Oxidative esterification of renewable furfural on gold-based catalysts: Which is the best support?

Article abstract of DOI:10.1016/j.jcat.2013.10.005

Gold-based catalysts over different supports were investigated in the oxidative esterification of furfural by employing an efficient and sustainable process. The catalytic performances follow the trend: Zirconia-Au > Ceria-Au a‰

Full text of DOI:10.1016/j.jcat.2013.10.005

Journal of Catalysis 309 (2014) 241–247
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Oxidative esterification of renewable furfural on gold-based catalysts:
Which is the best support?
a
a,
⇑
a
b
b
Federica Menegazzo , Michela Signoretto , Francesco Pinna , Maela Manzoli , Valentina Aina ,
b
b
Giuseppina Cerrato , Flora Boccuzzi
a
Dept. of Molecular Sciences and Nanosystems, Ca’ Foscari University Venice and Consortium INSTM, RU-Venice, Calle Larga S. Marta 2137, 30123 Venice, Italy
Dept. of Chemistry & NIS Centre of Excellence, University of Turin, Via P. Giuria 7, 10125 Turin, Italy
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Gold-based catalysts over different supports were investigated in the oxidative esterification of furfural
by employing an efficient and sustainable process. The catalytic performances follow the trend: Zirconia–
Au > Ceria–Au ꢀ Titania–Au. Zirconia came out to be the best support option to promote activity, selec-
tivity and also stability. The chemical and morphological properties observed for zirconia-supported
sample seem to fulfill a good compromise between high gold dispersion and the presence of suitable
acid–base properties, for good selectivity. Moreover, stability and recycling of the catalysts were also
investigated.
Received 29 August 2013
Revised 2 October 2013
Accepted 8 October 2013
Available online 7 November 2013
Keywords:
Gold catalyst
Furfural
Ó 2013 Elsevier Inc. All rights reserved.
Oxidation
Biomass
Effect of the support
1
. Introduction
With diminishing fossil resources, developing new technologies
ran (via hydrogenation of furan). However, additional transforma-
tions of furfural are highly desired [4]. There already have been
efforts to modify furfural-based derivatives by gold catalysis [5–
8]. The synthesis of alkyl furoates can open very interesting per-
spectives for the use of xyloses because they find applications as
flavor and fragrance component in the fine chemical industry.
Up to now, only a few studies have investigated the heteroge-
neous catalysis in oxidation of 2-FA [9,10]. Traditionally, the ester
is prepared by oxidizing furfural with potassium permanganate,
preferably using acetone as solvent, and reacting the furoic acid
so formed with methyl or ethyl alcohol, in the presence of sulfuric
acid. The use of these substances has a substantial negative impact
on the environment. It has been shown [9] that furfural can be con-
verted to methyl furoate under mild conditions by an oxidative
to utilize versatile and renewable biomass as the alternative feed-
stock of energy and chemical sources has been attracting more
attention than ever. In particular, the upgrading of lignocellulosic
biomass wastes into fuels and higher added-value chemicals is
one the most researched topics in the forthcoming concept of bior-
efinery [1].
The sustainability of bio-refineries derives from their ability of
exploiting every product. Furfural (2-FA), a C5 compound, is indus-
trially manufactured for a long time through hydrolysis of pentose
which comes from agricultural raw materials including corncobs,
oat, wheat bran, sawdust, etc. These materials are yearly renew-
able and not competitive with the food sector. Thus, exploring
the products derived from furfural as the replacements of fossil re-
sources is greatly attractive. In fact, utilization of furfural as the
starting material could synthesize a variety of chemicals including
more than 1600 commercial products [2].
3 3 2
esterification with NaCH O and CH OH on the Au/TiO reference
catalyst purchased from the World Gold Council (WGC) [11]. In
fact, gold supported on oxides or carbon, once considered catalyt-
ically inert, is now firmly established as an effective catalyst [12].
Very recently, we have observed [13] good catalytic performances
during esterification of furfural over a gold-supported sulfated zir-
Actually furfural has many different uses: it is considered an
excellent solvent for many organic materials, and it can also be
used as a feedstock to make gasoline, diesel, or jet fuel [3]. It is also
a precursor to other desired compounds such as furfuryl alcohol
2
conia support, especially when compared with the Au/TiO refer-
ence catalyst. Homogeneous oxidative esterifications by gold
have also been reported, but most unfortunately, these depend
on peroxides as oxidants [14]. Afterward, we have investigated a
(
via hydrogenation), furan (via decarbonylation) and tetrahydrofu-
series of Au/ZrO
50 °C up to 650 °C) in order to modulate the size of the gold nano-
particles, demonstrating that in this reaction, the catalytic activity
2
catalysts calcined at different temperatures (from
1
⇑
Corresponding author. Fax: +39 (0)41 2348517.
E-mail address: miky@unive.it (M. Signoretto).
0
021-9517/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcat.2013.10.005

2
42
F. Menegazzo et al. / Journal of Catalysis 309 (2014) 241–247
is strictly connected to the metal dispersion. In particular, the pres-
ence of highly dispersed gold clusters ability to activate atomic
oxygen is required for good catalytic performances. Moreover,
the stability of these new catalysts has been also studied [15].
We demonstrated that the catalytic activity can be completely
recovered when the organic residue of the exhausted sample is re-
moved from both gold and zirconia sites [15]. Such results sug-
Oxidative treatments of the exhausted catalysts were carried
out with a temperature rate of 2 °C/min from 25 °C to 300 °C in a
5%O /He flow (40 mL/min).
2
2.2. Catalyst characterization method
The sulfate content was determined by ion chromatography
(IC). Sulfate concentration was calculated as the average of two
independent analyses, each including two chromatographic
determinations.
The gold amount for both fresh and exhausted catalysts was
determined by atomic absorption spectroscopy (AAS) after micro-
wave disgregation of the samples (100 mg) using a Perkin–Elmer
Analyst 100.
gested that the support also plays
esterification reaction.
a role in the furfural
Therefore, we decided to investigate different oxidic supports
that are commonly used in catalysis. In particular, we examined
plain titania (TiO ), ceria (CeO ), and zirconia (ZrO ). TiO is widely
2 2 2 2
used for a variety of applications because of its high photocatalytic
activity, non-toxicity, good availability, low cost, and stability. Its
main characteristics strongly depend on its physicochemical prop-
erties, such as surface area, crystal structure (anatase, rutile, brook-
ite), crystallite size, and surface hydroxyl groups [16]. Ceria is
characterized by a high oxygen storage capacity and reducibility
Surface areas and pore size distributions were obtained from N
2
adsorption/desorption isotherms at ꢂ196 °C (using a Micromeri-
tics ASAP 2000 analyzer). Surface area was calculated from the
2
N adsorption isotherm by the BET equation, and pore size distri-
[
17]: we can take advantage of these properties in furfural esterifi-
bution was determined by the BJH method [23]. Total pore volume
was taken at p/p0 = 0.99.
cation reaction, in which atomic oxygen produced on gold species
play a fundamental role. Moreover, gold on cerium oxide support
has been shown by IR studies to stabilize gold(III) for heteroge-
neous catalytic applications [18]. Finally, the choice of zirconia as
a support is due to its intrinsic chemical and physical characteris-
tics that can be adjusted by choosing different precursors and syn-
thesis conditions [19].
The aim of the present work is to verify the role of the nature of
the support in the base free oxidative esterification of furfural cat-
alyzed by gold-based systems. In particular, the goal is to investi-
gate the above tested catalysts by employing the main
characterization techniques typically used in surface science
approach.
High-resolution transmission microscopy (HRTEM) analysis was
performed on all catalysts using a side entry Jeol JEM 3010 (300 kV)
microscope equipped with a LaB6 filament and fitted with X-ray
EDS analysis by a Link ISIS 200 detector. For analyses, the powdered
samples were deposited on a copper grid, coated with a porous car-
bon film. All digital micrographs were acquired by an Ultrascan
1000 camera, and the images were processed by Gatan digital
micrograph. A statistically representative number of particles were
counted in order to obtain the particle size distribution.
CO pulse chemisorption measurements were performed at
ꢂ116 °C in a lab-made equipment. Before the analysis, the follow-
ing pretreatment was applied: the sample (200 mg) was reduced in
2 2
a H flow (40 mL/min) at 150 °C for 60 min, cooled in H to room
temperature, purged in He flow, and finally hydrated at room tem-
perature. The hydration treatment was performed by contacting
the sample with a He flow (10 mL/min) saturated with a proper
amount of water. The sample was then cooled in He flow to the
temperature chosen for CO chemisorption (ꢂ116 °C) [24].
2
. Materials and methods
.1. Catalyst preparation
Zr(OH) was prepared by precipitation from ZrOCl
2
FTIR spectra were obtained on a BRUKER IFS28 spectrophotom-
ꢂ1
4
2
2
ꢁ8H O at
eter (resolution: 2 cm , MCT detector). All materials were in-
constant pH = 8.6 and then aged for 20 h at 90 °C [20]. Then, zirco-
nium hydroxide was calcined in air (30 mL/min STP) at 650 °C for
spected in the form of self-supporting pellets (about
ꢂ2
10 mg cm ). All samples were activated in a controlled atmo-
3
h.
Ceria support was synthesized by precipitation from (NH
Ce(NO by urea at 100 °C in aqueous solution [21,22]. The solu-
sphere at 300 °C for ꢃ1 h in a controlled atmosphere (O , ꢃ60 Torr)
2
4
)
2-
in quartz cells that were connected to a gas vacuum line and
3
)
6
equipped with mechanical and turbo molecular pumps (residual
ꢂ5
tion was mixed and boiled for 6 h at 100 °C, and the precipitate
was washed twice in boiling deionized water and dried at 110 °C
overnight. The material was then calcined in flowing air (50 mL/
min) at 650 °C for 3 h.
pressure p < 10 Torr). BT (IR beam temperature, ꢃ50 °C) CO
2
adsorption (ꢃ20 Torr) and desorption (up to 30 min of direct
pumping off in vacuo) measurements were carried out in a strictly
in situ configuration that allowed background subtraction and
spectra rationings.
Titanium hydroxide was precipitated at pH = 8.0 from 0.5 M
titanyl sulfate aqueous solution [6]. In particular, 40 g of TiOSO
xH SO yH O (Aldrich) was dissolved in 300 mL of distilled water
at room temperature under vigorous stirring. The Ti(OH) precipi-
4
TPO measurements were carried out in a lab-made equipment:
samples (100 mg) were heated with a temperature rate of 10 °C/
min from 25 °C to 600 °C in a 5%O /He flow (40 mL/min). The efflu-
2
2
4
2
4
tation was obtained by the drop wise addition of 9 M ammonia
solution under vigorous stirring. The suspension was magnetically
stirred at 60 °C for 20 h. Then, the precipitate was filtered, washed
ent gases were analyzed by a TCD detector and by a Genesys 422
quadrupole mass analyzer (QMS).
2
ꢂ
with distilled water in order to remove SO4 ions, and dried at
10 °C for 18 h. The absence of sulfates in the material was verified
by IEC analysis. Finally, the hydroxide was calcined in air flow at
2.3. Catalytic activity measurements
1
2-FA oxidative esterification with oxygen and methanol was
3
00 °C for 4 h.
investigated at 120 °C, without NaCH
ical stirred autoclave fitted with an external jacket [15]. Catalyst
(100 mg), 2-FA (Sigma Aldrich, >99%; 300 L) and n-octane (Sigma
Aldrich, >99%; 150 L), used as internal standard, were added to
the solvent (150 mL of methanol). The reactor was charged with
oxygen (6 bar) and stirred at 1000 rpm. The progress of the
reaction was determined after 90 min by gas-chromatographic
3
O addition, using a mechan-
In all cases, gold was added by deposition–precipitation (DP)
method at pH = 8.6. The oxide supports were suspended in an
aqueous solution of HAuCl O for 3 h, and the pH was con-
trolled by the addition of NaOH (0,5 M). After filtration, the sam-
ples were dried at 35 °C overnight and finally calcined in air for
1
l
4
ꢁ3H
2
l
h at 300 °C. Samples were denoted as Z–Au, C–Au, and T–Au.

F. Menegazzo et al. / Journal of Catalysis 309 (2014) 241–247
243
analysis of the converted mixture (capillary column HP-5, FID
detector).
Table 1
Physicochemical properties of the samples and catalytic performances.
Au/ZrO
Z–Au
2
Au/CeO
C–Au
2
Au/TiO
T–Au
2
3
. Results and discussion
2
Surface area (m /g)
39
21
0.26
1.0
0.24
105
5
0.16
2.2
0.06
166
4
0.21
1.2
Pore diameter (nm)
Pore volume (cm /g)
3
3.1. Texture, morphology, and structure of the catalysts
Au loading (%)
Chemisorption (molCO/molAu
)
0.004
N
2
physisorption analyses were carried out in order to deter-
mine surface areas and pore size distributions of both supports
and catalysts. No significant differences between the N physisorp-
2
tion analyses of the bare supports and those of the corresponding
gold-based samples have been evidenced. The isotherms for the
catalysts are shown in Fig. 1, while the corresponding data are re-
ported in Table 1.
The C–Au and T–Au samples exhibited isotherms with hystere-
sis loops typical of mesoporous materials with unimodal pore size
distribution. The adsorption for Z–Au sample shifted toward higher
values, indicating the presence of larger pores. Ceria and titania
2
based catalysts exhibited surface area values of 105 m /g and
2
1
66 m /g respectively, while zirconia exhibited the lowest surface
2
area (ꢄ40 m /g).
CO pulse chemisorption measurements were performed in or-
der to obtain information on the amount of highly uncoordinated
Au sites exposed at the surface of the different samples. This
way, it is possible to evidence the highly dispersed Au sites escap-
ing from HRTEM detection. We have previously developed a proce-
dure based on the combined use of pulsed CO chemisorption
measurements and FTIR spectroscopy of adsorbed CO in well con-
trolled experimental conditions to dose low coordination gold sites
2
1
1
00
50
00
Z-Au
2
1
1
0
0
,0
,5
,0
,5
,0
exposed at the surface of TiO
we have proved that this method is suitable also for the quantita-
tive determination of the gold sites on Au/ZrO catalysts [24].
2 2
and CeO oxides [25]. Subsequently,
2
Chemisorption data are reported in Table 1 as molCO/molAu ratio,
and they allow us to directly compare all samples, as they refer to
the number of moles of Au in each sample. CO chemisorption also
gives a measure of gold dispersion and it can be related to the
respective catalytic performances. A high molCO/molAu ratio indi-
cates the presence of clusters that are able to activate molecular
oxygen producing atomic O species and thereby rendering the cat-
alyst more active for furfural esterification reaction [15]. The volu-
metric experiments showed that the molar ratio between adsorbed
CO and gold on Z–Au is quite high, indicating the presence of gold
clusters. An intermediate value of the ratio was obtained for gold
supported on ceria, for which a small amount of Au clusters was
present. The sample supported on titania presented the lowest
molCO/molAu value, indicating that almost only large Au particles
were present on this catalyst.
A careful HRTEM analysis was carried out in order to focus on
the gold dispersion on the different supports. The results are re-
sumed in Fig. 2 where a HRTEM representative image and the me-
tal particle size distribution for each sample are shown. The
measurements revealed the presence of very small roundish Au
particles with average size of 2.3 ± 1.1 nm on the Z–Au catalyst
with a quite narrow and symmetric particle size distribution, indi-
cating that the gold nanoparticles had homogeneous size. How-
ever, we were not able to detect gold particles of even smaller
size by means of HRTEM, but we were aware of their presence from
CO chemisorption data.
50
1
10
100
Pore diameter,nm
0
0.0
0.2
0.4
0.6
0.8
1.0
p/p0
2
00
50
00
C-Au
1
0
0
0
.0
.8
.6
.4
1
1
0.2
.0
0
1
10
100
0.6
50
Pore diameter, nm
0
0
.0
0.2
0.4
0.8
1.0
p/p0
2
00
50
00
T-Au
2
1
1
0
0
,0
,5
,0
,5
,0
1
1
On C–Au, HRTEM measurements evidenced the presence of Au
particles with average size of 2.0 ± 1.4 nm. The particle size distri-
bution was quite narrow and symmetric and it was shifted toward
smaller sizes. However, chemisorption data revealed a small
amount of chemisorbed CO, indicating that on this sample the
amount of clusters was smaller than that of the zirconia-supported
catalysts.
1
10
100
50
Pore diameter, nm
T–Au contained gold nanoparticles with average diameter of
.2 ± 1.5 nm, meaning that large gold particles only were present
on this catalyst. This inference is in good agreement with CO chem-
isorptions. Moreover, when compared to those of the other sam-
ples, the particle size distribution of this sample was very broad
ranging between 2 and 11 nm.
0
4
0.0
0.2
0.4
0.6
0.8
1.0
p/p0
Fig. 1. N
distributions.
2
physisorption isotherms of the catalysts and (insert) their BHJ pore size

2
44
F. Menegazzo et al. / Journal of Catalysis 309 (2014) 241–247
4
3
2
1
0
0
0
0
0
Z-Au
5 nm
0.0
2.5
2.5
2.5
5.0
5.0
5.0
7.5
7.5
7.5
10.0
10.0
10.0
12.5
15.0
4
3
2
1
0
0
0
0
0
C-Au
5 nm
0
.0
12.5
15.0
40
5
nm
T-Au
30
20
10
0
0
.0
12.5
15.0
Particle size (nm)
Fig. 2. HRTEM images of calcined catalysts and gold particle size distribution. The images were taken at an original magnification of 300,000ꢅ.
3.2. Catalytic performances in the furfural oxidative esterification
reaction
The catalysts were tested in the oxidative esterification of furfu-
ral to methyl-2-furoate (see Scheme 1). Such reaction was carried
out in methanol without the addition of NaCH O, which would
3
potentially make the esterification process less environmentally
friendly and more expensive [10]. A part from 2-MF, the only by-
product that was found is the acetal derivate, as reveal by mass
spectroscopy.
In Fig. 3 the conversions and the 2-MF selectivities after 90 min
of reaction are reported. Z–Au catalyst showed very good catalytic
performances for both conversion (82%) and selectivity (92%). The
catalytic performance of this sample was due to the very high dis-
persion of the gold active phase, as indicated by CO chemisorption
selectivity, %
conversion, %
Fig. 3. Catalytic performances after 90 min of reaction.
data. The Au clusters are able to activate molecular oxygen produc-
ing atomic O species that render the catalysts more active for fur-
fural esterification reaction [15]. The C–Au catalyst showed a lower
conversion (66%) than Z–Au sample. The molCO/molAu ratio ob-
tained for this catalyst was 0.06, indicating the presence of small
gold nanoparticles. Selectivity was much lower if compared to
the zirconia catalyst. In fact, this sample presented a remarkable
formation of 2-furaldehyde-dimethyl-acetal. The conversion for
the T–Au catalyst was very low (20%), according to the presence
of big gold particles with size around 4 nm, even if the selectivity
was surprisingly good and on a par with the value attained by Z–
Au sample (above 90%).
2
-furaldehyde (2-FA)
reagent)
(
methyl-2-furoate (2-MF)
product)
(
2
-furaldehyde-dimethyl-acetal
(by-product)
Scheme 1.

F. Menegazzo et al. / Journal of Catalysis 309 (2014) 241–247
245
Table 2
1590 1417
Comparison between the ‘‘normalized’’ catalytic conversion and the integrated area of
the 1224 cm spectral component.
ꢂ1
0
.2
Conversion/SSAa
a.u.
Sample
Integrated area of the
ꢂ1
1
224 cm spectral component
1380
1310
1224
Z–Au
C–Au
T–Au
82/39 = 2.1
66/105 = 0.62
20/166 = 0.12
2.89
0.15
0.68
T-Au
Z-Au
a
1435
Obtained normalizing the catalytic conversion toward the BET-specific surface
area.
1622
B
1224
1190
B
B
1
680
All these results quite clearly indicate that gold dispersion is a
crucial factor controlling conversion. However, gold dispersion it-
self cannot explain the trend observed as for selectivity. Table 1 re-
ports the gold content for each sample, which is in the 1–2 wt%
range. Despite a lower Au loading, if compared to the other cata-
lysts, the zirconia-supported sample presented the best catalytic
results, indicating a vital role played by the support. To accommo-
date this anomaly, we normalized the catalytic conversion toward
a BET-specific surface area: see Table 2. It is well known from the
literature [26] that despite the much lower SSA exhibited by the
zirconia-based catalyst, it always exhibits the best catalytic perfor-
mance, as activity is strongly influenced by reactivity of the sup-
port. In order to shed some light on this aspect, we performed
1
687
1420
C-Au
1800
1600
1400
1200
-1
Wavenumber (cm )
2
FTIR spectra relative to the adsorption/desorption of CO at BT.
2
Fig. 4. Differential spectra relative to CO adsorption/desorption at BT onto the
different Au-supported catalysts. Black curves refer to the adsorption of the probe
molecule, whereas red curves refer to the desorption (i.e., pumping off) in vacuo for
3.3. FTIR spectra of adsorbed CO
2
30 min. (For interpretation of the references to color in this figure legend, the reader
is referred to the web version of this article.)
ꢂ
When present, CO
2
can interact with surface basic sites, like O2
anions if present, forming carbonate- and/or bicarbonate-like
(
hydrogen-carbonate) species [27]. These species are characterized
2
the CO molecule involved in linear coordination with surface cat-
ionic sites, i.e. Ti ions, present at the surface of the catalyst, thus
revealing a residual (somewhat important) Lewis acidity.
As for the C–Au system, carbonate-like species still formed, but
had little in common with those formed in the other Au-based
systems. This indicates that their structure is typical of the so-
ꢂ1
4+
by peculiar IR bands in the 2000–1000 cm . Their resistance to
outgassing is a clear indication of their stability. Fig. 4 provides
the main spectroscopic features indicating the interaction between
the probe molecule and the different Au-based catalysts. Prior to
adsorption, all samples have been thermally activated in order to
get rid of most of the contaminants (water, anions, hydro-carbona-
ceous residues) typically present on the surface of such oxidic sys-
tems. In Fig. 4 all the spectral patterns are reported as differential
spectra obtained by subtracting the contribution of the spectrum
relative to the plain activated sample from the spectra of the ad-
sorbed or desorbed probe molecule. Accordingly, all the bands cor-
respond to species which form or resist to adsorption or desorption
respectively. The detailed analysis of the various spectral patterns
reveals that on T–Au and Z–Au (see the corresponding curves in
Fig. 4) there is a fair formation of:Carbonate-like species, due to
called bi-dentate carbonates, which are characterized by the
ꢂ1
m
(O–C–O) modes in the 1680–1790 cm spectral range and below
ꢂ1
ꢃ1200 cm
. These species are structures that are bridging
through two O species of the carbonate molecule and one surface
cation, i.e. Me species, and are typical of the medium–high dehy-
n+
dration stages of oxides [29]. Moreover, one cannot exclude the
formation of hydrogen-carbonate species for this system; as very
weak absorptions are present in their typical positions (see the
broken-line arrows reported in Fig. 4). In summary, to obtain the
best performing catalyst activity, both gold dispersion (for good
conversion) and acid/base properties of the support (for good
selectivity) must be taken into account. In order to check if a trend
can be put into evidence, we took the spectral component located
the ‘‘side-on’’ coordination of CO
2
onto pairs of coordinatively
ꢂꢂ
nþ
unsaturated (cus) O2
M
ions; the spectral components corre-
sponding to the formation of these species are quite complicated
by many factors, among which are the natural width of the bands,
the presence of band pairs for each kind of carbonate (due to the
ꢂ1
at ꢃ1224 cm as a ‘‘measure’’ of the intrinsic reactivity of the var-
ious supports: the integrated area of this component has been re-
ported in Table 2. It can be observed that the trend is the following:
Z–Au ꢀ T–Au ꢀ C–Au. On the basis of all the obtained results, it
can be concluded that the presence of suitably located free OH
groups, which bring about the formation of hydrogen-carbonate
species, is vital to lead to a good catalytic activity. This is achieved
in the case of zirconia-supported catalyst, for which a high gold
dispersion is observed as well. In the case of the T–Au sample,
symmetric and antisymmetric
m of the O–C–O groups), and the
large variety of species formed;Bicarbonate-like species (indicated
by B and evidenced by the green dotted line), due to the interaction
of CO with suitably located free OH groups. B species exhibit, be-
2
sides the O–C–O stretching modes pair (visible at ꢃ1620 and
ꢂ1
ꢂ1
ꢃ1450 cm ), a rather narrow peak at ꢃ3610 cm
(stretching,
ꢂ1
m
OH; not reported for the sake of brevity), and a ꢃ1225 cm
4
+
(
bending,
r
OH) vibration.
which exhibits a quite good amount of Ti –OH species that can ex-
plain its intrinsic good selectivity, its low conversion value can be
justified by the worst Au dispersion. Finally, C–Au system pos-
sesses good conversion, due to its good Au dispersion, but exhibits
low selectivity due to the poor reactivity of its surface leading to
the formation of hydrogeno-carbonate species.
Moreover, it can be mentioned that the main carbonate-like
species formed mono-dentate species (with spectral components
ꢂ1
located at ꢃ1590, 1420–1430 and 1310 cm ) for all systems
[
28]. However, only in the case of the T–Au sample a peculiar band
ꢂ1
located at 1380 cm was evident: this was due to the
R
c+ mode of

2
46
F. Menegazzo et al. / Journal of Catalysis 309 (2014) 241–247
100
3
.4. Investigation on stability and reusability of the catalysts
1st
2nd
1 st
Catalysts deactivation is a major challenge in catalytic pro-
8
0
0
cesses. However, tracing the origin of deactivation is often difficult,
due to several simultaneously occurring mechanisms contributing
to the loss of activity and selectivity. A better understanding of the
deactivation processes is essential in order to minimize additional
costs and for improving and/or optimizing (i) process conditions,
1st
2 nd
6
(
ii) catalysts design and (iii) for preventing premature catalyst deg-
40
radation. Combined with the relatively high price of gold, im-
proved stability of the catalyst will add competitiveness to a
technology. The study of catalyst deactivation is mainly a charac-
terization oriented problem, as the relevant main mechanisms
are sintering, metal leaching and poisoning. Sintering, which is
considered to be an irreversible phenomenon, leads to a reduction
in the active surface area. Loss of active components by leaching
also leads to catalyst deactivation. Either reactants or products
generated from the main or side reactions may accumulate at the
active surface; carbon containing species may build up on the sur-
face and poison or physically block the active sites.
2nd
2
0
100
1st
2 nd
8
0
0
2
nd
1
st
6
Very recently we have investigated the stability of gold-based
catalysts supported on sulfated zirconia in the furfural oxidative
esterification [15]. We have ruled out metal leaching and gold sin-
tering problems throughout the reaction time as the cause of deac-
tivation of our catalysts. As revealed by both TPO and FTIR
analyses, the reason for catalysts deactivation is due to their ten-
dency to adsorb organic-like species that accumulate during the
reaction. The deactivation is reversible and, after heating in oxygen
atmosphere at a proper temperature (450 °C), the catalyst surface
is again set free by evolution of carbon dioxide [15]. Such oxidation
treatment of the exhausted catalyst allows to almost completely
recovering catalytic performances [15].
40
2
nd
1 st
20
0
Z-Au
C-Au
T-Au
Fig. 6. Catalytic performances in the reuse tests.
When compared with Z–Au sample, both conversion and selectiv-
ity of the T–Au sample are very low after the same thermal regen-
eration in oxygen. However, conversion increases slightly, while
selectivity drops down. Characterization data have shown that,
after the first reaction, the exhausted T–Au sample has only
The TPO profiles of the various catalysts after reaction have
been reported in Fig. 5. The presence of one broad band that starts
at about 150 °C and closes at 500 °C was formed in the curve of Z–
2
Au. The band is due to CO evolution, as confirmed by mass spec-
troscopy that generates from the decomposition of organic species
present at the surface of the catalyst. Therefore, the catalyst has
been heated in oxygen atmosphere at 450 °C, in order to eliminate
the contaminants. After regeneration at 450 °C in oxygen, the cat-
alytic performance of the Z–Au sample is almost completely recov-
ered, as reported in Fig. 6.
On the other hand, the TPO profile of the spent T–Au catalyst
presents two peaks centered at 250 °C (broad) and 360 °C (very
sharp), respectively, suggest that very low conversion data was ob-
served due to the poisoning of the surface large gold particles.
0
.81 wt% of gold, indicating that the catalyst is strongly affected
by metal leaching in the reaction medium. In this case the calcina-
tion treatment at 300 °C is then likely not sufficient to establish a
strong metal/support interaction. We can conclude that the T–Au
sample is not recyclable and it is not suitable for this catalytic
reaction.
The TPO of the exhausted C–Au sample presents two very weak
peaks at 180 °C (very broad) and 350 °C (sharp), indicating the
higher reducibility of the ceria support compared to the other sys-
tems. After regeneration, the catalytic performances are similar to
that of the prepared sample, as shown in Fig. 6. This result might be
due a partial covering of the surface by contaminants. The C–Au
catalyst is therefore recyclable, and presents no gold leaching dur-
ing the reaction.
Z-Au
C-Au
4
. Conclusion
Gold-based catalysts were investigated in the oxidative esterifi-
cation of furfural by an efficient and sustainable process. The furo-
ate ester can be obtained with optimal yields by a process more
environmentally friendly than the actual one.
A comparison among metals, such as Zirconium, Cerium, and
Titanium, as a support for Au-based catalysis was performed. Cat-
alytic performances followed the trend of Z–Au > C–Au ꢀ T–Au.
Both chemical and morphological properties, such as (i) high dis-
persion of Au, (ii) specific surface area of the support, and (iii)
proper surface sites on the support itself, influence the esterifica-
tion process. While the surface area of the support influenced the
conversion, surface sites affected the selectivity in the process. In
T-Au
0
100
200
300
400
500
600
700
800
Temperature, °C
Fig. 5. TPO of the exhausted catalyst after 90 min of reaction.

F. Menegazzo et al. / Journal of Catalysis 309 (2014) 241–247
247
this study, zirconium proved to be the ideal support for Au-based
oxidation of furfural as it is active, selective, recyclable, and appli-
cable in a biomass based renewables producing industry.
[11] Sample number 17C, supplied by World Gold Council. <http://www.gold.org>.
[
[
12] A.S.K. Hashmi, G. Hutchings, Angew. Chem. Int. Ed. 45 (2006) 7896–7936.
13] F. Pinna, A. Olivo, V. Trevisan, F. Menegazzo, M. Signoretto, M. Manzoli, F.
Boccuzzi, Catal. Today 203 (2013) 196–201.
[
[
14] A.S.K. Hashmi, C. Lothschutz, M. Ackermann, R. Doepp, S. Anantharaman, B.
Marchetti, H. Bertagnolli, F. Rominger, Chem. Eur. J. 16 (2010) 8012–8019.
15] M. Signoretto, F. Menegazzo, L. Contessotto, F. Pinna, M. Manzoli, F. Boccuzzi,
Appl. Catal. B 129 (2013) 287–293.
Acknowledgments
We thank Mrs. Tania Fantinel for technical assistance. Financial
support to this work by MIUR (Cofin 2008) is gratefully
acknowledged.
[16] V. Trevisan, M. Signoretto, F. Pinna, G. Cruciani, G. Cerrato, Chem. Today 30
2012) 25–28.
(
[
[
17] A. Trovarelli, Catal. Rev. – Sci. Eng. 38 (1996) 439–520.
18] S. Carrettin, C. Blanco, A. Corma, A.S.K. Hashmi, Adv. Synth. Catal. 348 (2006)
1283–1288.
[
[
19] X. Song, A. Sayari, Catal. Rev. – Sci. Eng. 38 (1996) 329–412.
20] M. Manzoli, F. Boccuzzi, V. Trevisan, F. Menegazzo, M. Signoretto, F. Pinna,
Appl. Catal. B 96 (2010) 28–33.
21] L. Kundakovic, M. Flytzani-Stephanopoulos, J. Catal. 179 (1998) 203–221.
22] F. Menegazzo, P. Burti, M. Signoretto, M. Manzoli, S. Vankova, F. Boccuzzi, F.
Pinna, G. Strukul, J. Catal. 257 (2008) 369–381.
23] S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and Porosity, second ed.,
Academic Press, 1982.
24] F. Menegazzo, F. Pinna, M. Signoretto, V. Trevisan, F. Boccuzzi, A. Chiorino, M.
Manzoli, Appl. Catal. A 356 (2009) 31–35.
25] F. Menegazzo, M. Manzoli, A. Chiorino, F. Boccuzzi, T. Tabakova, M. Signoretto,
F. Pinna, N. Pernicone, J. Catal. 237 (2006) 431–434.
26] F. Moreau, G.C. Bond, Catal. Today 122 (2007) 215–221.
27] G. Busca, V. Lorenzelli, Mater. Chem. 7 (1982) 89–126.
28] V. Bolis, G. Magnacca, G. Cerrato, C. Morterra, Top. Catal. 19 (2002) 259–269.
29] C. Morterra, G. Emanuel, G. Cerrato, G. Magnacca, J. Chem. Soc., Faraday Trans.
References
[
1] I. Agirrezabal-Telleria, J. Requies, M.B. Güemez, P.L. Arias, Appl. Catal. B 115–
16 (2012) 169–178.
[
[
1
[
[
2] H. Guo, G. Yin, J. Phys. Chem. C 115 (2011) 17516–17522.
3] R. Weingarten, J. Cho, C. Conner, G.W. Huber, Green Chem. 12 (2010) 1423–
[
[
[
1
429.
4] S. Fernando, S. Adhikari, C. Chandrapal, N. Murali, Energy Fuels 20 (2006)
727–1737.
5] A.S.K. Hashmi, T.M. Frost, J.W. Bats, J. Am. Chem. Soc. 122 (2000) 11553–
1554.
[
[
1
1
[
[
[
[
6] A.S.K. Hashmi, T.M. Frost, J.W. Bats, Org. Lett. 3 (2001) 3769–3771.
7] A.S.K. Hashmi, T.M. Frost, J.W. Bats, Catal. Today 72 (2002) 19–27.
8] A.S.K. Hashmi, M. Wölfle, J.H. Teles, W. Frey, Synlett 11 (2007) 1747–1752.
9] E. Taaring, I.S. Nielsen, K. Egeblad, R. Madsen, C.H. Christensen, ChemSusChem
[
[
[
[
1
(2008) 75–78.
88 (1992) 339–348.
[
10] O. Casanova, S. Iborra, A. Corma, J. Catal. 265 (2009) 109–116.