M.N. Gatti et al.
Catalysis Today xxx (xxxx) xxx–xxx
oxidic supports to make them less susceptible to hydrolytic attack at
200 °C. The approach involves the deposition of a thin film of carbon
obtained by controlled pyrolysis of sugars on commercial SiO2. The
results obtained showed a reduced sintering of the metallic particles
and almost no leaching. Even though the results are interesting, the
difficulty of the method lies in controlling the carbon thickness that is
deposited to avoid covering the particle.
on the activity and selectivity of the catalyst. The nickel catalysts were
characterized by different techniques to correlate their activities with
the structure of the solids. We also studied the effect of the main
operating variables of this reaction, such as temperature, glycerol
concentration, reaction time and hydrogen partial pressure.
2. Experimental
[15,23,24] have been reported in the literature. Even though the
catalysts based on noble metals show adequate levels of activity and
selectivity, these precious elements are limited in their application due
to their low abundance and high costs. Besides, noble metal catalysts
generally present an undesired activity towards the cleavage of CeC
bonds.
Xiong et al. [11] reported a deposition–precipitation–carbonization
method to prepare highly dispersed Pd/niobia/carbon catalysts with
improved hydrothermal stability.They found that the niobia/carbon
compounds helped prevent the growth of oxide crystallite size and
preserved the Pd dispersion.
2.1. Catalyst synthesis
The gelling property of TEOS (SILBOND 40-AKZO Chemicals) was
used in an alcoholic medium (Ethanol 96% from Anedra) to include a
phenol-formaldehyde liquid resin (RL 43003, ATANOR, Argentina) in
its structure. With subsequent curing and pyrolisis in reducing atmo-
sphere, this resin leaves a high amount of residual carbon.
TEOS and RL 43003, with a 1:1 mass ratio, were mixed on a
magnetic stirrer until obtaining an emulsion, to which ethanol was
gradually added. Afterwards, pregellification occurred at room tem-
perature for 24 h, drying at 50 °C for another 24 h and complete
polymerization by heating to 180 °C (heating rate 10 °C h−1) for 3 h.
Finally, calcination took place in a reducing atmosphere during 3 h on
an electrical oven at 1580 °C (at 5 °C min−1). This material was
denominated SC-095.
A portion of the SC-095 composite was treated with a solution at
10 wt% HF (48 wt% from Aldrich) for 30 min. This solid was washed
with distilled water until obtaining a value of neutral pH, filtered and
dried at 120 °C for 24 h. The presence of residual H2F6Si was eliminated
by heating at 400 °C during 1 h. The material thus obtained was
denominated SC-015.
The catalysts were prepared by incipient wetness impregnation of
Ni (NiCl2·6H2O from Sigma-Aldrich) on supports SC-095 and SC-015.
Ethanol was used as solvent. The NiCl2·6H2O concentration in ethanol
was calculated so as to obtain 5 wt% of Ni in the final solid. Then, the
solid obtained was dried at 120 °C during 12 h and activated in H2
(50 cm3 min−1) at 400 °C for 90 min (10 °C min−1).
Gallegos-Suarez et al. [25] conducted the hydrogenolysis of glycerol
(10 wt%, 8 MPa of H2 and 180 °C) over Ru-based catalysts supported on
activated carbon (AC), high surface area graphite (HSAG), multiwalled
carbon nanotubes (CNT) and KL-zeolite. They found that in Ruthenium
catalysts supported on HSAG and CNT the electron donor character of
the respective supports stabilizes electron-rich metal species (Ruδ−
)
favoring the cleavage of the CeO bond.
Non-noble metal catalysts are mainly based on Cu [15,26–33] and,
of the addition of P2O5 (2,3 and 4 wt%) to a Cu/SiO2 (Cu 20 wt%)
catalyst. The increase of catalytic activity in glycerol hydrogenolysis at
220 °C and 5 MPa H2 was attributed to a combination of electronic and
geometric effects induced by the presence of P2O5 on the metallic
particles of Cu. The Cu-P interaction increased the distance Cu–Cu and,
consequently, the electronic density was increased, thus provoking the
increase in catalytic activity. This happened with low contents of (2 wt
%), while with higher contents of P2O5, the formation of aggregates led
to the loss of active sites and a consequent decrease in activity.
Dasari et al. [15] evaluated a Ni/C commercial catalyst and
obtained a conversion of 40% and a selectivity towards 1,2-PG of
69% at 24 h on stream operating at 200 °C and 1.4 MPa, employing
glycerol at 80 wt%. W. Yu et al. [34] employing catalysts of Ni/AC
(activated carbon) (Ni 10 wt.%), demonstrated that high nickel disper-
sions and the acidity generated by the surface oxygenated groups of
carbon have a synergic effect on the activity of the hydrogenolysis
reaction.
Chau et al. [36] synthesized bifunctional catalysts of Ni and
silicotungstic acid (HSiW) supported on Al2O3 by a sequential incipient
wetness impregnation method (Ni 1 and 10 wt%) for the hydrogeno-
lysis of glycerol (240 °C and 6 MPa H2). Due to the strong acid character
of the catalyst, the main product was 1-propanol. The authors
concluded that in order to obtain a desired product selectively, the
control of reaction conditions and catalyst properties such as acid
strength, the amount of appropriate acid sites, and metal hydrogenation
activity will be needed.
Jiménez-Morales et al. [37] prepared and characterized a catalyst of
Ni (10 wt% Ni) supported on a support of non-acid SBA-15 mesoporous
silica promoted by the addition of cerium (2.5 and 10 wt.% of Ce). The
effect of cerium oxide is to act as a strong acid site promoting the
formation of acetol, which is later reduced by H2.
From the above-mentioned studies, it can be concluded that the
development of catalytic materials, hydrothermally stable and selective
to the desired products, is still a challenge. The aim of the present work
is to prepare a nickel catalyst with a metal loading of 5 wt% Ni
supported on a SiO2-C composite, to be used in the liquid-phase glycerol
hydrogenolysis reaction. For this purpose, two SiO2-C composites with
different composition were studied to analyze the effect of the support
2.2. Characterization
The Ni content of samples was determined by atomic absorption
spectrometry (Spectrophotometer AA-6650 Shimadzu).The equipment
utilized was an IL Model 457 spectrophotometer, with a single channel
and double beam.
Adsorption-desorption measurements were performed for a textural
characterization.
Brunauer–Emmett–Teller (BET) multipoint method and textural analy-
sis were obtained using a Micromeritics ASAP 2020 equipment. The
samples were pretreated under vacuum in two 1 h stages at 100 and
300 °C.
XRD patterns were recorded on a Philips 3020 powder diffract-
ometer, using Cu Kα radiation (λ = 1.5418 Å, intensity = 40 mA, and
voltage = 35 kV). The patterns were recorded in the range of
2θ = 20°−70°. The crystallite sizes of metallic nickel in the reduced
samples were calculated using Scherreŕs equation:
Surface
area
measurements,
the
K.λ
B.cos(θ)
dXRD
=
Where K was taken as 0.89 and B was the full width of the diffraction
line at half of the maximum intensity in radians (Nickel (111) planes).
The acid base properties of supports were determined by the test of
isopropanol decomposition (IPA). This reaction was tested in a con-
tinuous-flow fixed-bed reactor between 200 and 300 °C, atmospheric
pressure, feed 4.5% IPA in Helium, with a flow of 40 cm3 min−1
.
The acidic groups content was determined according to Boehm’s
method. The titration results allowed obtaining the concentration of
carboxylic, lactonic and phenolic groups present in the supports,
assuming that NaHCO3 neutralizes carboxylic groups, Na2CO3 neutra-
2