Hydrodeoxygenation of guaiacol over ReS2/activated carbon catalysts. Support and Re loading effect

Article abstract of DOI:10.1016/j.apcata.2014.01.057

Two activated carbons (GAC and CGran Norit) were used as supports to prepare Re/C catalysts (0.4 atoms of Re per nm² of support). In addition, Re(x)/GAC catalysts with loadings of 0.5, 0.6, 0.7 and 0.8 atoms of Re per nm² of support were prepared. The activity difference between Re(0.4)/CGran and Re(0.4)/GAC catalysts was related to the amount of surface oxygen functional groups on the support that is resistant to sulfidation. Meanwhile, the activity of the Re(x)/GAC catalysts as a function of Re content was correlated with ReS₂ dispersion; the most active catalyst was the one with Re loading of 0.6 atoms per nm² of support. On the other hand, the selectivity, expressed in terms of the phenol/catechol ratio, did not change with Re loading but was rather mainly influenced by the acid strength of the support.

Full text of DOI:10.1016/j.apcata.2014.01.057

Applied Catalysis A: General 475 (2014) 427–437
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Hydrodeoxygenation of guaiacol over ReS2/activated carbon catalysts.
Support and Re loading effect
C. Sepúlvedaa, R. Garcíaa, P. Reyesa, I.T. Ghampsonb, J.L.G. Fierroc, D. Laurentid,
M. Vrinatd, N. Escalonaa,∗
a Universidad de Concepción, Facultad de Ciencias Químicas, Casilla 160C, Concepción, Chile
b Unidad de Desarrollo Tecnológico, Universidad de Concepción, Casilla 4051, Concepción, Chile
c Instituto de Catálisis y Petroquímica, CSIC, Cantoblanco, 28049 Madrid, Spain
d IRCELYON, UMR 5256 CNRS, Université Lyon 1, 2 avenue A. Einstein, 69626 Villeurbanne cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 1 October 2013
Received in revised form 21 January 2014
Accepted 30 January 2014
Available online 7 February 2014
Two activated carbons (GAC and CGran Norit) were used as supports to prepare Re/C catalysts (0.4 atoms
of Re per nm2 of support). In addition, Re(x)/GAC catalysts with loadings of 0.5, 0.6, 0.7 and 0.8 atoms
of Re per nm2 of support were prepared. The activity difference between Re(0.4)/CGran and Re(0.4)/GAC
catalysts was related to the amount of surface oxygen functional groups on the support that is resistant to
sulfidation. Meanwhile, the activity of the Re(x)/GAC catalysts as a function of Re content was correlated
with ReS2 dispersion; the most active catalyst was the one with Re loading of 0.6 atoms per nm2 of support.
On the other hand, the selectivity, expressed in terms of the phenol/catechol ratio, did not change with
Re loading but was rather mainly influenced by the acid strength of the support.
Keywords:
Hydrodeoxygenation
2-Methoxyphenol
Activated carbon
ReS2
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
In general, the catalysts used for HDO of bio-oil or guaiacol
were mainly sulfided Co-Mo and Ni-Mo supported on alumina
Due to limited oil reserves and the high demand for fuel and
petrochemical feedstock, it has become necessary to find new,
efficient and economical sources of hydrocarbons. The incentive
to look for an alternative energy is not limited to economic and
energy security issues; in fact, the threat of climate change has also
spurred countries worldwide to explore environmentally benign
energy sources. Lignocellulosic biomass is one of the most impor-
tant renewable and sustainable sources of hydrocarbons on earth
from which transportation fuel (biofuel) can be obtained. One of
the main processes to produce biofuel from biomass is pyroly-
sis. This pyrolytic so-called “bio-oil” has high oxygen (35–40 wt.%
dry base) and water (15–25 wt.%) contents. The presence of oxy-
genated compounds renders the oil thermally unstable, corrosive
and with a low heating value (25 MJ kg−1). For this reason, the qual-
ity of bio-oil must be improved through partial or total removal
of the oxygenated functionalities. One method for bio-oil upgrad-
ing is catalytic hydrodeoxygenation (HDO). Several studies on HDO
of bio-oil and guaiacol (or 2-methoxyphenol, as a model of bio-oil
refractory component) have been carried out in the last years [1–4].
[1,2]. Other supports such as activated carbon have received less
attention even though it has the potential to improve the over-
all catalytic reactivity. For instance, Centeno et al. [3] found that
activated carbon-supported catalysts favored direct elimination of
methoxy groups of guaiacol with negligible coke formation. The
authors suggested that appropriate modifications of these cata-
lysts can lead to more effective bio-oil stabilization reactions. De
la Puente et al. [4] studied the nature of the surface of activated
carbon supports on the activity and selectivity of CoMo catalysts
for the HDO of guaiacol. They found that the surface chemistry of
the support influences the degree of precursor/support interaction
and therefore the nature of the phases present on the sulfide cata-
lyst after activation. The formation of different metal environments
may be responsible for the differences in selectivity displayed by
these catalysts. The results suggest that the catalytic conversion of
biomass-derived oil can be controlled by appropriate changes to
the surface chemistry of the activated carbon supports.
Recently, Sepúlveda et al. [5] studied the HDO of guaiacol over
activated carbons-supported Mo2N catalysts and found that the
Mo2N dispersion was related to the surface chemistry of the sup-
port, and that catalytic activity was partially related to support
porosity.
Different noble metals such as Ru, Pd, Rh, and Pt catalysts
on different supports have been studied in HDO of guaiacol,
∗
Corresponding author. Tel.: +56 41 2207236; fax: +56 41 2245374.
E-mail address: escalona.nestor@gmail.com (N. Escalona).
0926-860X/$ – see front matter © 2014 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2014.01.057

428
C. Sepúlveda et al. / Applied Catalysis A: General 475 (2014) 427–437
Table 1
Physical characteristics, isoelectric point (IEP), zero point charge (ZPC), total acidity and acid strength (Eo) of supports.
*
**
***
Supports
SBET (m2 g−1
)
Vp (cm3 g−1
)
Vm (cm3 g−1
)
Vo (cm3 g−1
)
ZPC (pH)
IEP (pH)
ZPC-IEP (pH)
Total acidity
Eo (mV)
(mmol g−1
)
CGran
GAC
1477
1039
1.11
0.62
0.72
0.17
0.39
0.45
4.7
5.4
3.2
4.1
1.5
1.3
0.80
0.30
171.5
−45.1
*
Vp = total volume.
Vm = mesopore volume.
Vo = micropore volume.
**
***
displaying similar catalytic activities which were comparable or
higher than conventional sulfide catalysts [6]. This study demon-
strated the importance of exploring other active phases beside
conventional metal sulfides to improve the catalytic upgrading pro-
cess. Among unconventional metal sulfide catalysts supported ReS2
have been rarely studied despite having been demonstrated to dis-
play high activity in hydrodesulfurization process [7–9]. Recently,
Ruiz et al. [10] studied the effect of the sulfiding mixture (H2S/H2
and H2S/N2) over Re/ZrO2 and Re/ZrO2-sulfated catalysts in the
hydrodeoxygenation of 2-methoxyphenol. They found that the
reaction rates for both catalysts increased upon sulfiding in H2S/N2
mixture and that Re/ZrO2 catalyst sulfided with H2S/N2 mixture
was 2.8 times more active than the NiMoS/Al2O3 catalyst. The same
behavior was observed previously by Escalona et al. in hydrodesul-
furization reactions [11].
The objective of this work was to study the effect of the textural
and surface properties of activated carbon-supported ReS2 cata-
lysts in the HDO of guaiacol. Additionally, the effect of Re content
was also studied. We believe that determining the factors that con-
tribute to the catalytic activity of ReS2/activated carbon catalysts is
important in the exploration, design and development of catalysts
for the biorefinery process.
properties of the catalysts are presented in Table 2. The Re con-
tent was determined by ICP using Re emission line of 221.426 nm.
Before reaction the catalysts were sulfided at 673 K for 4 h under
flowing 15 vol.% H2S in N2, similarly to previous studies [10,11] that
clearly demonstrated the advantages of activation under an H2S/N2
mixture instead of an H2S/H2 mixture.
2.2. Characterization of supports and catalysts
The BET “apparent” specific surface area (SBET) and pore vol-
umes of the supports and catalysts were determined from nitrogen
adsorption measurements at 77 K using a Micromeritics TriStar II
3020 equipment. Prior to the measurements, the samples were out-
gassed at 573 K for 2 h. Micropore volume (Vo) was determined by
the Dubinin–Radushkevich method while the total pore volume
(Vp) was recorded at nitrogen adsorption at a relative pressure
of 0.99. Mesopore volume (Vm) is obtained from the difference
between Vp and Vo.
Temperature-programmed decomposition (TPD) analyses of the
activated carbons were carried out under an inert atmosphere at
a heating rate of 18 K min−1 up to 1373 K and a helium flow of
50 mL min−1. The amounts of gases evolved (CO2 and CO) were
measured by a thermal conductivity detector (TCD) coupled to a
non-dispersive infrared device.
2. Experimental
FT-IR analyses of the supports were performed on a Nicolet
Nexus FT-IR in the middle range (4000–400 cm−1) and averaged
after 64 scans. The tablets were prepared using KBr as a support in
a 1:100 mg proportion of sample to support.
2.1. Catalyst preparation
The ReS2/C catalysts were prepared using two commercial acti-
vated carbons (obtained from Norit Americas, Inc) NORIT GAC 1240
Plus and NORIT CGran supports whose characteristics are reported
in Table 1. The supports were ground and sieved to obtain between
80 and 125 ␮m particle size. The catalysts were prepared by incip-
ient wetness impregnation using aqueous solution of NH4ReO4
(Aldrich, 99%) to obtain catalysts with a nominal metal loading
of 0.4 Re atoms per nm2 on each support. Furthermore, catalysts
with varying Re loadings (0.5, 0.6, 0.7 and 0.8 Re atoms per nm2
of support) were also prepared using NORIT GAC 1240 support [7].
After impregnation, the samples were left at room temperature for
20 h and further dried at 383 K for 6 h. The catalysts are denoted
by the number of metal atoms per nm2 of the initial surface area,
e.g. Re (0.36) contains 0.36 Re atoms (10.34 wt.% of Re2O7) per nm2
of GAC activated carbon support. The Re contents and the physical
X-ray powder diffraction (XRD) patterns were obtained by
means of a Rigaku diffractometer equipped with a nickel-filtered
˚
CuK␣1 radiation (ꢀ = 1.5418 A).
Electrophoresis measurement (EM) was carried out in a Zeta-
Meter 3.0+ apparatus using 20 mg of sample suspended in 200 mL
of a 1 × 10−3 mol L−1 KCl solution. The pH was adjusted with either
0.1 mol L−1 HCl or KOH solutions. The IEP (isoelectric point) was
obtained from the plot of pH vs. zeta potential. The Zero Point
Charge (ZPC) of the two carbons was determined by mass titra-
tion using the ASTM method D3838-05. Approximately 50 mL of
hot distilled water was added to carbon in a flask. The flask was
heated on a hot plate and a thermometer was used to ensure that
no false boiling, due to trapped gases, occurred. The flask was
then removed from the hot plate and its contents were filtered
Table 2
Composition and physical characteristics of oxidic catalysts.
*
**
***
Catalysts
Re loading
(wt.% Re2O7)
Re loading
SBET (m2 g−1
)
Vp (cm3 g−1
)
Vm (cm3 g−1
)
Vo (cm3 g−1
)
(atoms nm−2
)
Re(0.4)/CGran
Re(0.4)/GAC
Re(0.5)/GAC
Re(0.6)/GAC
Re(0.7)/GAC
Re(0.8)/GAC
14
10
13
16
18
20
0.4
0.4
0.5
0.6
0.7
0.8
906
664
639
625
615
444
0.79
0.40
0.33
0.29
0.28
0.20
0.42
0.10
0.05
0.05
0.04
0.02
0.37
0.30
0.28
0.24
0.24
0.18
*
Vp = total volume.
Vm = mesopore volume.
Vo = micropore volume.
**
***

C. Sepúlveda et al. / Applied Catalysis A: General 475 (2014) 427–437
429
immediately through pre-moistened filter paper. The pH measure-
ment was performed at 323 5 K.
The total acidity measurements of the carbon supports and cat-
alysts were carried out potentiometrically by titrating a suspension
of carbon in acetonitrile with n-butylamine using an Ag/AgCl elec-
trode [12].
900
750
600
450
300
150
0
GAC
CGran
X-ray photoelectron spectra (XPS) of sulfide catalysts were
obtained on an Escalab 200R electron spectrometer using Mg K␣
(1253.6 eV) as the excitation source. Spectra were recorded at 45◦
take-off angles by a concentric hemispherical analyzer operating in
the constant pass energy mode at 50 eV. Under these conditions the
Au 4f7/2 line was recorded with 1.34 eV FWHM at a binding energy
of 84.0 eV. The spectrometer energy scale was calibrated using Cu
2p3/2, Ag 3d5/2 and Au 4f7/2 photoelectron lines at 932.7, 368.3 and
84.0 eV, respectively. Charge referencing was done against adventi-
tious carbon (C 1s 284.8 eV). The catalyst samples were pre-sulfided
ex situ under flowing 15 vol.% H2S/N2 at 673 K for 4 h. The samples
were then cooled to room temperature, flushed with N2 and stored
in flasks containing N2, and then transferred to the pre-treatment
chamber of the spectrometer. Then, sulfide catalysts were mounted
on a sample holder and kept overnight in high vacuum in the
preparation chamber before they were transferred to the analy-
sis chamber of the spectrometer. The C1s, S2p and Re4f energy
regions were scanned with several sweeps until a good signal-to-
noise ratio was observed. The pressure in the analysis chamber was
maintained lower than 5 × 10−9 mbar.
0,0
0,2
0,4
0,6
0,8
1,0
Relative pressure (P/Po)
Fig. 1. N2 adsorption–desorption isotherms at 77 K of activated carbon supports.
taken periodically during the reaction and identified by Perkin-
Elmer Clarus 680 a coupled to GCMS-SQ8T equipped and quantified
by a Perkin-Elmer (Clarus 400) gas chromatograph equipped with
a Flame Ionization Detector (FID) and a CP-Sil 5 column (Agilent,
30 m × 0.53 mm × 1.0 ␮m film thickness). The specific rate for the
total conversion of guaiacol was deduced from the initial slope of
conversion vs. reaction time plot. The intrinsic rate was calculated
from the specific rate according to the following equation:
Spectra were decomposed with the Casa XPS program (Casa
Software Ltd., UK) by using a Gaussian/Lorentzian (90/10) prod-
uct function after subtraction of a linear background. Atomic
ratios were calculated by using peak areas normalized on the
basis of acquisition parameters, sensitivity and transmission factors
provided by the manufacturer, and determined from the corre-
sponding peak intensities, corrected with tabulated sensitivity
factors, with a precision of 7%.
The theoretical dispersion of the Re phase over the microporous
and micro–mesoporous catalysts were calculated according to the
model described by Kerkhof and Moulijn [13]. The model assumes
that the catalyst consists of sheets of the support of thickness t
with cubic crystallites of the supported phase of a dimension c in
between. For a monolayer dispersion of the supported metal phase,
the predicted XPS metal (Im) to support (Is) intensity ratio can be
expressed as in equation (1):
Sr
Ir =
× N
(2)
nRe
where Ir is the intrinsic rate (i.e. molecules of guaiacol transformed
per Re atom per second and expressed in molec. Re at−1 s−1), and Sr
is the specific rate (i.e. the moles of guaiacol−tr1ansformed per gram
of catalyst per second and expressed in mol gcat s
−1), nRe represents
the number of Re atoms per gram of catalyst, and N is the Avogadro
number. The deoxygenation (HDO) rate was calculated from the
conversion of guaiacol to O-free compounds (benzene, cyclohex-
ane, cyclohexene and methylated analogs). The HDO rate was also
calculated according to Eq. (2), where the specific activity repre-
sents the conversion of guaiacol into deoxygenated compounds.
On the other hand, the selectivities (%) were determined at 10% of
guaiacol conversion.
Im
Is
(m/s)(Dm/Ds)(ꢁm/ꢁs)(ˇs/2)[1 + exp(−ˇm)]
[1 + exp(−ˇm)]
=
(1)
where ˇs = t/ꢀs (ꢀs is the support photoelectron escape depth, t is
the thickness of the support), ˇm = t/ꢀm (ꢀm is the metal photoelec-
tron escape depth) and m/s is the metal/support atomic ratio. Ds and
Dm are the detector efficiencies, and ꢁs and ꢁm are the cross sections
for photoelectron emission of the support and metal, respectively.
The thickness of the sheets (t) can be estimated from the density
and the surface area of the support: t = 2/ꢂsS.
3. Results and discussion
3.1. Supports characterization
The nitrogen adsorption–desorption isotherms for the carbon
supports are shown in Fig. 1. The isotherms of the supports are
type I according to IUPAC classification, which are characteristics
of microporous materials [14]. However, at high relative pres-
sure an increase in the amount of the nitrogen adsorbed, due to
the presence of mesopores, can be observed. This mesoporosity
is more prominent in the CGran support. Also, Fig. 1 shows that
both isotherms displayed a type H4 hysteresis loop, suggesting
lamellar-shaped pores. The textural properties of the supports are
summarized in Table 1: CGran support displayed higher surface
area than the GAC support. Also, the micropore volume (Vo) of the
CGran support represents about 35% of the total pore volume (Vp)
while that of GAC support occupies about 72% of the total pore
volume.
2.3. Catalytic tests
The HDO of guaiacol was carried out in an autoclave reactor
operating in batch mode. The liquid reactant feed, consisting of
guaiacol (0.232 mol L−1) in n-dodecane (80 mL) with hexadecane
(0.0341 mol L−1) as an internal standard, was introduced into the
reactor. Then, approximately 0.200 g of freshly sulfided catalyst was
added. The system was closed, and to avoid any air contamination,
N2 was bubbled through the solution for 10 min. Still under N2 the
reactor was heated to the reaction temperature of 573 K under stir-
ring. When the reaction temperature was reached, the pressure was
adjusted to 5 MPa by H2 introduction into the reactor and kept con-
stant during the course of the experiment. Condensed samples were
Differences in the chemical nature of the oxygen functional
groups of the supports were determined by means of TPD profiles,
as shown in Fig. 2. The CGran support displayed a weak peak at

430
C. Sepúlveda et al. / Applied Catalysis A: General 475 (2014) 427–437
25000
25000
20000
15000
10000
5000
0
Re(0,4)/GAC
Re(0,4)/CGran
b)
a)
20000
15000
CO
10000
CO
5000
CO2
CO2
0
400
600
800
1000
1200
1400
400
600
800
1000
1200
1400
Temperature (K)
Temperature (K)
Fig. 2. Thermal-programmed decomposition (TPD) of a) GAC, b) CGran activated carbons supports.
1300–900 cm−1, assigned to C–O–C vibrations or C–O stretching of
ether, lactonic carboxylic or phenolic groups [19]. The CGran sup-
port has a more intense band at 1573 cm−1 which can be assigned
to quinonic groups, in agreement with its TPD profile. Fig. 3 clearly
shows that the CGran support displayed the most intense bands,
suggesting that it has the highest amount of oxygen functional
groups, in agreement with results deduced from TPD profiles.
The IEP and ZPC results of the GAC and CGran activated carbon
supports are summarized in Table 1. It can be observed that the
higher ZPC value was presented by the GAC activated carbon sup-
port. Also, Table 1 shows that CGran support displayed the lowest
IEP of 3.2, while the GAC presented an IEP value of 4.1. Consider-
ing that IEP values only represent carbon external surface charges
whereas ZPC values represent the total net charge, their differ-
ence (ZPC-IEP) can be interpreted as the surface charge distribution
of porous carbons measurement [20–22]. From the interpretation,
values higher than zero indicates that the external surfaces are
more negatively charged than the internal surfaces, while values
close to zero correspond to a more homogeneous charge distribu-
tion in the pores. Table 1 shows that the difference in the (ZPC-IEP)
values are small and similar for the two supports, suggesting that
the charges are homogeneously distributed in the external and
internal surfaces of both supports.
low temperatures (473–573 K). In addition, both supports (GAC and
CGran) displayed a very pronounced peak between 873 and 1150 K.
Finally, the CGran support displayed a high-temperature peak at
1150–1250 K. The peaks observed during decomposition of carbon
can be assigned to various surface groups according to Figuereido
et al. [15]: lactonic (473–973 K) [16,17]; carboxylic (473–573 K)
[16]; phenolic (873–973 K) [16]; carbonyls (1073–1253 K) [16];
quinones (973–1273 K) [17]. The decomposition of these groups
(and the release of either CO2 or CO) can give an indication of the
strength of the acid sites present on the carbon surface. The con-
tributions from CO and CO2 from Fig. 2 were obtained from TPD
profiles coupled to a non-dispersive infrared device. The decon-
volution profiles obtained were similar to the previously reported
by Haydar et al. [18]. The CGran support (Fig. 2b) contains strong
acidic sites (which are associated with the release of CO2 during
decomposition of surface carboxyl groups), while the GAC support
(Fig. 2a) contains weak acidic sites probably associated to pheno-
lic and quinonic surface groups (which release CO upon thermal
decomposition).
Fig. 3 shows the FT-IR spectra of the CGran and GAC supports.
The two supports displayed bands in the region of 3500–3300 cm−1
assigned to phenolic structures, 1600–1500 cm−1 attributed to C
O
stretching vibrations of lactonic, quinonic or carboxylic groups, and
The acidity of the activated carbons estimated from potentio-
metric measurements is summarized in Table 1. It shows that the
CGran support showed higher initial electrode potential (E0) than
the GAC support. On the basis of these results, the GAC support pre-
sented very weak acid sites with E0 < 0 mV [12], while the CGran
support has strong acid sites with E0 > 100 mV [12]. These results
are in agreement with the interpretations from TPD profiles and
FT-IR spectra.
100
98
96
94
92
90
88
86
3.2. Catalysts characterization
Table 2 summarizes the textural properties for all the catalysts.
It shows that the SBET of the two catalysts with the same Re content
decreased by about 38% in comparison to the SBET of their respec-
tive supports. On the other hand, there was a 33% decrease in Vo
of the GAC support after Re(0.4) impregnation while the Vo of the
CGran support decreased by only 5% after Re(0.4) impregnation.
In contrast, a decrease in mesoporosity of about 40% was similarly
observed for the two catalysts. These results suggest that during
84
GAC
CGran
82
80
3500
3000
2500
2000
1500
1000
−
impregnation ReO4 species were deposited mainly in the meso-
-1
pores of CGran support, whereas it was deposited in the micropores
and mesopores for the GAC support. This behavior can be attributed
to the presence of oxygen surface groups. These groups confer a
Wavenumber (cm )
Fig. 3. FT-IR of the activated carbon supports.

C. Sepúlveda et al. / Applied Catalysis A: General 475 (2014) 427–437
431
25000
20000
15000
10000
5000
0
ReS2/GAC
ReS2/CGran
Re(0,8)/GAC
Re(0,7)/GAC
Re(0,4)/GAC
400
600
800
1000
1200
1400
Temperature (K)
GAC
Fig. 4. Thermal-programmed decomposition (TPD) of Re(0.4)/GAC and
Re(0.4)/CGran sulfide catalysts.
-
ReO
4
hydrophobic character to CGran support, disfavoring the migra-
tion of ReO4− species inside the support micropores. Similar results
were observed previously by Lagos et al. [9].
10
20
30
40
50
60
70
80
90
Also, Table 2 summarizes the surface area, micro and meso-
porous volumes of the Re(x)/GAC catalysts. It shows a gradual
decrease in the surface area, micro and mesoporous volumes with
increase in the Re content, consistent with the deposition of Re
species in the pores of the support.
2
Fig. 5. X-ray diffraction patterns of Re(x)/GAC catalysts.
TPD profiles of Re(0.4)/GAC and Re(0.4)/CGran sulfide catalysts
are shown in Fig. 4. The absence of peaks below 800 K, in compari-
son to the TPD profile of the original supports, suggests that the GAC
and CGran supports lost most of the lactonic and carboxylic oxy-
genated functional groups after sulfidation, in agreement with XPS
results. On the other hand, only the Re(0.4)/CGran catalyst showed
peaks above 800 K (with a lower peak area in comparison to that of
the support), suggesting a decrease in the oxygen surface functional
groups on the support after sulfidation. This behavior indicates that
the oxygen surface groups exhibited by the CGran support (from
Fig. 2) between 800 and 1300 K attributed to phenolic [16], car-
bonylic [16] and quinonic [17] oxygen groups were more resistant
to the sulfidation conditions than the GAC support.
The XRD patterns of Re(x)/GAC catalysts shown in Fig. 5 revealed
−
ReO4 species and peaks attributed to the species present in the
GAC carbon support. The result shows that at low Re content (0.4
Re atoms per nm2 of support) ReO4 diffraction peaks were not
−
observed, suggesting the presence of small Re crystallites on the
GAC support below the detection limit. On the other hand, the
catalysts with high Re contents (0.7 and 0.8 Re atoms per nm2
Table 3
XPS binding energies (eV) and surface atomic ratio of sulfide catalysts.
Catalysts
C 1s eV
O1s eV
Re 4f7/2 eV
42.1
S 2p eV
Re/C at/at
0.014
S/Re at/at
Re(0.4)/CGran
284.8 (60)
286.2 (22)
287.7 (10)
289.6 (8)
284.8 (58)
286.2 (27)
287.7 (10)
289.6 (5)
284.8 (62)
286.2 (24)
287.7 (9)
289.3 (5)
284.8 (67)
286.2 (23)
287.7 (7)
289.3 (3)
284.8 (60)
286.2 (26)
287.7 (10)
289.3 (4)
284.8 (66)
286.2 (22)
287.7 (8)
289.3 (4)
532.8 (61)
534.2 (39)
162.6 (60)
164.3 (40)
3.29 (1.97)*
Re(0.4)/GAC
Re(0.5)/GAC
Re(0.6)/GAC
Re(0.7)/GAC
Re(0.8)/GAC
532.8 (51)
534.2 (49)
42.0
42.1
42.1
42.0
42.0
162.5 (68)
163.7 (32)
0.011
0.012
0.018
0.027
0.030
2.93 (1.89)*
2.67 (1.99)*
2.70 (1.84)*
2.78 (2.17)*
2.29 (1.88)*
532.8 (55)
533.7 (45)
162.6 (69)
164.0 (31)
532.6 (51)
534.2 (49)
162.6 (68)
164.0 (32)
532.8 (50)
534.2 (50)
162.6 (78)
164.1 (22)
532.4 (58)
534.1 (42)
162.5 (82)
164.3 (18)
*
S/Re ratio calculated from S 2p (162.6 0.2 eV).

432
C. Sepúlveda et al. / Applied Catalysis A: General 475 (2014) 427–437
Re4f
ReS2(0.4)/C GAC
Re4f
ReS2(0.4)/C Gran
b)
a)
40
44
BE (eV)
48
40
44
BE (eV)
48
C-C
O1s
Re(0.4)/GAC
c)
O-C
d)
C1s
Re(0.4)/GAC
O=C
C-O
C=O
COO
280
284
288
BE (eV)
292
528
532
BE (eV)
536
540
Fig. 6. XPS spectra of Re(0.4)/GAC and Re(0.4)/CGran sulfide catalysts.
−
of support), clearly showed ReO4 diffraction peaks. These results
S 2p peaks with BEs of 162.6 0.2 and 164.0 0.3 eV: the BE of
162.6 0.2 eV was assigned to S2− type species [25], while the BE
of 164.0 0.3 eV was assigned to elemental sulfur species [26,27].
Also, Table 3 shows two values for the S/Re atomic ratio; the val-
ues in parenthesise were calculated from S2− type species, while
the other S/Re atomic ratio values were calculated taking the two S
2p peaks. The values in parenthesis were around 2.0, suggesting a
complete sulfidation. On the contrary, the other S/Re atomic ratio
values were over 2.0 indicating excess of sulfur in the surface, in
agreement with the formation of elemental sulfur species.
−
suggest the formation of ReO4 aggregates on the surface of sup-
port.
XPS results of the Re/GAC and Re/CGran sulfide catalysts are
summarized in Table 3. The XPS spectra (Fig. 6a and b) of the two
catalysts in the Re 4f region exhibited only a single doublet belong-
ing to the Re 4f7/2 of the 4f level, similar to previous work [9]. The
binding energy (BE) of the main Re 4f7/2 component for all the cat-
alysts was 42.1 0.1 eV, assigned to ReS2 according to previously
reported values [23,24]. These observations confirm that the con-
dition used for sulfidation of the supported-Re oxide species was
optimal. Also, Table 3 shows that all the catalysts displayed two
Fig. 6c shows the XPS spectra C 1s region of Re(0.4)/GAC cat-
alyst. This catalyst displayed four C 1s peaks with BEs at 284.8,

C. Sepúlveda et al. / Applied Catalysis A: General 475 (2014) 427–437
433
100
100
80
60
40
20
0
0,04
0,03
0,02
0,01
a)
HDO
Calculated
Experimental
Phenol
80
Catechol
Guaiacol
60
40
20
0
0
4000
8000
12000
16000
Time (s)
0,4
0,5
0,6
0,7
0,8
100
80
60
40
20
0
100
80
60
40
20
0
Re content (atoms nm-2)
b)
HDO
Fig. 7. Calculated and measured XPS Re/C atomic ratios vs. the nominal surface
Phenol
densities of Re(x)/GAC catalysts.
Catechol
Guaiacol
286.2, 287.7 and 289.6 0.1 eVs. The BE at 284.8 eV was assigned
to C C and/or C C bonds of aromatic and aliphatic carbon [16,28].
The BE of 286.2 0.1 eV was attributed to C–O bonds [29] in phe-
nolic or ether groups [30]. The peak of BE at 287.7 0.1 eV was
assigned to C O bonds in carbonyl groups [29], and finally the BE
at 289.6 0.1 eV is associated to C–O–OR bonds in carboxyl or ester
groups [29]. Similar BEs were displayed by all catalysts, and are
summarized in Table 3.
0
4000
8000
12000
16000
The relative intensities of the two C 1s peaks were calculated
and the values obtained are also shown in parentheses in Table 3.
It can be seen (comparing the relative proportion of these two
peaks) that all the catalysts displayed similar proportions of C 1s
peaks, suggesting that even after sulfidation the catalysts contains
similar proportions of surface oxygenated groups. However, XPS
analysis probes the chemical surface composition of the support
and is limited in its ability to uncover the properties of inside the
pores. For this reason, the disparity between results obtained from
TPD profiles and XPS analysis suggests the presence of resistive
oxygen functional groups inside the support pores after which are
undetected by XPS.
Time (s)
Fig. 8. Variation of the transformation of guaiacol and the yield of products with
time for a) Re(0.4)/GAC and b) Re(0.4)/CGran sulfide catalysts.
Table 3 also shows that Re/C atomic ratio increases with the Re
content for the GAC carbon-supported catalysts. This relationship is
depicted in Fig. 7 together with theoretical Re/C atomic ratio values
calculated according to the Kerkhof and Moulijn model [13], which
assumes a complete and homogeneous dispersion of the deposited
catalytic phase on the support surface. Fig. 7 shows that the exper-
imental Re/C atomic ratio increased linearly with Re content up to
0.6 atoms per nm2 of support. This trend is slightly below the the-
oretical Re/C atomic ratio, suggesting that the Re sulfide is highly
dispersed on the GAC surface in a monolayer up to this loading.
At higher loading, the experimental Re/C atomic ratios were sig-
nificantly higher than the calculated values. One reason for this
deviation may be the re-dispersion of rhenium species over the sur-
face of the GAC support. However, the textural properties (Table 1)
showed a pronounced decrease in support surface area after Re
impregnation with a loading of 0.8 atoms per nm−2 of support,
suggesting the formation of aggregates in the pore mouth of the
support. Therefore, the increase in experimental Re/C atomic ratios
for the catalysts with loadings above 0.6 atoms per nm2 of sup-
port cannot be attributed to re-dispersion of rhenium species but
by the formation of small three-dimensional ReS2 particles in the
pore mouth; these particles are homogeneously dispersed on the
carbon surface. The overestimated experimental Re/C atomic ratios
observed for the catalysts with higher loadings were probably due
to the formation of ReS2 aggregates over the ReS2 monolayer, lead-
ing to a lower quantity of exposed carbon surface and consequently
considerable increase in Re/C ratio. This behavior has been pre-
viously observed and reported by Lagos et al. [9]. Unfortunately,
HRTEM can not be used to estimate the size of sulfide because the
lamellar structure decomposes into spherical particles of metallic
Fig. 6d shows the XPS spectra O 1s region of Re(0.4)/GAC cata-
lyst. This catalyst displayed two O 1s peaks with BEs at 531.9 and
533.6 0.1 eV. The BE at 531.9 eV was assigned to C O bond of car-
bonyl groups [16,29], while the BE at 533.6 0.1 eV was attributed
to C–O bonds in phenolic or ether groups [29]. These results are in
agreement with XPS spectra of C 1s and TPD profile. Similar BEs
were displayed by all catalysts, and are summarized in Table 3.
The XPS Re/C surface atomic ratios of all the catalysts are shown
in Table 3. The catalysts containing 0.4 Re atoms per nm2 of support
displayed different Re/C atomic ratios: the Re(0.4)/CGran catalyst
displayed a higher value in comparison to the Re(0.4)/GAC catalyst.
These results suggest that ReS2 was better dispersed on the CGran
support compared to the GAC support. The lowest dispersion dis-
played by the Re(0.4)/GAC catalyst can be related to the support
morphology: the GAC support is mainly microporous with ReS2
deposited inside these pores, as suggested by nitrogen porosime-
try results. On the contrary, the higher dispersion displayed by
the Re(0.4)/CGran catalyst can be attributed to two factors: a high
mesoporosity of support together with the large number of sur-
face oxygen functional groups. These surface groups led to high
acid strength (higher hydrophilic character) which favors a homo-
−
geneous diffusion of the aqueous solution of ReO4 precursor into
the support during impregnation.

434
C. Sepúlveda et al. / Applied Catalysis A: General 475 (2014) 427–437
Fig. 9. Reaction scheme of conversion of guaiacol over Re/C catalysts.
rhenium under an electron bea−m [23]. However, the XRD patterns
showed the formation of ReO4 aggregates at high Re content, in
agreement with XPS measurements.
catalyst displayed a higher initial and intrinsic rates than the
Re(0.4)/CGran catalyst; however, the specific HDO rates of the two
catalysts were similar. The intrinsic rates did not correlate with
ReS2 dispersion obtained by XPS results. In fact, the Re(0.4)/CGran
catalyst was less active even though it displayed a higher exper-
imental Re/C atomic ratio. Furthermore, the higher mesopore
contribution of the Re(0.4)/CGran catalyst, as indicated by nitrogen
porosimetry results, suggests that its lower intrinsic activity can-
not be attributed to the diffusional limitations of guaiacol to active
sites inside the pores of the support. The only logical explanation
for the observed behavior is the amount of surface oxygen groups
on the carbon support. Fig. 4 showed that Re(0.4)/CGran catalyst
contains a high quantity of oxygen surface groups after sulfidation;
these surface groups could impede the adsorption and/or migration
of guaiacol to homogeneously distributed catalytic active centers
of ReS2 species present inside the CGran support, and thereby
decrease its activity. It is important to also point out that the influ-
ence on the activity caused by changes in ReS2 morphology induced
by the support, as has been observed for MoS2 catalysts [34], cannot
be discounted.
3.3. Catalytic activity
3.3.1. Effect of the support on activity and selectivity of the
catalysts
Fig. 8 shows the conversion and products yield for the HDO
of guaiacol over the Re(0.4)/GAC and Re(0.4)/CGran catalysts as
a function of time. It can be observed that the products yield
obtained were similar for the two catalysts studied, with phenol as
the principal reaction product, while catechol and deoxygenated
compounds (HDO) such as benzene, cyclohexene and cyclohex-
ane were detected in less amount. Possible secondary products
such as methylcatechols, methylphenols, toluene and methylcy-
clohexene compounds were not detected. The conversion scheme
of guaiacol to these compounds is shown in Fig. 9, in agreement
with the reaction scheme initially proposed by Laurent and Del-
mon [1,2] and Bui et al. [31]. On the basis of this scheme, guaiacol
can undergo demethylation (DME) to form catechol, which is sub-
sequently deoxygenated to form phenol. Also, guaiacol can undergo
demethoxylation (DMO) to directly form phenol without catechol
as an intermediate [32,33].
The Re(0.4)/GAC catalyst presented the highest phenol/catechol
ratio as shown in Table 4. If we consider that the same type of active
sites on sulfide catalysts are responsible for HDO of guaiacol, similar
to the phenomenon previously observed for HDS and hydrogena-
tion reactions [35], then the changes in selectivity can be attributed
to the nature of the support. The supports may modify the nature
Table 4 summarizes the total initial rate and intrinsic activity
of the catalysts for the conversion of guaiacol. The Re(0.4)/GAC

C. Sepúlveda et al. / Applied Catalysis A: General 475 (2014) 427–437
435
Table 4
Catalytic activity of the ReS2/activated carbon catalysts.
Catalysts
At. Re nm−2
Specific initial rate
Intrinsic rate (×103
R(phenol/catechol)
(10% conv.)
HDO rate
(×106 mol g−1 s−1
)
molec. at. Re−1 s−1
)
(×106 mol g−1 s−1
)
cat
cat
Re (0.4)/CGran
Re(0.4)/GAC
Re(0.5)/GAC
Re(0.6)/GAC
Re(0.7)/GAC
Re(0.8)/GAC
0.4
0.4
0.5
0.6
0.7
0.8
2.5
7.3
10
14
6.7
6.4
3.0
12
13
16
6.8
5.8
3.5
5.7
5.5
5.3
5.3
5.3
0.19
0.17
0.18
0.21
0.07
0.11
100
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
b)
a)
HDO
HDO
Phenol
Catechol
Guaiacol
80
60
40
20
Phenol
Catechol
Guaiacol
0
0
4000
8000
12000
16000
0
4000
8000
12000
16000
Time (s)
Time (s)
100
100
80
60
40
20
0
100
80
60
40
20
0
100
d)
c)
HDO
HDO
Phenol
Catechol
Guaiacol
80
60
40
20
Phenol
Catechol
Guaiacol
80
60
40
20
0
0
0
4000
8000
12000
16000
0
4000
8000
time (s)
12000
100
16000
Time (s)
100
80
60
40
20
0
e)
HDO
80
60
40
20
0
Phenol
Catechol
Guaiacol
0
4000
8000
12000
16000
Time (s)
Fig. 10. Variation of the transformation of guaiacol and the yield of products with time on Re(x)/GAC sulfide catalysts a) (0.4), b) (0.5) c) (0.6) d) (0.7), e) (0.8).

436
C. Sepúlveda et al. / Applied Catalysis A: General 475 (2014) 427–437
was attributed to the presence of oxygen surface groups resistant
18
16
14
12
10
8
to sulfidation which could limit the diffusion and/or adsorption of
guaiacol to active sites, leading to diminished catalytic activity. On
the other hand, the selectivity, expressed by phenol/catechol ratio,
was higher for the Re(0.4)/GAC catalyst than for the Re(0.4)/CGran
catalyst. The difference in selectivity was attributed to the acid
strength of the support resulting from different amount of sur-
face oxygen groups: the Re(0.4)/GAC catalyst presented low acid
strength, favoring the DMO pathway, while the Re(0.4)/CGran cat-
alysts displayed high acid strength which favors the DME pathway
as the first step of guaiacol transformation. This result confirms
that the surface chemical properties of the activated carbon influ-
enced the pathway during the conversion of guaiacol. The activity of
the Re(x)/GAC catalysts increased with Re content up to 0.6 atoms
per nm2 support. The XPS results showed that this trend can be
attributed to an increase of active sites due to an increase in Re
dispersion on the support. The decrease in activity at higher Re
loadings above 0.6 atoms of Re per nm2 of supports was attributed
to loss of active sites due to the formation of aggregates. The result
further indicates the importance of determining optimal catalytic
properties for the conversion of guaiacol. The phenol/catechol ratio
did not change with Re loading, suggesting that the nature of the
active sites was not modified by an increase in Re content at low
conversion.
6
4
2
0
0.8
0.6
Re content
0.5
0.7
0.4
Fig. 11. Intrinsic activity of Re(x)/GAC catalysts as a function of Re content.
of the catalytic sites on ReS2 responsible for C–O bond breaking
and aromatic ring hydrogenation. In fact, the difference in selec-
tivity can be attributed to the acid strength of the support as the
Re(0.4)/CGran catalyst displayed a higher acid strength than the
Re/GAC catalyst. This suggests that strong acid sites favor the for-
mation of catechol (DME pathway) while weak acid sites favor the
formation of phenol (DME route), in agreement with Bui et al. [36].
Acknowledgments
3.3.2. Effect of the Re metallic content over the activity of the
catalysts
C. Sepúlveda is indebted to CONICYT for doctoral grant at the
University of Concepción. The authors also thank CONICYT-Chile
for FONDECYT N◦ 1100512, ACT-130, PFB-27, FONDEQUIP EQM
120096 grants and Red Doctoral REDOC.CTA, MINEDUC project
UCO1202 at the University of Concepción.
Fig. 10 shows the products yield and the conversion of guaiacol
as a function of time for the Re(x)/GAC catalysts with different Re
contents. The products were phenol, catechol, lights compounds
(methanol was not detected) and HDO compounds. Phenol is the
principal product while catechol and HDO products were detected
in smaller amounts. Table 4 summarizes the catalytic activity of
the different Re(x)/GAC catalysts. The initial reaction rate, intrinsic
reaction rates and HDO rates displayed similar trends and shows
that the activity increases with the Re content up to about 0.6 atoms
of Re per nm2 of supports, followed by a gradual decrease in activ-
ity with further increase in Re loading. This trend is illustrated
in Fig. 11. This behavior can be related to the (Re/C) atomic ratio
derived from XPS result. The increase in intrinsic activity when the
Re loading is increased from 0.3 to 0.6 atoms per nm2 of support
can be attributed to increase in ReS2 dispersion. Fig. 7 suggested
that ReS2 particles were homogeneously distributed on the catalyst
with a Re loading of 0.6 atoms per nm2 of support, with corre-
sponding increase in the number of active sites and, consequently,
increase in HDO catalytic activity. On the other hand, the decrease
in catalytic activities at higher Re loadings was attributed to the for-
mation of aggregates which decrease the number of Re active sites
or their accessibility. Similar results have been previously reported
by Escalona et al. [8] and Lagos et al. [9]. Table 4 also shows that the
phenol/catechol ratio does not change with Re content, suggesting
that the changes in dispersion and/or the formation of Re sulfide
aggregates does not significantly modify the nature and number of
the active sites on ReS2 at low conversion.
References
[1] E. Laurent, B. Delmon, Appl. Catal. A 109 (1994) 97–115.
[2] E. Laurent, B. Delmon, Appl. Catal. A 109 (1994) 77–96.
[3] A. Centeno, E. Laurent, B. Delmon, J. Catal. 154 (1995) 288–298.
[4] G. De la Puente, A. Gil, J.J. Pis, P. Grange, Langmuir 15 (1999) 5800–5806.
[5] C. Sepúlveda, K. Leiva, R. García, L.R. Radovic, I.T. Ghampson, W.J. DeSisto, J.L.G.
Fierro, N. Escalona, Catal. Today 172 (2011) 232–239.
[6] A. Gutierrez, R.K. Kaila, M.L. Honkela, R. Slioor, A.O.I. Krause, Catal. Today 147
(2009) 239–246.
[7] C. Sepulveda, V. Bellière, D. Laurenti, N. Escalona, R. García, C. Geantet, M. Vrinat,
Appl. Catal. A: Gen. 393 (2011) 288–293.
[8] N. Escalona, J. Ojeda, R. Cid, G. Alves, A. López Agudo, J.L.G. Fierro, F.J. Gil Llam-
bías, Appl. Catal. A: Gen. 234 (2002) 45–54.
[9] G. Lagos, R. García, A. López Agudo, M. Yates, J.L.G. Fierro, F.J. Gil Llambías, N.
Escalona, Appl. Catal. A Gen. 358 (2009) 26–31.
[10] P.E. Ruiz, K. Leiva, R. Garcia, P. Reyes, J.L.G. Fierro, N. Escalona, Appl. Catal. A:
Gen. 384 (2010) 78–83.
[11] N. Escalona, F.J. Gil Llambias, M. Vrinat, T.S. Nguyen, D. Laurenti, A. Lopez Agudo,
Catal. Commun. 8 (2007) 285–288.
[12] R. Cid, G. Pechi, Appl. Catal. 14 (1985) 15–21.
[13] F.P.J.M. Kerkhof, J.A. Moulijn, J. Phys. Chem. 83 (1979) 1612–1619.
[14] S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and Porosity, Academic Press,
London, 1982.
[15] J. Figueiredo, M. Pereira, M. Freitas, J. Orfao, Carbon 37 (1999) 1379–1389.
[16] U. Zielke, K.J. Huttinger, W.P. Hoffman, Carbon 34 (1996) 983–998.
[17] B. Marchon, J. Carrazza, H. Heinemann, G.A. Somorjai, Carbon Interface Anal.
15 (1990) 627–634.
[18] S. Haydar, C. Moreno-Castilla, M.A. Ferro-García, F. Carrasco-Marín, J. Rivera-
Utrilla, A. Perrard, J.P. Joly, Carbon 38 (2000) 1297–1380.
[19] P.E. Fanning, M.A. Vannice, Carbon 31 (1993) 721–730.
[20] M.O. Corapcioglu, C.P. Huang, Carbon 25 (1987) 569–578.
[21] C.A. Leon y Leon, A.A. Lizzio, L.R. Radovic, On the chemical characterization of
porous carbon surfaces, in: International Carbon Conference, Paris, 1990.
[22] G. Newcombe, R. Hayes, M. Drikas, Colloids Surf. A 78 (1993) 65–71.
[23] D. Laurenti, K.T. Ninh Thi, N. Escalona, L. Massin, M. Vrinat, F.J. Gil Llambías,
Catal. Today 130 (2008) 50–55.
[24] J.P.R. Vissers, C.H. Groot, E.M. Van Oers, V.H.J. de Beer, R. Prins, Bull. Soc. Chim.
Belg. 93 (1984) 813–821.
[25] J.A. Rodriguez, S.Y. Li, J. Hrbek, H.H. Huang, G.Q. Xu, Surface Sci. 370 (1997)
85–95.
4. Conclusions
ReS2/support catalysts were prepared using activated carbons
with different textural and surface chemistry properties as sup-
ports: CGran activated carbon was mainly mesoporous, while
the GAC activated carbon was predominantly microporous. The
catalytic system studied exhibited activity for the conversion of
guaiacol. Catalytic activity was strongly affected by the nature of
the supports. The lower activity displayed by Re/CGran catalyst

C. Sepúlveda et al. / Applied Catalysis A: General 475 (2014) 427–437
437
[26] R.P.W.J. Struis, T.J. Schildhauer, I. Czekaj, M. Janousch, S.M.A. Bionaz, Chr. Lud-
wig, Appl. Catal. A: Gen. 362 (2009) 121–128.
[31] V.N. Bui, D. Laurenti, P. Afanasiev, C. Geantet, Appl. Catal. B Environ. 101 (2011)
239–245.
[27] T. Weber, J.C. Muijsers, J.H.C. van Wolput, C.P.J. Verhagen, J.W. Niemantsver-
driet, J. Phys. Chem. 100 (1996) 1411–1414.
[28] J.P.R. Vissers, S.M.A. Bouwens, V.H.J. de Beer, R. Prins, Carbon 25 (1987)
485–493.
[32] M. Ferrari, R. Maggi, B. Delmon, P. Grange, J. Catal. 198 (2001) 47–55.
[33] M. Ferrari, B. Delmon, P. Grange, Carbon 40 (2002) 497–511.
[34] H. Topsøe, B.S. Clausen, F.E. Massoth, in: J.R. Anderson, M. Boudart (Eds.), Catal-
ysis Science and Technology, Springer, Berlin, 1996.
[29] S. Biniak, G. Szymanki, J. Siedlewski, A. Swiatkowski, Carbon 35 (1997)
1799–1810.
[35] M. Vrinat, R. Bacaud, D. Laurenti, M. Cattenot, N. Escalona, S. Gamez, Catal.
Today 107–108 (2005) 570–577.
[30] S.D. Gardner, C.S.K. Singamsetty, G.L. Booth, G.R. He, C.U. Pittman, Carbon 33
(1995) 587–595.
[36] V.N. Bui, D. Laurenti, P. Delichère, C. Geantet, Appl. Catal. B Environ. 101 (2011)
246–255.