diameter below
.2 nm. Nevertheless, the difference in the Pt particles size is not
−1
of W/F lower than 11.3 gcath (gcresol
)
were obtained by increasing
the cresol flow rate and maintaining the catalyst mass constant.
The selectivity of catalysts with different Pt contents were com-
pared at conversions near 80%, which was obtained changing the
0
1
significantly different between these two catalysts.
−1
W/F between 80 and 0.1 gcath (gcresol
)
.
The TPR profile of Pt(1.7)/␥Al O (not shown) displayed two
2
3
Conversion levels and product yields at different temperatures
◦
◦
peaks, one at 240 C and the second at 414 C, which can be assigned
to the bulk phase of the PtOx and to highly dispersed particles with
strong interaction with the support, as previously reported [38–40].
Results of surface area, pore volume and pore diameter are
shown in Table 2. The ␥-Al O catalytic surface area was 200
m g . The preparation procedure had little effect on the BET area,
the pore volume, and pore diameter, and consequently, it can be
expected that the activity results are not influenced by these neg-
ligible changes.
were analyzed with Pt(1.7)/␥Al O catalysts. The experiments were
2
3
carried out with the same catalyst sample, increasing the temper-
◦
◦
◦
ature from 200 C to 450 C, in 50 C steps. Afterwards the reactor
was progressively cooled down, following the reverse procedure,
using the same temperature values to determine the catalytic activ-
ity. This procedure was adopted to check the catalyst stability by
comparison of the activities obtained in the direct and the reverse
temperature sequence.
2
3
2
−1
The absence of internal and external mass transfer lim-
itations was verified using different catalyst particle sizes
−
1
3
(
W/F = 0.4 gcath (gcresol
)
), and H2 flow rates (5.5 and 50 cm /min)
3.2. Effect of reduction temperature
−
1
at two different W/F (W/F = 0.15 and 37 gcath (gcresol
)
). Under
◦
these conditions, which are the extremes values used in this study
in each variable, similar conversions and product distributions
were obtained. No heat-transfer limitations are expected since the
heats of reaction are relatively small.
The platinum catalysts were reduced at 300 and 500 C. The
activity results obtained in both cases were very similar (results
not shown). Because of this, throughout this study results obtained
◦
with the catalysts reduced at 300 C are presented.

1
2
M.S. Zanuttini et al. / Catalysis Today 213 (2013) 9–17
Table 2
Metallic particle sizes, calculated dispersions, BET surface, pore volume (Vg) and average pore diameter (dp) values for the fresh catalysts.
BET, Sg (m2
g
−1
)
Vg (ml g
−1
)
dp (Å)
Catalyst
TEM particle size, Dva (nm)
Calculated dispersion (%)
Pt(1.7%)/␥-Al2O3
Pt(0.5%)/␥-Al2O3
Pt(0.1%)/␥-Al2O3
Pt(0.05%)/␥-Al2O3
1.50
1.43
1.41
1.33
67
80
80
86
214.0
195.0
189.6
193.1
0.488
0.490
0.492
0.462
87.7
87.8
87.9
88.7
Table 3
XPS results.
3.4. Catalytic activity
◦
Activation temperature
A blank experiment was carried out at 500 C without catalyst
in order to determine conversion due to thermal decomposition.
◦
◦
C
3
00
C
500
In this case, only cresol was observed at the reactor outlet. Pure
Binding energy Pt 4d5/2 (eV)
%
314.0
314.1
317
54.7
45.3
◦
alumina was tested at 300 C, W/F = 11.3 h, and H /HC = 510. In
2
3
16.9
0
this case, the m-cresol conversion also was zero. The amount of
coke deposits detected by TPO analysis was 5.14%, with a maxi-
Pt
59.0
41
PtOx
◦
mum in the TPO profile at 490 C. Nevertheless, the alumina had a
light-pink colour after the reaction, while the coked catalysts were
grey or black, what indicates that there is a different nature of the
3.3. XPS studies
coke deposits between the alumina and the Pt/Al O3 catalysts. It
2
Fig. 2 shows the XPS spectra obtained with the Pt(1.7%)/␥-
is very interesting to highlight, that the amount of coke deposited
on alumina is five times higher than that on Pt(1.7)/Al O , what
◦
Al O catalyst after reduction at 300 and 500 C. Although the most
2
3
2
3
intense photoemission lines of platinum are those arising from the
f levels, this energy region is overlapped by the Al 2p line of ␥-
Al O . Therefore, in this article the Pt 4d line was used. This one is
demonstrates the importance of the metallic function not only in
changing the relative rates of the reactions involved in this system,
but also in the coking mechanism. Evidently, the fast deoxygen-
ation that occurs on the metal function leading to the formation of
toluene (as will be shown in the following sections) and the hydro-
gen spillover, lead to an important improvement in the catalyst
stability.
4
2
3
weaker but it is not overlapped by spectral lines of other compo-
nents. This assignment of the Pt oxidation state is difficult and the
information in literature is contradictory [41,42]. The low levels of
Pt in the samples, along with the broad nature of Pt 4d features,
introduce considerable uncertainty in the determination of bind-
ing energies. Fig. 2 presents the spectral region of Pt 4d5/2 line with
The conversion and yields of products vs. time on stream (TOS)
for different contact times (W/F) are shown in Fig. 3 for m-cresol
substration of a Shirley background spectrum of Pt(1.7%)/Al O
reduced one hour at 300 C and at 500 C. The decomposition of
◦
2
3
deoxygenation at 300 C. The main reaction products were toluene
Tol), benzene (Bz), methylcyclohexane (MCH) and light hydro-
◦
◦
(
the spectra to individual components revealed the presence of two
carbons (LH) including mainly methane, ethane, propane, butane,
and small amounts of n-pentane, isopentane, 2 and 3 methyl-
hexane and hexane. Phenol (Ph) was found in very low amounts.
GC–MS analyses of the condensed samples confirmed the GC prod-
ucts identification. Methylcyclohexanol and methylcyclohexene
were not observed. Small amounts of dimethyl phenyl, dimethyl
diphenyl, and a dimethyl benzophenone were also observed by
GC–MS. The conversion was lower than 100% only in the case of
◦
platinum states: the reduced state (Pt ) with a binding energy (E )
b
of 314.0–314.1 eV and the oxidized state with a E of 316.9–317 eV.
b
Table 3 gives the Pt 4d5/2 binding energies observed in the sam-
◦
ples after reduction at 300 and at 500 C, and the percentages of
the different states of platinum. These results show that there are
no significant differences between the two treatments. As stated
◦
by Serrano-Ruiz et al. [41], even after reduction at 500 C, Pt seems
+
to maintain some ı character, although this oxidation might occur
−1
W/F = 0.15 gcath (greactant) . These results are in good agreement
also during the analysis. These analyses explain the activity results
that showed that the cresol conversion and product distribution
were practically the same with both pretreatments.
with the conversion reported by Lobo et al. [48], who found that
◦
−1
at 260 C and W/F = 0.58 gcath (greactant) using Pt(1.7 wt.%)/Al O3
2
catalyst, the conversion was 38%.
A mixture of 50 wt.% phenol and 50 wt.% cresol was used in
order to elucidate the reaction path for benzene formation. If ben-
zene comes from toluene and from phenol, the benzene/toluene
ratio would be higher when feeding the mixture of phenol plus
cresol, since the only source of toluene is cresol. The contact time
2
2
1
1
500
000
500
000
−
1
(W/F) for this experiment was 0.1 gcath (greactant) . The conver-
sion levels and the benzene/toluene ratio corresponding to both
feeds, i.e., cresol and the mixture phenol-cresol, obtained with
Pt(1.7)/␥Al O , are shown in Fig. 4A and B. The conversion as a func-
5
00ºC
2
3
tion of time followed the same behaviour for both feeds. However,
the benzene/toluene ratio was much higher when feeding the mix-
ture of cresol and phenol, confirming that benzene was formed by
dehydroxylation of phenol at a high rate. Since the amount of ben-
zene is rather small under all the reaction conditions with m-cresol,
another important conclusion that can be drawn from these results
is that phenol formation rate on the Pt/␥Al2O3 catalysts is much
slower than the cresol dehydroxylation rate that forms toluene.
This is a very important result to be highlighted, since the selec-
tivity towards deoxygenated products would be very high. These
results also indicate that the transalkylation of m-cresol forming
5
00
3
00ºC
0
3
05
310
315
320
325
Binding energy (eV)
Fig. 2. Pt 4d spectra obtained by XPS for fresh Pt(1.7%)/␥-Al2O3 catalyst reduced at
◦ ◦
00 C (grey) and at 500 C (black).
3

M.S. Zanuttini et al. / Catalysis Today 213 (2013) 9–17
13
W/F= 37.8 h (gcat)(g )-1
W/F= 11.5 h (gcat)(gcr)-1
100
cr
1
00
80
60
40
20
8
0
6
4
2
0
0
0
0
0
100 200 300 400
TOS (min)
0
100 200 300 400
TOS (min)
0
0
20
40
60
80
TOS (min)
W/F= 4.3 h (gcat)(g )-1
W/F= 0.15 h (gcat)(gcr)-1
cr
1
00
Fig. 5. Toluene conversion and product yield vs. time on stream. Pt(1.7%)/␥Al2O3,
at 300 C, H2/toluene molar ratio = 50, W/F = 0.1 gcath (greactants) . References: (ꢀ)
toluene conversion, (×) LH, (ꢂ) benzene, (ꢃ) phenol, and (᭹) methylcyclohexane.
◦
−1
8
0
0
0
0
Table 4
Coke content on spent catalysts obtained by TPO.
6
4
2
−
1
Catalyst
W/F gcat h (gcresol)
H2/cresol molar ratio TPO%C
Pt(1.7%)/␥-Al2O3
0.15
65
197
510
510
1.93
1.19
1.05
0.39
4
.3
1.3
37.7
1
0
Toluene was fed in another experiment, with the Pt(1.7)/␥Al O3
0
100 200 300 400
0
100 200 300 400
2
catalyst, in order to investigate the other possible reaction
path for benzene formation, and also for methylcyclohexane
TOS (min)
TOS (min)
◦
production. The experimental conditions were T = 300 C and
Fig. 3. Cresol conversion and product yield vs. time on stream for different contact
times (W/F) over Pt(1.7%)/␥Al2O3 at 300 C. References: (ꢀ) m-cresol conversion,
ꢁ) toluene, (×) LH, (ꢂ) benzene, (ꢃ) phenol, and (᭹) methylcyclohexane.
◦
−1
W/F = 0.1 gcath (greactant) . Fig. 5 shows the results. Methylcy-
(
clohexane was the main product. This result indicates that the
Pt(1.7)/␥Al O3 catalyst was more selective to ring hydrogenation
2
phenol and dimethylphenol does not occur at an appreciable rate
due to the weak acidity of the alumina, in contrast to the result
obtained with a high acidity catalyst [43].
than hydrocracking under these reaction conditions, in agreement
with results obtained in the hydrogenation of phenol [43].
Coke deposits were determined by TPO analyses of the spent cat-
alysts. Two maxima were observed in all the TPO profiles, around
Benzene formation from phenol could occur by dehydration on
the acid sites of the alumina, and by CO bond hydrogenolysis on
the metal. There are not direct evidences regarding which site is
more active under the conditions used in these experiments. How-
ever the fast change in the benzene/toluene ratio shown in Fig. 4B,
and the low amount of coke deposited on the alumina as will be
shown below, suggests that the metal is playing an important role.
Nevertheless, this issue needs further studies.
◦
◦
250 C and 450 C respectively (not shown). The first peak is sharp,
◦
with a width of 50 C. These two maxima correspond to carbon
deposited on the metal or near the metal particles and on the
support, respectively [44–46]. The carbon contents on the spent
Pt(1.7)/␥Al2O3 catalyst as a function of the space velocity are sum-
marized in Table 4. The H2/cresol ratio was not constant in these
experiments. However, the last two points shown in the table have
the same H /cresol ratio, with higher amount of coke deposited
2
at shorter contact time, indicating that the deactivation follows a
parallel mechanism, with the coke being formed faster from cresol
than from the products. Most probably, the compounds deposited
on the catalyst are products of cresol condensation, similar to the
dimethyl benzophenone, detected in low amounts. The other com-
pounds present in the system, either oxygenated or not, leads to the
formation of coke, as it has been determined years ago in the study
of deactivation of reforming catalysts. Nevertheless, the relative
rates of coke formation is very different among these hydrocarbons,
and according to the results shown above, the cresol is the main
precursor. Nevertheless, since these data was obtained at 100% con-
version, additional information is needed in order to determine the
exact mechanism for coke deposition.
6
5
4
3
2
1
0
1
00
A
B
8
0
0
0
0
6
4
2
0
0
50 100 150 200
0
50 100 150 200
TOS (min)
TOS (min)
3.5. Effect of metal loading on product distribution
Fig. 4. (A) Cresol conversion; (B) benzene–toluene yield ratio; Feed: (ꢀ) cresol; (ꢀ)
◦
The cresol deoxygenation reaction was evaluated with dif-
ferent Pt loadings. Toluene is a preferred product compared to
phenol:cresol (1:1) mixture, Catalyst: Pt (1.7%)/␥Al2O3 at 300 C, H2/m-cresol molar
−1
ratio = 100, W/F = 0.1 gcath (greactants)
.

1
4
M.S. Zanuttini et al. / Catalysis Today 213 (2013) 9–17
Table 5
Products distribution obtained with different catalyst. W/F = 37.7 gcath (greactants)−1, H2/cresol molar ratio = 510. Temperature: 300 ◦C.
%
Pt
Conversion
MCH
LH
TOL
Bz
MCH/Tol
0
0
0
1
.05
.1
.5
.7
100
100
100
100
3.27
3.83
4.61
5.71
45.9
41.8
36.7
46.4
51.2
56.8
86.11
4.04
3.02
1.85
1.28
0.08
0.07
0.07
0.07
6.88
Table 6
Products distribution obtained with different catalysts. Temperature: 300 C.
◦
%
Pt
Conversion
MCH
LH
TOL
Bz
MCH/Tol
H2/Cresol
0
0
0
1
.05
.1
.5
.7
82.0
90.0
74.0
83.5
1.0
2.0
4.4
3.7
71.3
55.1
1.8
8.0
22.0
66.7
77.2
0.0
10.9
1.1
0.125
0.091
0.066
0.048
1483
1483
197
65
1.1
1.5
methylcyclohexane since it has a higher octane number. The Pt
loadings used in these experiments were 1.7%, 0.5%, 0.1%, and 0.05%.
The activity results are summarized in Table 5. All the catalyst
reached 100% conversion, what shows the very high activity of
this catalyst for m-cresol conversion. The catalyst with the lower
metal loading produced a significant amount of light hydrocar-
bons, as compared to the catalyst with 1.7% Pt, consequently, the
toluene yield was much lower in the former. As above described,
the catalyst with 0.05%Pt has smaller particles. It has been reason-
ably established, that the smaller the metal particle, the higher the
hydrogenolytic activity, nevertheless, the difference in size among
the catalysts was not large enough to explain the different prod-
uct distribution observed with each of them. On the other hand,
the catalyst with the higher Pt loading has a much higher ratio of
metal/acid sites. Consequently, the relative rates of the reactions
involved in the mechanism for cresol conversion are different, thus
affecting the product distribution. As the metal content increases,
the metal-catalyzed transformation of cresol to toluene occurs at
a faster rate than the acid-catalyzed condensation reaction, thus
forming a more stable compound and consequently decreasing the
rate of light hydrocarbons formation. It is important to emphasize
that each reaction involves several steps and different catalytic
functions. For example, the m-cresol dehydroxilation to toluene
involves a first hydrogenation step on the metal and then the dehy-
dration on the acid [48]. According to the activity results shown
above, the metal function controls the reaction rate of m-cresol
dehydroxylation.
The methylcyclohexane/toluene (MCH/Tol) ratio was practically
the same in all four catalysts The benzene/toluene ratio decreased
as the Pt loading increased, being 0.1 for the Pt(0.5%) and 0.01 for
the Pt(1.7%) (data from Table 5). Even though, under the condi-
tions used to obtain the data shown in this table the conversion
was 100%, this large difference between the two catalysts indicates
that there is a significant difference in the relative rates of dehy-
droxylation and demethylation, as the metal content increased. As
the Pt loading increased, the reaction rate of dehydroxylation form-
ing toluene increased relative to the rate of demethylation forming
phenol and then benzene. To verify this behaviour, experiments
100
8
0
60
4
0
0
0
2
0
100
200
300
400
H /Cresol molar ratio
2
Fig. 6. Cresol conversion and product yield vs. H /cresol molar ratio. Cata-
2
◦
−1
lyst: Pt(1.7%)/␥Al2O3 at 300 C, P = 1 bar, W/F = 6.4 gcath (gcresol) . References: (ꢀ)
m-cresol conversion, (ꢁ) toluene, (×) LH, (ꢂ) benzene, (ꢃ) phenol, and (᭹) methyl-
cyclohexane.
high selectivity to light hydrocarbons, the benzene production was
negligible.
In summary, as the metal content increased, the deoxygenation
reaction of m-cresol to toluene became the main reaction path,
being faster than both, the demethylation of cresol to phenol, and
the condensation followed by hydrocracking reactions.
3.6. Effect of the H2/cresol ratio on the product yields and stability
The effect of the H2/cresol ratio on the product yields was stud-
◦
−1
ied at 300 C on Pt(1.7)/␥Al2O3. The W/F was 6.4 gcath (gcresol) and
the H2/cresol ratio was varied by changing the H2 partial pres-
sure in a current of N2, and keeping constant the total flow rate
−
1
at 30 mL min . The conversion and product yields, after 20 min
of reaction, are shown in Fig. 6. The coke deposited on the cata-
lyst after 2.25 h of reaction was determined by TPO (see Table 7).
were carried out changing the W/F and the H /m-cresol ratio in
2
order to obtain conversions lower than 100%. Results are shown
in Table 6. It was possible to measure the activity with conver-
sions in the order of 80%, although working with very different W/F
Table 7
Value of %C obtained for spent Pt(1.7%)/␥Al2O3 catalyst at different H2/cresol molar
ratios and W/F 6.4 gcath (gcresol)
−
1
.
and H /m-cresol ratios. It is important to highlight that the toluene
2
2
H /cresol molar ratio
H2 partial pressure
%C
selectivity increases as the metal content increased, in agreement
with results shown in Table 5 using other reaction conditions.
The benzene/toluene ratio also follows the same trend as that
shown in Table 5, i.e., it decreased as the metal content increased,
except in the case of the Pt(0.05%) catalyst in which, due to the
0
4
114
250
25
0.00
0.13
0.22
0.50
0.83
1.50
1.22
0.80
0.40
0.33
6
4

M.S. Zanuttini et al. / Catalysis Today 213 (2013) 9–17
15
The higher the H /cresol ratio, the lower the amount of carbon
deposited on the catalyst. Fig. 6 shows that toluene yield was
100
80
2
A
B
higher and benzene yield was lower as H /cresol ratio increased.
2
Methylcyclohexane was observed only at the highest H /cresol
2
ratio. These results makes it possible to conclude that as hydrogen
partial pressure increases, the hydrodeoxygenation rate (forming
toluene) increases relative to the demethylation reaction (forming
phenol). Because of this, lower amount of benzene was observed
60
40
20
0
at higher H /cresol ratio, since as indicated by the results shown in
2
Figs. 4 and 5 the rate of toluene transformation to benzene is slower
than the phenol dehydroxylation to benzene.
200 300 400 500
200 300 400 500
3.7. Effect of temperature on product selectivity
Temperature (ºC)
Temperature (ºC)
Conversion levels and product yields at different tempera-
Fig. 7. Cresol conversion and product yield vs. temperature. Catalyst: Pt
−
1
(
4
1.7)/␥Al2O3. W/F = 37.7 gcath (greactants) . (A) Increasing temperature from 200 to
50 C. (B) Decreasing temperature from 450 to 200 C. References: (ꢀ) m-cresol
tures were analyzed using Pt(1.7)/␥Al O catalyst. Results obtained
2
3
◦
◦
while increasing the temperature are shown in Fig. 7A. The yields
obtained at each temperature while decreasing its value were not
exactly the same as those obtained while increasing the temper-
ature, indicating that there was deactivation. Three temperature
ranges can be clearly distinguished. In the lower temperature
conversion, (ꢁ) toluene, (×) LH, (ꢂ) benzene, (ꢃ) phenol, and (᭹) methylcyclohex-
ane.
reaction can be successfully carried out at different tempera-
ture levels and atmospheric pressure, according to the desired
products.
◦
range (below 250 C), the main product was methylcyclohexane.
Dehydroxylation also occurred in the lower temperature range,
but no hydrocracking took place. In the intermediate tempera-
◦
ture range (250–350 C) the cresol dehydroxylation was favoured,
3.8. Cresol deoxygenation routes. Catalyst stability
being toluene the main product; and in the higher temperature
◦
range (above 350 C), benzene and light hydrocarbons were the
The reaction routes proposed for m-cresol deoxygenation are
shown in Scheme 1, in which each step may involve one or more
intermediates not observed in this study due to their very low con-
centrations in the gas phase. Toluene is formed by dehydroxylation
of cresol. Phenol is formed by hydrogenolysis of the cresol methyl
group. Two different routes generate benzene. Hydrogenolysis of
toluene methyl group produces benzene and methane, while dehy-
droxylation of phenol forms benzene. The latter reaction can occur
both on the metal and on the acid sites [48]. It has been proposed
main reaction products. It is important to highlight that deac-
tivation led to an increase in selectivity towards toluene, and
a decrease towards light hydrocarbons and methylcyclohexane
compare Fig. 7A and B).
Methane was formed by the hydrocracking of cresol condensa-
tion products, leading to a methane yield higher than the benzene
yield, as observed in the activity data obtained while increasing the
temperature. It can be concluded that the cresol deoxygenation
(
OH
OH
COKE
+
CH3
CH3
OH
OH
Condensation
products
+ H2
CH3
-CH4
m- Cresol
Phenol
-
H O
+ H2
- H O
+ H2
2
2
Light
Hydrocarbons
+
H2
+
H2
CH3
-CH4
Toluene
Benzene
CH
3
Methylcyclohexane
Scheme 1. Proposed reaction routes for m-cresol deoxygenation over Pt/␥Al2O3.

1
6
M.S. Zanuttini et al. / Catalysis Today 213 (2013) 9–17
that this reaction occurs on the metal after the aromatic ring is
partially hydrogenated, and then the C O bond breakage occurs
followed by dehydrogenation of the ring [47]. The hydrogenoly-
sis of the toluene methyl group was confirmed by feeding toluene
to the reaction system. In this reaction, benzene and methylcy-
clohexane were found as the reaction products, although benzene
appeared in very low amounts. The other reaction path for ben-
zene formation, i.e. phenol dehydroxylation, was confirmed by
feeding the mixture of phenol and cresol. This reaction was com-
pared with the reaction carried out under the same conditions
but feeding only cresol. As seen in Fig. 4, the benzene/toluene
ratio for the same conversion level was much higher when feed-
ing the mixture phenol–cresol, confirming that benzene was also
formed from phenol but at a higher rate than the demethylation of
toluene. Finally, methylcyclohexane was formed by hydrogenation
of toluene.
language editing. Thanks are also given to ANPCyT for the
purchase of the SPECS multitechnique analysis instrument (PME8-
2003).
References
[
[
1] Y. Sheu, R.G. Anthony, E.J. Soltes, Fuel Processing Technology 19 (1988)
31–50.
2] E.G. Baker, D.C. Elliott, in: A.V. Bridgwater, J.L. Kuester (Eds.), Thermo chemi-
cal Biomass Conversion, Elsevier Application Sciences, London, UK, 1988, pp.
883–895.
[3] F.A. Agblevor, S. Besler, Energy & Fuels 10 (1996) 293–298.
[
4] W. Craig, E. Coxworth, in: C. Granger, B.C. Richmond (Eds.), Proceedings of the
Sixth Canadian Bioenergy R&D Seminar Canada, Elsevier Application Sciences,
London, 1987, pp. 407–411.
[5] R. Maggi, B. Delmon, in: A.V. Bridgwater (Ed.) Advances in Thermochemical
Biomass Conversion, London, 1993, 1185 pp.
[
6] R. Maggi, A. Centeno, B. Delmon, Biomass energy environment, in: P. Chartier
Ed.), Proceedings of the 9th European Bioenergy Conference, Elsevier, Oxford,
1996, p. 327.
(
The conversion levels were 100% for contact times between
−1
3
7.7 and 4.34 gcath (gcresol
)
(Fig. 3). However, the product yield
[7] P. Grange, A. Centeno, R. Maggi, B. Delmon, Thermochemical biomass process,
in: A.V. Bridgwater, E.N. Hogan (Eds.), Bio-oil Prod. Util. Proc. EU-Can Workshop
Therm Biomass Process 2nd 1995, CPL Press, Newbury, 1996, p. 186.
changed with time on stream, indicating the progressive catalyst
deactivation. In all cases it was observed that as the reaction pro-
ceeded, benzene and methane yields decreased while toluene yield
increased. A significant amount of methylcyclohexane was only
observed for the highest contact time. As the reaction proceeded,
carbon deposits were formed both on metal and acid sites and, con-
sequently, the benzene and methylcyclohexane yields decreased.
Besides, the coplanar type adsorption needed in order to hydro-
genate toluene to methylcyclohexane was also more difficult as
coke deposition occurred. It is very important to highlight that
according to the TPO profile, coke was formed mainly on the
metal particles, and therefore, the observed deactivation might be
related to the activity loss of the metallic function. Other prod-
ucts found by GC–MS in low amounts were dimethyl diphenyl and
dimethyl benzophenone. Methyl groups from these compounds
can be hydrocracked, and because of this the light hydrocarbon
fraction (which contained an important amount of methane) yield
was higher than the benzene yield. According to these results,
the secondary hydrocracking reactions deactivated faster than the
cresol dehydroxylation reaction. Therefore, as the reaction pro-
ceeded less methane was produced and its yield (not shown)
followed the same trend as the benzene yield.
[
8] C.A. Fisk, T. Morgan, Y. Ji, M. Crocker, C. Crofcheck, S.A. Lewis, Applied Catalysis
A-General 358 (2009) 150–156.
[
9] Y. Wang, T. He, K. Liu, J. Wu, Y. Fang, Bioresource Technology 108 (2012)
2
80–284.
10] J.D. Adjave, N.N. Bakhshi, Fuel Processing Technology 45 (1995) 161–183.
[11] S.P.R. Katikaneni, J.D. Adjave, N.N. Bakhshi, Energy & Fuels 9 (1995) 1065–1078.
[
[
[
[
12] P.T. Williams, P.A. Horne, Fuel 74 (1995) 1839–1851.
13] J.D. Adjave, N.N. Bakhshi, Fuel Processing Technology 45 (1995) 185–202.
14] J.D. Adjave, S.P.R. Katikaneni, N.N. Bakhshi, Fuel Processing Technology 48
(1996) 115–143.
[
[
15] F. Janssen, Catalysis Science Series (1999) 1–4.
16] V.N. Bui, D. Laurenti, P. Afanasiev, C. Geantet, Applied Catalysis B-
Environmental 101 (2011) 239–245.
[17] W. Wang, Y. Yang, H. Luo, T. Hu, W. Liu, Catalysis Communications 12 (2011)
36–440.
18] Y. Romero, F. Richard, Y. Reneme, S. Brunet, Applied Catalysis A-General 353
2009) 46–53.
4
[
(
[19] E. Odebunmi, D.F. Ollis, Journal of Catalysis 80 (1983) 56–64.
[20] R.K.M.R. Kallury, T.T. Tidwell, D.G.B. Boocock, D.H.L. Chow, Canadian Journal of
Chemistry 62 (1984) 2540–2545.
[
21] C.R. Lee, J.S. Yoon, Y-W. Suh, J-W. Choi, J-M. Ha, D.J. Suh, Y-K. Park, Catalysis
Communications 17 (2012) 54–58.
[22] B. Yoosuk, D. Tumnantong, P. Prasassarakich, Fuel 91 (2012) 246–252.
[
[
[
[
[
23] B. Yoosuk, D. Tumnantong, P. Prasassarakich, Chemical Engineering Science 79
2012) 1–7.
24] Y. Romero, F. Richard, S. Brunet, Applied Catalysis B: Environmental 98 (2010)
213–223.
25] V.N. Bui, D. Laurenti, P. Delichere, C. Geantet, Applied Catalysis B: Environmen-
tal 101 (2011) 246–255.
26] Eun-Jae Shin, M.A. Keane, Industrial & Engineering Chemistry Research 39
(
4
. Conclusions
(
2000) 883–892.
27] J.D. Adjaye, N.N. Bakhshi, Biomass & Bioenergy 8 (3) (1995) 131–149.
Cresol deoxygenation was successfully carried out with
[28] P.D. Chantal, S. Kaliaguine, J.L. Grandmaison, Applied Catalysis. 18 (1985)
33–145.
1
Pt/␥Al O catalysts at atmospheric pressure; toluene, benzene and
2
3
[
29] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, J. Bilbao, Industrial & Engineer-
ing Chemistry Research 43 (2004) 2610–2618.
methylcyclohexane being the main reaction products depending
upon the reaction conditions. The yield of the desired product can
[30] A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, J. Bilbao, Industrial & Engineer-
ing Chemistry Research 43 (2004) 2619–2626.
be regulated by changing the metal loading, the H /cresol ratio and
2
[
31] R.N. Olcese, M. Bettahar, D. Petitjean, B. Malaman, F. Giovanella, A. Dufour,
Applied Catalysis B-Environmental 115-116 (2012) 63–73.
the reaction temperature. The toluene hydrogenation to methylcy-
clohexane is favoured at low temperature, while the toluene yield
[32] M.A. González-Borja, D.E. Resasco, Energy & Fuels 25 (2011) 4155–4162.
[33] J. Filley, C. Roth, Journal of Molecular Catalysis A: Chemical 139 (1999) 245–252.
[34] A. Ausavasukhi, Y. Huang, A.T. To, T. Sooknoi, D.E. Resasco, Journal of Catalysis
has a maximum at the intermediate temperature levels (around
◦
3
00 C), since at high temperatures toluene is transformed into
290 (2012) 90–100.
benzene by demethylation. Catalyst deactivation is probably due
mainly to cresol condensation products deposited on the catalyst
surface. The amount of coke deposited during m-cresol deoxy-
[
35] P.C. Aben, Journal of Catalysis 10 (1968) 224–229.
[36] M.G. White, Heterogeneous Catalysts, Prentice Hall, Englewood Cliffs, NJ, 1991,
pp. 88.
[
37] S.C. Fung, C.A. Querini, Journal of Catalysis 138 (1992) 240–254, 48.
genation on the Pt/␥Al O3 is rather low, in the order of 1 wt.%,
2
[38] C. Meephoka, C. Chaisuk, P. Samparnpiboon, P. Praserthdam, Catalysis Commu-
nications 9 (2008) 546–550.
with an important fraction of coke formed on the metallic parti-
cles.
[
39] M.J. Tiernan, O.B. Finlayson, Applied Catalysis B-Environmental 19 (1998)
3–35.
2
[
40] A.C.S.F. Santos, S. Damyanova, G.N.R. Teixeira, L.V. Mattos, F.B. Noronha,
F.B. Passos, J.M.C. Bueno, Applied Catalysis A-General 290 (2005)
Acknowledgments
123–132.
[
[
41] J.C. Serrano-Ruiz, G.W. Huber, et al., Journal of Catalysis 241 (2) (2006) 378–388.
42] A.S. Ivanova, E.M. Slavinskaya, et al., Applied Catalysis B: Environmental 97
The authors wish to acknowledge the financial support received
from ANPCyT, CONICET and CAID+D-UNL. Thanks are also given
to Guillermo Pitich, Diego López Delzar and José Ignacio Lobos
for the technical support, and to Elsa Grimaldi for the English
(
1–2) (2010) 57–71.
[
43] A.K. Talukdar, K.G. Bhattacharyya, Applied Catalysis A-General 96 (1993)
229–239.
[44] J. Barbier, E. Churin, P. Marecot, Journal of Catalysis 126 (1990) 228–234.

M.S. Zanuttini et al. / Catalysis Today 213 (2013) 9–17
17
[
[
[
45] J. Barbier, in: G.F. Forment, B. Delmon (Eds.), Catalyst deactivation, Elsevier,
Amsterdam, 1987, p. 1.
46] J. Barbier, E. Churin, J.M. Parera, J. Riviere, Reaction Kinetics and Catalysis Letters
[48] P.T.M. Do, A.J. Foster, J. Chen, R. Lobo, Green Chemistry 14 (2012) 1388.
[49] A.J. Foster, P.T.M. Do, R. Lobo, Topics in Catalysis 55 (2012) 118–128.
[50] H. Huang, R.V. Chaudhari, B. Subramaniam, Topics in Catalysis 55 (2012)
129–139.
2
9 (1985) 323–330.
47] R. Taylor, K.H. Lundlum, Journal of Physical Chemistry 76 (1972) 2882–2886.