Furfuryl alcohol from furfural hydrogenation over copper supported on SBA-15 silica catalysts

Article abstract of DOI:10.1016/j.molcata.2013.11.034

Vapor phase furfural hydrogenation has been investigated over Cu supported on SBA-15 silica catalysts. These SBA-Cu catalysts, with variable Cu loadings (8, 15 and 20 wt%), have been prepared by impregnation and characterized by N₂ sorption, XRD, XPS, N₂O decomposition and TEM techniques. Compared with copper chromite, SBA-Cu catalysts showed a better catalytic performance, reaching a furfural conversion of 54 mol% and a selectivity to furfuryl alcohol of 95 mol%, after 5 h of time-on-stream at 170 °C, with the 15 wt% Cu catalyst. The studies of the used catalysts by CNH analysis and thermo-programmed oxidation (TPO) evidenced a lower amount of carbonaceous deposits on this used SBA-15Cu catalyst. Moreover, the study of the copper dispersion by XPS, before and after the catalytic test, revealed that this intermediate copper loading gives rise to the most stable copper particles. The evaluation of the effect of different reaction parameters, such as reaction temperature (170-270 °C), catalyst loadings and furfural concentration and H₂ flow, on the catalytic performance has demonstrated that higher conversion are attained at low reaction temperature, and, as expected, by using high catalyst weight and low furfural feed.

Full text of DOI:10.1016/j.molcata.2013.11.034

Journal of Molecular Catalysis A: Chemical 383–384 (2014) 106–113
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Furfuryl alcohol from furfural hydrogenation over copper supported
on SBA-15 silica catalysts
a
b
b
b
D. Vargas-Hernández , J.M. Rubio-Caballero , J. Santamaría-González , R. Moreno-Tost ,
b,∗
a
b
a
J.M. Mérida-Robles , M.A. Pérez-Cruz , A. Jiménez-López , R. Hernández-Huesca ,
b
P. Maireles-Torres
a
Centro de Investigaciones de la Facultad de Ciencias Químicas de la Benemérita Universidad Autónoma de Puebla, Bvrd. 14 Sur y Av. San Claudio Ciudad
Universitaria, 72570 Puebla, Mexico
b
Departamento de Química Inorgánica, Cristalografía y Mineralogía (Unidad Asociada al ICP-CSIC), Facultad de Ciencias, Universidad de Málaga, Campus
de Teatinos, 29071 Málaga, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 10 October 2013
Received in revised form
Vapor phase furfural hydrogenation has been investigated over Cu supported on SBA-15 silica catalysts.
These SBA-Cu catalysts, with variable Cu loadings (8, 15 and 20 wt%), have been prepared by impreg-
nation and characterized by N2 sorption, XRD, XPS, N2O decomposition and TEM techniques. Compared
with copper chromite, SBA-Cu catalysts showed a better catalytic performance, reaching a furfural con-
2
5 November 2013
Accepted 28 November 2013
Available online 6 December 2013
◦
version of 54 mol% and a selectivity to furfuryl alcohol of 95 mol%, after 5 h of time-on-stream at 170 C,
with the 15 wt% Cu catalyst. The studies of the used catalysts by CNH analysis and thermo-programmed
oxidation (TPO) evidenced a lower amount of carbonaceous deposits on this used SBA-15Cu catalyst.
Moreover, the study of the copper dispersion by XPS, before and after the catalytic test, revealed that this
intermediate copper loading gives rise to the most stable copper particles. The evaluation of the effect of
Keywords:
Cu-SBA-15
Furfural
Hydrogenation
Furfuryl alcohol
Copper
◦
different reaction parameters, such as reaction temperature (170–270 C), catalyst loadings and furfural
concentration and H2 flow, on the catalytic performance has demonstrated that higher conversion are
attained at low reaction temperature, and, as expected, by using high catalyst weight and low furfural
feed.
©
2013 Elsevier B.V. All rights reserved.
1
. Introduction
ether, which are employed as biofuels or fuel additives. Thus, for
instance, the potential of 2-methyl furan as biofuels has been
corroborated by Lange et al. [2]. Furfural is usually obtained by
acid-catalyzed dehydration of pentoses, a five carbon sugar, such
as xylose and arabinose. Furfural is also produced by fast pyrolysis
of biomass. However, the high oxygen content makes it unstable
against condensation and polymerization reactions that provide
furfural undesirable storage properties. Hydrogenation of furfural
to furfuryl alcohol is a method used to avoid these undesired reac-
tions [3].
The direct conversion of furfural into furfuryl alcohol, methyl-
furan and furan via metal-catalyzed hydrogenation, reduction and
decarbonylation reactions, is a strategic and ultimate industrial
source for the production of a wide range of derivatives. Thus, fur-
furyl amine, furoic acid, ␣-methylfurfuryl alcohol can be produced
in one step from furfural. Other important fine chemicals derived
from furfural are 2-acetylfuran, 2,5-dimethoxydihydrofuran and
Humanity is facing the need of finding new energy sources,
due to the increasing energy demand, growing environmental con-
cerns and mainly the declining of petroleum reserves. In this sense,
research attention is being paid to the exploration of new no-fossil
carbon resources, in order to take advantage of the technology
developed as well as obeying the more severe constraints about
greenhouse gas emissions. The research regarding catalytic trans-
formations of different biomass-derived compounds is currently
very active. Lignocellulose is a potential sustainable and renewable
organic carbon source. It also provides opportunities to secure the
local supply energy when it is produced locally [1,2].
Furfural is a promising platform as a building block for fuels
and chemicals. Furfural can be catalytically converted into 2-
methyl furan or 2-methyl tetrahydrofuran, used in innovative
P-series fuels, or ethyl levulinate, ␥-valerolactone, ethyl furfuryl
5
-dimethylaminomethylfurfuryl alcohol [2].
Furfuryl alcohol is an important chemical well known in poly-
∗ Corresponding author. Tel.: +34 952132021; fax: +34 952131870.
E-mail address: jmerida@uma.es (J.M. Mérida-Robles).
mer industry, mainly used for the production of thermostatic
resins, liquid resins for strengthening ceramics, as well as in the
1
381-1169/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcata.2013.11.034

D. Vargas-Hernández et al. / Journal of Molecular Catalysis A: Chemical 383–384 (2014) 106–113
107
production of various synthetic fibers, rubbers-resins and farm
chemicals. Besides, as chemical intermediate, it is employed for the
manufacture of lysine, vitamin C, lubricants, dispersing agents and
plasticizers [4].
Powder X-ray diffraction (XRD) patterns were obtained by using
a Siemens D5000 automated diffractometer, over a 2ꢀ range with
Bragg–Brentano geometry using the Cu K␣ radiation and a graphite
monochromator.
Furfuryl alcohol can be produced through hydrogenation of
furfural, either in liquid or vapor phase. The vapor phase hydro-
genation is usually preferred because it can be carried out at
atmospheric pressure. However, in both cases, the main drawback
is the low selectivity to furfuryl alcohol, where the catalyst can
play a key role. Copper chromite has been used in the furan indus-
try for the selective hydrogenation of furfural to furfuryl alcohol for
decades [5]. However, its moderate activity and high toxicity, which
causes severe environmental pollution, have increased the scien-
tific interest for developing new Cr-free catalysts which exhibit
high selectivity for furfuryl alcohol [6].
Hydrogenation of furfural in liquid phase has been widely stud-
ied, using supported metals and amorphous alloys as catalysts,
mainly copper chromite, Cu, Ni, Mo, Co, Pt, Rh, Ru and Pd [3–5,7–14]
and Ni-P, Ni-B and Ni-P-B Ultrafine Materials [15]. However, a
second metal or promoter is added sometimes for improving the
activity or/and the selectivity, by increasing the surface area or
acting as Lewis acid site to polarize the C O bond. Among them,
systems based on Ni or Co modified with Mo [16], Ni-Fe, Cu-Ca
bimetallic catalysts [11,17], promoters such as Co, Zn, Fe, Cr, Pd
and Ni in Cu-MgO [18], Ce [19], Ni-Fe-B [20], or heteropolyacids
have showed high selectivity (98%) to furfuryl alcohol and conver-
sion close to 100%. Nevertheless, the main drawback is that most
of them cannot be reused [21].
The morphology of catalysts was evaluated by Transmission
Electron Microscopy (TEM). Before TEM analysis, samples were
◦
reduced ex situ in pure H (60 ml/min) at 350 C for 2 h, and stored
2
in cyclohexane (Sigma–Aldrich, 99% purity).
X-ray photoelectron spectroscopy (XPS) studies were per-
formed with a Physical Electronics PHI 5700 spectrometer
equipped with a hemispherical electron analyzer (model 80-365B)
and a Mg K␣ (1253.6 eV) X-ray source. High-resolution spectra
◦
were recorded at 45 take-off angle by a concentric hemispherical
analyzer operating in the constant pass energy mode at 29.35 eV,
using a 720 ␮m diameter analysis area. Charge referencing was
done against adventitious carbon (C 1s at 284.8 eV). The pressure in
−
6
the analysis chamber was kept lower than 5 × 10 Pa. PHI ACCESS
ESCA-V6.0 F software package was used for data acquisition and
analysis. A Shirley-type background was subtracted from the sig-
nals. Recorded spectra were always fitted using Gauss–Lorentz
curves in order to determinate more accurately the binding energy
of the different element core levels. The samples underwent the
same aforementioned ex situ treatment before XPS analysis.
◦
N adsorption–desorption isotherms at −196 C of calcined cat-
2
alysts were obtained using an ASAP 2020 model of gas adsorption
analyzer from Micromeritics, Inc. Prior N2 adsorption, samples
◦
−2
were evacuated overnight at 200 C and 1 × 10 Pa. Pore size dis-
tributions were calculated with the BJH method [23].
While hydrogenation of furfural in liquid phase requires high
pressures, it is carried out under mild conditions in gas phase,
reducing costs. In this sense, some works concerning hydrogena-
tion of furfural to furfuryl alcohol in gas phase have been recently
reported. Although different types of supports, such as MgO and
The reducibility of the calcined samples was determined by H2
temperature-programmed reduction (H -TPR). For these measure-
2
ments, 80 mg of sample was placed in a quartz reactor and heated at
◦
100 C under a He flow of 35 ml/min, and held at this temperature
◦
for 1 h. The reactor was then cooled down to 50 C and the sam-
SiO , have been used in order to increase the surface area and
ple exposed to a stream of 47 ml/min of 10% H /Ar. Subsequently,
2
2
◦
◦
to enhance the metal-support interaction, Cu has been the most
employed metal, due to its high activity and selectivity toward fur-
furyl alcohol [3,4,7,9,12,14,17,18]. However, an active catalyst to
substitute copper chromite catalyst has not been developed yet,
since most of them are unsuitable for industrial applications owing
to severe deactivation phenomena.
the sample was heated up to 700 C at a heating rate of 10 C/min.
The amount of hydrogen consumed as a function of temperature
was monitored on-line on a TCD detector. The water formed dur-
◦
ing reduction was collected with a cryogenic trap at −85 C before
chromatographic analysis.
Copper surface area and dispersion were calculated by N O
2
The aim of this work is to investigate the furfural catalytic hydro-
genation in vapor phase over mesoporous silica-supported Cu
catalysts. The catalytic behavior of these environmentally friendly
materials was compared with that of a copper chromite. The effect
of different experimental parameters, such as catalyst loading,
reaction temperature, weight of catalyst and furfural feed, on the
catalytic performance was studied.
decomposition method. This method is based on the formation of a
0
monolayer of Cu O by oxidation of Cu with a N O flow, according
2
2
0
to the reaction: 2Cu + N O → Cu O + N . Before analysis, the CuO
2
2
2
◦
phase is reduced with a flow mixture of 10 vol.% H /Ar at 5 C/min
2
◦
at 350 C. Then, the catalyst is purged under He and cooled down
to 60 C. The oxidation of Cu to Cu is carried out by chemisorp-
tion of N O (5 vol.% N O/He) at 60 C during 1 h. Then, the catalyst
◦
0
+
◦
2
2
was again purged with an Ar flow and cooled to room tempera-
ture. After this, a temperature programmed reduction was carried
◦
2
. Experimental
out similarly to TPR, raising the temperature up to 350 C on the
freshly oxidized Cu O surface in order to reduce Cu O to Cu.
2
2
2
.1. Catalyst synthesis and characterization
Temperature-programmed oxidation analyses (TPO) were per-
formed in a METTLER TOLEDO TGA/DSC 1 apparatus equipped with
a HT1600 furnace and a MX5 balance with a DSC HSS2 Pt-Rh sen-
The synthesis of SBA-15 silica was carried out as reported by
◦
◦
Zhao et al. [22]. The SBA-Cu catalysts were prepared by incipient
sor. Temperature was varied from 30 to 400 C at a rate of 5 C/min,
wetness impregnation with aqueous solutions of Cu(NO ) ·3H O
under an O flow of 50 ml/min.
3
2
2
2
(
Aldrich, >99% purity). After impregnation, catalysts were dried
overnight at room temperature, and finally calcined for 6 h at
2.2. Catalytic activity
◦
◦
4
00 C, with a heating rate of 1 C/min. The metal loading ranges
between 8 and 20 wt% of Cu. The catalysts were labeled as SBA-
xCu, where x is the wt% of Cu. For comparison, a commercial copper
chromite (Cu-Cr) was also studied (CuCr O ·CuO, Aldrich).
The vapor-phase reduction of furfural was conducted in a
tubular quartz reactor. The pelletized catalyst (325–400 ␮m)
was placed at the center of the reactor tube between two lay-
ꢀ
ꢀ
¼
2
4
Elemental analysis was performed on a PERKIN-ELMER 2400
CHN with a LECO VTF900 pyrolysis oven. Cu contents have been
determined by atomic adsorption spectroscopy (AAS) by using a
Varian SPECTRAA 50.
ers of glass beads and quartz wool. Catalysts were reduced in situ
◦
under a H2 flow (60 ml/min, Airgas, 99.99%) for 2 h at 350 C,
prior to the catalytic test. After reduction, the catalysts were
cooled down to the selected reaction temperature under a H flow
2

1
08
D. Vargas-Hernández et al. / Journal of Molecular Catalysis A: Chemical 383–384 (2014) 106–113
Fig. 1. N2 adsorption–desorption isotherms: (a) SBA-15, (b) SBA-8Cu, (c) SBA-15Cu
and (d) SBA-20Cu catalysts.
(
10 ml/min). When the temperature was reached, a flow of fur-
fural (Sigma–Aldrich) solution in cyclopentyl methyl ether (CPME)
Sigma–Aldrich) (5 vol.%) was continuously injected by using a
(
HPLC pump. CPME is an environmentally friendly solvent that has
been used in different organic reactions, thus, for instance, demon-
strating to be a green co-solvent for the selective dehydration of
lignocellulosic pentoses to furfural [24].
The reaction products were analyzed by gas chromatography
(
Agilent model 14A) with a Flame Ionization Detector. The product
Fig. 2. Low angle XRD patterns of (a) SBA-15, (b) SBA-8Cu, (c) SBA-15Cu, (d) SBA-
20Cu catalysts.
yield and selectivity were calculated and defined as follows:
mol of furfural converted
Conversion (%) =
Selectivity (%) =
× 100
× 100
mol of furfural fed
the (1 0 0) reflection of the SBA-15 silica, confirming that the meso-
porous structure is preserved after copper incorporation. However,
the presence of copper on the support gives rise to a decrease in
the intensity of this low angle peak, which can be explained by the
scattering effect of the metallic copper particles.
mol of the product
mol of furfural converted
0
3
. Results and discussion
.1. Characterization of copper catalysts
The Cu contents of the different catalysts, as determined by AAS,
The presence of metallic copper (Cu ) on the surface of catalysts,
fresh (reduced) and spent (after reaction), can be deduced from XRD
(Fig. 3). Thus, the characteristic diffraction lines of Cu (fcc struc-
ture) are clearly visible at 2ꢀ ( ) = 43.4 and 50.5 (JCPDS 04-0836),
excepting for the catalyst with the smallest amount of Cu (SBA-
8Cu), and their intensities, as expected, increase with the copper
loading. However, after reaction, copper chromite also shows the
0
3
◦
are displayed in Table 1. Thus, the Cu content of the synthesized
SBA-xCu catalysts varies between 7.8 and 20.2 wt%, while for the
Cu-Cr sample was much higher (39.6 wt%).
◦
presence of weak peaks at 2ꢀ ( ) = 31.4, 36.5 and 62.4, which point to
N2 adsorption–desorption isotherms for SBA-15 and calcined
SBA-xCu are presented in Fig. 1. All of them exhibited a Type-IV
isotherm according to the IUPAC classification, with a H1 hystere-
sis loop. The shape of the isotherm of pure SBA-15 was preserved
even for the catalyst with the highest copper content, which means
that, after the impregnation and subsequent calcination of copper
precursors, the mesoporous nature of the solid was maintained.
The evaluation of the textural properties of calcined samples from
the formation of CuCrO crystallites (JCPDS-39-0247). These results
2
suggest that the metallic copper in copper chromite is oxidized
+
to Cu , which reacts with Cr O to form CuCrO . This behavior
2
3
2
of copper chromite, after reduction, has been already observed by
Deutsch and Shanks in the study of the active species derived from
Cu-chromite in C O hydrogenolysis of furfuryl alcohol [25].
It is well known that reducibility of copper species depends on
the preparation method, the nature of the support and precursor,
as well as the experimental conditions used for the reduction step.
Fig. 4 shows the TPR profiles of the calcined SBA-xCu catalysts and
Cu-chromite. The quantification of H2 consumption confirms the
◦
nitrogen adsorption–desorption isotherms at −196 C reveals that
the BET surface area decreases progressively with the loading of
copper, with a drastic reduction for the material with the highest
Cu content (SBA-20Cu) (Table 1). However, excepting for the latter
sample, the average pore diameters barely change in comparison
with the mesoporous silica, used as support, thus pointing to the
presence of copper oxide particles blocking partially the entrance
to the mesoporous framework.
◦
complete reduction of CuO to Cu(0) at the temperature of 350 C.
The existence of several defined ranges of hydrogen consumption
in TPR profiles points to the presence of different Cu2 species in
+
calcined precursors. Two broad H consumption bands of different
2
intensity can be observed in the H -TPR spectra of the supported
2
2+
The low-angle powder XRD patterns of the SBA-15 support and
SBA-xCu catalysts (Fig. 2) exhibit a well-resolved diffraction peak,
which can be indexed in a hexagonal unit cell as corresponding to
copper catalysts, which can be assigned to the reduction of Cu
to Cu . According to the literature [26], the low temperature band
could be associated to the presence of finely dispersed CuO particles
0

D. Vargas-Hernández et al. / Journal of Molecular Catalysis A: Chemical 383–384 (2014) 106–113
109
Table 1
Physicochemical characterization of SBA-xCu and Cu-Cr catalysts.
SBET (m2
g
−1
)
dp (nm)
Vp (cm3
g
−1
)
DM (%)
SM(m2
g )
−1
dM (nm)
Catalyst
Cu loading (wt%)
SBA-15
–
7.8
14.1
20.2
39.6
581
3.6
3.9
4.0
4.3
18.0
0.465
0.326
0.267
0.202
0.217
–
–
–
SBA-8Cu
SBA-15Cu
SBA-20Cu
Cu-Cr
369
307
179
46
58.0
38.5
9.7
389.3
260.4
65.8
1.7
2.6
10.3
4.1
24.5
165.8
and the second one to larger particles of bulk CuO. These findings
are in agreement with the XRD data of reduced catalysts, where the
presence of diffraction peaks associated to metallic copper is not
detected in the SBA-8Cu catalyst. With the increase of Cu loading,
the relative intensity of the hydrogen consumption peaks change
gradually, by increasing the second one which could be explained
by the formation of large copper oxide particles. This agrees with
previous studies, where the presence of two reduction peaks in
Therefore, it could be inferred from the TPR profiles that complete
reduction of copper in this series of catalysts can be expected after
a reduction treatment at 350 C.
◦
Regarding the copper dispersion, the results of the titration with
◦
N O at 60 C are presented in Table 1. The SBA-xCu catalysts showed
2
high dispersion (DM) and metal area (SM) values that decreased
with Cu content. It was possible to estimate the average particle
size (dM) assuming cubic metal particles and using a surface den-
15
−2
the H -TPR curves of Cu/SiO2 and co-precipitated Cu-MgO cata-
sity of Cu atoms of 1.08 × 10 at cm . Thus, the estimated values
range between 1.7 and 10.3 nm, indicating that small metal copper
particles are formed by reduction of the samples with the highly
dispersed CuO phases present on the SBA surface, as previously
showed by XRD patterns. As expected, the average particle sizes
increase with the copper loading.
2
lysts have been shown [3,10]. On the other hand, copper(II) in
◦
Cu-chromite is fully reduced at temperatures lower than 250 C.
Catalysts have been characterized by X-ray photoelectron spec-
troscopy to obtain more information about the Cu oxidation state
and the chemical composition of the catalyst surface. The binding
energy values of Cu 2p3/2, O 1s and Si 2p along with the Si/Cu atomic
ratios, of both fresh and spent catalysts, are presented in Table 2.
The Si 2p binding energy (BE) value, in all the SBA-xCu catalysts, is
very similar, and the average value of 103.5 eV is typical of silica.
The O 1s signal is symmetrical and appears at 533 eV in the SBA-
xCu catalysts, lowering until 530 eV in copper chromite. These BE
values for O 1s are characteristics of silica and metal oxide, respec-
tively. The Cu 2p3/2 signal at 932–933 eV is due to the presence
of metallic copper and/or Cu(I) species, whereas surface copper(II)
species are not detected, as confirmed by the absence of XPS sig-
nals at 934–935 eV and the typical shake-up satellite at 940–945 eV,
related with unoccupied 3d states (Fig. 5). However, the presence
of Cu(II) is observed in the reduced copper chromite. For that rea-
son, the Cu LMM kinetic energy of the Auger spectra has been used
to distinguish between Cu(I) and Cu(0) (Fig. 6). In all cases, a broad
band is observed with maxima at ca. 919 and 917 eV, which can be
assigned to the contributions of Cu(0) and Cu(I), respectively [27].
Fig. 3. XRD patterns for fresh (A) and used (B) catalysts: (a) SBA-15, (b) SBA-8Cu,
0
(
c) SBA-15Cu (d) SBA-20Cu and (e) Cu-Cr. The following phases are denoted: Cu (*)
Fig. 4. Temperature programmed reduction profiles (TPR) of (a) SBA-8Cu, (b) SBA-
15Cu, (c) SBA-20Cu and (d) and Cu-Cr catalysts.
◦
and CuCrO2( ).

1
10
D. Vargas-Hernández et al. / Journal of Molecular Catalysis A: Chemical 383–384 (2014) 106–113
Table 2
XPS data of SBA-xCu and Cu-Cr catalysts.
Catalyst
Binding energy (eV)
Atomic ratio
Si 2p
O 1s
Cu 2p3/2
Si(Cr)/Cu (fresh)
Si(Cr)/Cu (spent)
Si(Cr)/Cu (theor)
SBA-8Cu
SBA-15Cu
SBA-20Cu
Cu-Cr
103.6
103.4
103.5
–
533.1
532.9
533.0
530.2
933.2
932.2
932.0
932.5
53.3
30.4
13.1
2.66
32.1
29.4
20.8
2.79
12.5
6.4
4.2
1.0
In general, the Cu LMM bands shift to lower kinetic energy values
after the catalytic reaction, thus evidencing the partial oxidation
of Cu(0) to Cu(I), which has been proposed by some authors as the
active sites for hydrogenation of furfural to furfuryl alcohol [12,13].
Concerning the surface Si/Cu atomic ratios, their values are
always higher than the theoretical bulk ratios determined by ICP-
AES, and they decrease, as expected, with the copper loading.
This difference can be explained by considering not only the pres-
ence of copper particles into the mesoporous channels but also
the existence of large copper particles on the external surface of
the mesoporous support, as inferred from the presence of narrow
diffraction peaks in the corresponding XRD patterns.
copper chromite becomes almost inactive, whereas the supported
catalyst with 15 wt% of copper loading still maintains a conversion
higher than 50%. Hydrogenation of furfural can lead to different
products, including furfuryl alcohol, 2-methylfuran, tetrahydrofur-
furyl alcohol and other ring-opening products, depending on the
catalysts employed and the reaction conditions. In this work, fur-
furyl alcohol (FOL) was obtained as the main product along with
2-methylfuran (MF) (Fig. 9). These results are in agreement with
previous reports of furfural conversion over Cu-based catalysts,
which show that furfuryl alcohol is the main product [4,6,11,14,18].
Sitthisa et al. [3] found that the conversion of furfural on Cu cata-
lyst involves the interaction of the surface preferentially with the
carbonyl group, rather than the furanic ring. Thus, DFT calculations
have given evidence for a strong repulsion between the surface Cu
(1 1 1) and the furan ring, presumably due to the overlap of the 3d
band of Cu atoms and the anti-bonding orbital of the aromatic furan
ring [3,28].
The TEM studies corroborate the results obtained by XRD
and H -TPR analysis. The TEM micrographs of the SBA-8Cu and
2
chromite catalysts, fresh and spent, are presented in Fig. 7. It can
be observed that the metallic copper clusters are well dispersed on
◦
the silica support after reduction at 350 C. As the copper loading
increases, a more heterogeneous distribution of particles sizes is
observed; thus, some metal particles of 35 nm are detected on the
catalysts with the highest copper loading. In the case of the Cu-Cr
catalyst, agglomerates of copper particles can be distinguished in
the corresponding micrographs.
All the catalysts exhibit high selectivity toward FOL, which is the
highest for a Cu content of 15 wt%. Nagaraja et al. have proposed
that the presence of more Cu , or less CuO species, gives rise to
0
Fresh catalysts
3.2. Catalytic activity studies
SBA-8Cu
SBA-15Cu
SBA-20Cu
Cu-Cr
This series of copper-based catalysts has been tested in the vapor
phase hydrogenation of furfural (FAL), at atmospheric pressure.
Previously, the stability of the cyclopentyl methyl ether, used as sol-
◦
vent of furfural, was studied by feeding it into the reactor at 170 C,
in the presence of the SBA-15Cu catalyst. The results indicate that
the solvent is stable, since it was fully recovered after 5 h of reac-
tion time, without the presence of any new chemical compound.
The results of catalytic activity as a function of time on stream
◦
(
TOS), at 170 C, reveal a strong deactivation (Fig. 8), which is more
pronounced in the case of copper chromite. Thus, after 5 h of TOS,
9
05
910
915
920
925
930
Cu LMM (eV)
SBA-8Cu
SBA-15Cu
SBA-20Cu
Cu-Cr (:5)
SBA-8Cu(sp)
SBA-15Cu(sp)
SBA-20Cu(sp)
Cu-Cr(sp)
Spent catalysts
9
05
910
915
920
925
930
9
70
960
950
940
BE (eV)
930
920
Cu LMM (eV)
Fig. 5. XPS spectra in the Cu 2p region of reduced copper catalysts.
Fig. 6. Cu LMM Auger spectra of fresh and spent (sp) copper catalysts.

D. Vargas-Hernández et al. / Journal of Molecular Catalysis A: Chemical 383–384 (2014) 106–113
111
Fig. 7. TEM micrographs of the SBA-8Cu and Cr-Cu catalysts, before and after catalytic reaction.
a high catalytic activity toward the selective formation of FOL by
hydrogenation of FAL [14]. A simple bimolecular surface reaction
The deactivation process could be explained by the active copper
site coverage by coke or adsorbed reactants/products, as revealed
the high carbon percentages found in the spent catalysts by CHN
analysis (Table 6), rather than the copper particle sintering or loss
+
as a rate-determining step by correlating Cu sites with activity
+
0
and they proposed that the presence of both Cu and Cu species
are required for optimum performance [4,13]. However, lack of
required number of smaller Cu particles could make the SBA-20Cu
catalyst less active in furfural hydrogenation, where the agglomer-
ation of copper particles, as inferred from the surface Si/Cu molar
ratio of the spent catalyst, could be responsible for its low activity
compared to that of SBA-15Cu.
+
of the active Cu species by reduction. The carbon deposition is the
lowest for the SBA-15Cu material, which is that with the highest
conversion values with TOS (Fig. 8). By comparing the Cu LMM spec-
tra of fresh and used catalysts, it can be inferred that the amount
of Cu+ is more important after reaction, and the XPS results con-
firm that the existence of sintering could only be suggested for the
◦
Fig. 8. Conversion of furfural over copper catalysts, at 170 C, as a function of
Fig. 9. Selectivity toward furfuryl alcohol and 2-methyl furan (MF) over copper cat-
alysts, at 170 C, as a function of time on stream. Reaction conditions: mcat = 150 mg,
◦
time on stream. Reaction conditions: mcat = 150 mg, H2 flow = 10 ml/min, feed
flow = 2.3 mmol/h.
H2 flow = 10 ml/min, feed flow = 2.3 mmol/h.

1
12
D. Vargas-Hernández et al. / Journal of Molecular Catalysis A: Chemical 383–384 (2014) 106–113
Table 3
Influence of the reaction temperature on the furfural hydrogenation over the SBA-
5Cu catalyst (experimental conditions: mcat = 150 mg, H2 flow = 10 ml/min, feed
Table 5
Influence of the furfural feed on FAL hydrogenation over SBA-15Cu (experimental
conditions: m = 150 mg, T = 170 C, H2 flow = 10 ml/min, t = 5 h).
◦
1
flow = 2.3 mmol/h, t = 1 h).
Furfural feed (mmol/h)
Conversion (%)
Selectivity (%)
Yield (%)
FOL
◦
T ( C)
Conversion (%)
Selectivity (%)
Yield (%)
FOL
FOL
MF
MF
FOL
MF
MF
6.4
12.6
21.9
34.0
37.2
37.5
1.6
58.3
54.1
31.9
84.8
95.8
91.0
15.2
4.2
9.0
49.5
51.8
29.0
8.9
2.3
2.9
170
190
210
230
250
270
91.5
84.4
81.2
82.7
77.8
66.5
93.1
85.1
73.0
58.9
52.1
43.5
6.9
14.9
27.0
41.1
47.9
56.5
85.4
72.0
59.4
48.8
40.6
28.9
2.3
3.4
Table 6
Data of TOF (reaction conditions: mcat = 150 mg, H2 flow = 10 ml/min, feed
flow = 2.3 mmol/h, t = 5 h) and C analysis for the used SBA-xCu and Cu-Cr catalysts.
Catalyst
C (%)
TOFt = 0 (FAL molecules
converted s Cu(0) )
Table 4
−
1
−1
Influence of the catalyst weight on furfural hydrogenation over the SBA-
◦
−3
1
5Cu catalyst (experimental conditions: T = 170 C, H2 flow = 10 ml/min, feed
SBA-8Cu
SBA-15Cu
SBA-20Cu
Cu-Cr
10.3
6.3
9.7
2.66 × 10
flow = 2.3 mmol/h, t = 5 h).
2.03 × 10−3
−
−3
3
1.32 × 10
0.72 × 10
Catalyst weight (mg)
Conversion (%)
Selectivity (%)
Yield (%)
FOL
4.8
FOL
MF
MF
1
1
2
00
50
00
33.9
54.1
69.4
82.4
95.8
89.5
17.6
4.2
10.5
27.9
51.8
62.0
5.9
2.3
7.3
Taking into account the differences in Cu dispersion and load-
ing among the catalysts, it has been estimated initial TOF values
in order to compare the intrinsic activity of the Cu metal sites.
−
1
Table 6 shows the turnover frequency (TOF, s ) values, defined as
0
catalyst with the highest copper loading (SBA-20Cu). It is note-
worthy that the best catalytic performance is achieved for an
intermediate copper loading, which is the catalyst that main-
tained the atomic Si/Cu ratio unchanged. A similar evolution of the
catalytic activity as a function of the Cu loading in the FAL hydro-
genation reaction has been observed by using Cu-MgO catalysts
4].
In the case of copper chromite, although the presence of car-
bonaceous deposits is detected, its strong deactivation could be
molecules of FAL converted per second and Cu site, calculated by
extrapolation at zero time. From this study it is clear that TOF value
is higher in SBA-xCu catalysts than in the Cu-Cr one. The highest TOF
value was obtained for SBA-8Cu material, which showed a strong
deactivation that can be associated to the deposition of coke and/or
adsorbed reactants/products, as inferred from the highest C content
found by CNH analysis in the spent catalyst. The SBA-20Cu catalyst
exhibited the lowest TOF value of this series of supported copper
[
0
catalysts, which could be due to a partial sintering of Cu particles
+
also due to the formation of a CuCrO2 phase from Cr O3 and Cu ,
as evidenced the increase of the Si/Cu atomic ratio, obtained by XPS
analysis, of the spent catalyst. Thus, the SBA-15Cu catalyst seems
to be that with the appropriated loading and dispersion of Cu0 to
produce the highest conversion and FOL selectivity with TOS, with
less deactivation, neither by C deposition nor metal sintering.
Finally, by feeding low FAL flows, a higher initial conversion is
attained and the catalyst is more resistant to deactivation (Table 5).
However, it is observed a decrease in FOL selectivity, favoring the
formation of MF, thus resulting that a high feed of furfural inhibits
MF production.
2
as evidenced the XRD pattern of the spent copper chromite (Fig. 3).
This fact has been previously observed by Liu et al. in their work
devoted to the study of the deactivation mechanism of this copper
chromite for the same catalytic reaction at 200 C, where the cov-
erage of Cu sites by this CuCrO phase is proposed as an additional
◦
2
cause of catalyst deactivation [29].
On the other hand, the study of the influence of the reaction
◦
temperature, in the range 170–270 C (Table 3), on conversion and
product distribution over SBA-15Cu reveals a clear decrease in con-
version as temperature rises, whereas MF production increases in
detriment of FOL. Liu et al. [29] have reported the same behavior for
copper chromite catalysts and the conversion decrease is attributed
to the polymerization and adsorption of FOL molecules on the cat-
alyst surface, which is favored at high temperatures. Therefore, it
is apparent that furfural is first converted into FOL, which subse-
quent deoxygenation leads to MF, a reaction pathway previously
proposed by Sitthisa et al. [3]:
4. Conclusions
A series of supported copper catalysts has been prepared by
impregnation and subsequent calcination of mesoporous SBA-15
silica with different loading of copper nitrate. The study of the
◦
Therefore, the lowest temperature, 170 C, was chosen to opti-
mize other reaction parameters. The study of the catalytic behavior
at lower temperatures was not possible due to the boiling point of
furfural (161.7 C).
Regarding the influence of the catalyst weight, the FAL conver-
sion is ameliorated by increasing the amount of catalyst (Table 4).
However, although conversion is improved, a higher amount of
catalyst does not allow reaching much better FOL yield.
furfural hydrogenation over SBA-xCu and Cu-Cr catalysts indicated
0
+
that all of them are active, and that the presence of Cu -Cu species
on the catalyst surface is responsible for the high activity and
selectivity toward furfuryl alcohol. The SBA-15Cu catalyst shows
◦
the highest activity and selectivity toward the desired product at
◦
170 C, when a H flow 10 ml/min, 150 mg of catalyst and a furfural
2
feed flow of 0.1875 ml/h were employed.

D. Vargas-Hernández et al. / Journal of Molecular Catalysis A: Chemical 383–384 (2014) 106–113
113
All catalysts undergo deactivation with time on stream, being
[7] S. Sitthisa, D.E. Resasco, Catal. Lett. 141 (2011) 784–791.
[
8] V. Vetere, A.B. Merlo, J.F. Ruggera, M.L. Casella, J. Braz. Chem. Soc 21 (2010)
more drastic for copper chromite. In the case of SBA-xCu, the stud-
ies of the used catalysts by CNH analysis and thermo-programmed
oxidation (TPO) evidenced the presence of carbonaceous deposits,
which are lower for the used SBA-15Cu catalyst. Moreover, the
study of the copper dispersion by XPS, before and after the catalytic
test, revealed that this intermediate copper loading gives rise to the
most stable copper particles.
914–920.
[
9] W. Huang, H. Li, B. Zhu, Y. Feng, S. Wang, S. Zhang, Ultrason. Sonochem. 14
(2007) 67–74.
[
[
10] B.M. Nagaraja, V. Siva Kumar, V. Shasikala, A.H. Padmasri, B. Sreedhar, B. David
Raju, K.S. Rama Rao, Catal. Commun. 4 (2003) 287–293.
11] X. Chen, H. Li, H. Luo, M. Qiao, Appl. Catal. A 233 (2002) 13–20.
[12] R.S. Rao, R.T. Baker, M.A. Vannice, Catal. Lett. 60 (1999) 51–57.
[13] R.S. Rao, R.T. Baker, M.A. Vannice, J. Catal. 171 (1997) 406–419.
14] B.M. Nagaraja, A.H. Padmasri, P. Seetharamulu, K. Hari Prasad Reddy, B. David
Raju, K.S. Rama Rao, J. Mol. Catal. A 278 (2007) 29–37.
[15] S.P. Lee, Y.W. Chen, Ind. Eng. Chem. Res. 38 (1999) 2548–2556.
16] S. Sitthisa, W. An, D.E. Resasco, J. Catal. 284 (2011) 90–101.
17] J. Wu, Y. Shen, C. Liu, H. Wang, C. Geng, Z. Zhang, Catal. Commun. 6 (2005)
[
Acknowledgments
[
[
The authors are grateful to financial support from Spanish Min-
istry of Economy and Competitiveness (CTQ2012-38204-C03-02),
FEDER funds and Junta de Andalucía (P09-FQM-5070). RMT for the
financial support under the Program Ramón y Cajal (RYC-2008-
3387) and Consejo Nacional de Ciencia y Tecnología (CONACyT,
México) for financial support via scholarships (219821).
633–637.
[18] B.M. Nagaraja, A.H. Padmasri, B.D. Raju, K.S. Rama Rao, Int. J. Hydrogen Energy
36 (2011) 3417–3425.
19] H. Li, S. Zhang, H. Luo, Mater. Lett. 58 (2004) 2741–2746.
[
[
20] H. Li, H. Luo, S. Zhang, W. Dai, M. Qiao, J. Mol. Catal. A: Chem. 203 (2003)
0
2
67–275.
[21] A.B. Merlo, V. Vetere, J.F. Ruggera, M.L. Casella, Catal. Commun. 10 (2009)
665–1669.
1
[
22] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky,
Science 279 (1998) 548–552.
References
[
23] E.P. Barrett, L.J. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373–380.
[
[
1] L. Faba, E. Díaz, S. Ordó n˜ ez, Catal. Today 164 (2011) 451–456.
2] J.P. Lange, E. van der Heide, J. van der Buijtenen, R. Price, ChemSusChem 5 (2012)
[24] M.J. Campos Molina, R. Mariscal, M. Ojeda, M. López-Granados, Bioresour. Tech-
nol. 126 (2012) 321–327.
1
50–166.
3] S. Sitthisa, T. Sooknoi, Y. Ma, P.B. Balbuena, D.E. Resasco, J. Catal. 277 (2011)
–13.
4] B.M. Nagaraja, A.H. Padmasri, B. David Raju, K.S. Rama Rao, J. Mol. Catal. A 265
2007) 90–97.
5] J. Kijenski, P. Winiarek, T. Paryjczak, A. Lewicki, A. Mikołajska, Appl. Catal. A
33 (2002) 171–182.
6] L. Baijun, L. Lianhai, W.B. Bingchun, C. Tianxi, K. Iwatani, Appl. Catal. A 171
1998) 117–122.
[25] K.L. Deutsch, B.H. Shanks, J. Catal. 285 (2012) 235–241.
[
[
[
[
[26] G.E. Walrafen, S.R. Samanta, J. Chem. Phys. 69 (1978) 493–495.
[27] W.L. Dai, Q. Sun, J.F. Deng, D. Wu, Y.H. Sun, Appl. Surf. Sci. 177 (2001) 172–179.
[28] D.E. Resasco, S. Sitthisa, J. Faria, T. Prasomsri, M.P. Ruiz, Furfurals as chemical
platform for biofuels production, in: D. Kubi cˇ ka, I. Kubi cˇ ková (Eds.), Chapter in
Heterogeneous Catalysis in Biomass to Chemicals and Fuels, 2012, pp. 155–188.
[29] D. Liu, D. Zemlyanov, T. Wu, R.J. Lobo-Lapidus, J.A. Dumesic, J.T. Miller, C.L.
Marshall, J. Catal. 299 (2013) 336–345.
1
(
2
(