M.L. Dieuzeide et al.
Catalysis Today xxx (xxxx) xxx–xxx
hydrogenolysis of glycerol, some of them are Al
2
O
3
[8,14,15,18,22–25];
, ZnO/ZrO [13,28], ZnO
29] and MgO [16]. The role of the support on the hydrogenolysis of
performed with Siemens D5000 equipment, employing Cu Kα radiation.
Temperature programmed reduction (TPR) of fresh samples (after
calcination) were performed in a Micromeritics Autochem II 2920, with
a thermic conductivity detector (TCD). The samples (100 mg) were
placed in a quartz U-shaped reactor. Previously to temperature pro-
grammed reduction, samples were pre-treated under a flow of Ar
(50 mL/min) at 200 °C for 1 h. TPR was performed from 50 °C to 800 °C
at a heating rate of 10 °C/min, under a flow of 100 mL/min of a mixture
SiO
2
[9,10,13,26,27]; ZnO/Al
2
O
3
, ZnO/TiO
2
2
[
glycerol to 1,2-PDO has been extensively discussed. Some authors have
proposed that the acid or basic character of the support determine the
reaction mechanism, when the reaction is carried out in liquid phase,
and affects 1,2-PDO selectivity [4,6,17]. On the other hand, it has been
proposed that, when the reaction is performed in vapor phase, metal-
support interactions must be considered and the support has an im-
portant role in promoting the dispersion of the metallic phase [4]. In
2%H
2
/Ar. Hydrogen consumption was determined by a TCD detector.
O adsorption method was performed in a
The dissociative N
2
fact, in a previous study of our group it was proved that Al
activity in vapor phase hydrogenolysis of glycerol [25].
2
O
3
has no
Micromeritics Autochem II 2920 in order to determine copper metallic
area and dispersion. The catalysts sample (100 mg) was placed in a U-
shaped quartz reactor and was pre-treated in flowing Ar (50 mL/min) at
100 °C for 30 min, followed by cooling at room temperature. The cat-
alyst pre-reduction, was performed increasing the temperature to
Considering, as mentioned before, that the catalyst support has a
great influence on the metal-support interaction as well as on metal
dispersion [4], we have considered employing mesoporous alumina as
support for copper based catalysts. Yuan et al. [30] proposed a facile
synthesis of highly ordered mesoporous alumina with high thermal
stability and tunable pore size by self-assembly of Pluronic 123,
500 °C with a ramp of 10 °C/min under a 2% H
for 30 min. Then the sample was cooled to 50 ± 5 °C in Ar flow
(50 mL/min) and sequentially was exposed to a 50% N O/Ar flow
(100 mL/min) for 1 h, in order to oxidize the Cu° to Cu O by dis-
sociative adsorption of N O. Finally, after the purge of the sample under
Ar flow (50 mL/min) at 50 °C for 15 min, the TPR was carried out, in
order to reduce the Cu O species to metallic copper. This stage was
performed in a 2% H /Ar flow (100 mL/min) and temperature was
2
/Ar (100 mL/min) flow
2
(
EO)20(PO)70(EO)20, triblock copolymer and alumina precursors in
2
ethanolic solutions in the presence of additives such as nitric acid. This
synthesis was then extended to obtain mesoporous alumina-supported
noble metals or metal oxides [31–34]. Nevertheless there are few stu-
dies considering the impregnation of CuO over mesoporous alumina
2
2
2
[
35,36], these studies concluded that the textural and structural char-
increased to 500 °C with a 10 °C/min ramp. The copper metallic area
and dispersion, were calculated based on bibliography [37,38], con-
sidering that the number of superficial copper atoms per unit surface
acteristic of mesoporous alumina have great influence both on copper
metallic dispersion and on copper-support interactions.
Therefore, the aim of the present work is to analyse the effect of
copper loading on catalytic activity in the hydrogenolysis of glycerol in
vapor phase at atmospheric pressure, and its correlation with metallic
copper dispersion.
area is 1.47 × 1019 atoms/m and the density of copper is 8.92 g/cm .
2
3
2.3. Catalytic activity
The hydrogenolysis of glycerol was carried out at atmospheric
2. Experimental
pressure in
a stainless-steel continuous flow fixed bed reactor
(
Ø = 12 mm) placed in an electric furnace equipped with temperature
2
.1. Catalyst preparation
controllers. Reaction temperature was measured with a k-type ther-
mocouple, placed in the middle of the catalytic bed. For all catalytic
tests, the liquid stream was fed with an HPLC bomb (Eldex 1HM) and
vaporized in the initial third of the reactor. The liquid stream consisted
of a water glycerol solution with a molar ratio (R = nH O/nC H O )
Mesoporous alumina was prepared following a similar procedure to
that reported by Yuan et al. [30] and Morris et al. [31]. For this
synthesis, 25 g of (EO)20(PO)70(EO)20 (Pluronic 123 of Sigma Aldrich)
triblock copolymer were dissolved in 100 mL of anhydrous ethanol
2
3 8 3
R = 9:1 (35 wt% glycerol); with a liquid feed rate of 2.4 mL/h. The
catalytic runs were performed isothermally at 240 °C, 0.5 g of catalyst
were employed. Catalyst was diluted within an inert material in a ratio
1:5, in order to avoid temperature gradients. The hydrogen-glycerol
molar ratio was 65:1. The feed stream was completed with Ar, as car-
rier. Both hydrogen and argon were fed to the reaction system by mass
flow controllers (Brooks 0254), being the gaseous feed rate 360 mL/
min.
(
99.5% Cicarelli). Then, 50 g of aluminum isopropoxide (98% Sigma
Aldrich) were dissolved in 40 mL of nitric acid (65 v/v% Cicarelli) and
60 mL of ethanol. Once both solutions were dissolved they were
4
combined, employing additionally 20 mL of ethanol, in order to transfer
the solution of aluminum isopropoxide. The combined solution was
kept under stirring for 24 h. Solvent evaporation was done at 60 °C for
4
8 h under air without stirring. Finally mesoporous alumina was ob-
tained after calcining the precursor at 600 °C for 5 h.
Copper impregnation was performed by incipient wetness impreg-
3 2 2
nation method with aqueous solutions of Cu(NO ) ·3H O (99.5%
The total flow rate and particle diameter were chosen in order to
guarantee the absence of diffusional resistance during reaction tests.
Catalysts are reduced in situ at 500 °C (heating ramp of 10 °C/min)
Merck), with concentrations ranging between 0.06 M and 6.3 M.
Previously to impregnation, mesoporous alumina was ground and
sieved, in order to obtain particles with diameters between
2
under a flow of 50% H /Ar (100 mL) during 30 min and under a flow of
pure hydrogen (100 mL) for another 30 min. Then the catalytic bed
temperature was set at reaction temperature (240 °C) under an Ar flow.
Both liquid feed samples and condensed samples were analysed by a
GC (Agilent Technologies 7890A, DB-5, 30 m × 0.320 mm × 0.5 μm).
Liquid samples were collected every hour during reaction. The internal
standard method was used for the quantification of the results, being n-
butanol the standard. The liquid products analysed were: 1,2-propa-
nediol (propylene glycol), 1,3-propanediol, ethylene glycol and hy-
droxyacetone (acetol); no propanol was detected in the condensed
stream. Gas stream was analysed by a GC (Agilent Tecnologies 6890N,
Carboxen™ 1010 Plot, 30 m × 0.53 mm), however no gaseous products
were detected except for non-reacted hydrogen. Since hydrogen is in
excess respect to glycerol, it was not possible to estimate its con-
sumption by chromatography.
4
4 μm < dp < 88 μm. After impregnation with copper solutions, sam-
ples were dried at 120 °C for 6 h and then calcined at 400 °C for another
h. Both drying and calcination of impregnated samples were carried
on in a stove under air atmosphere, with a temperature ramp of 10 °C/
min. Fresh alumina is denoted as m-Al and the catalysts were de-
noted as: CuO(x)Al, being x the nominal CuO (wt%) loading, between
6
2 3
O
3
wt% and 30 wt%.
2.2. Catalyst characterization
Catalysts were characterized by several techniques.
Textural characterization was performed by N sorptometry in a
2
Micromeritics equipment ASAP 2020, employing 20 mg of sample.
In order to analyse the catalytic results, the following parameters
were considered:
Characterization by powder X-ray diffraction (PXRD) was
2
Dieuzeide
