S. Torres et al.
Applied Catalysis A, General 621 (2021) 118199
have been widely studied for converting glycerol to lactic acid at mod-
erate reaction temperatures at different NaOH/glycerol mole ratio. More
recently, cobalt-based catalysts have also received attention for selec-
tively transforming glycerol into different chemicals [37,38]. Cerium
using the BET method and pore volume and pore size distribution were
determined using the BJH method.
The XRD patterns of the catalysts were obtained using a Panalytical
diffractometer equipped with a Pixcel-3D solid-state detector (Empy-
oxide supported Co
3
O
4
particles is reported to be active in the selective
α
rean Series 2), using CuK radiation (1.5406 Å), operating at 40 kV and
◦
◦
conversion of glycerol to lactic acid in aqueous medium and basic pH at
40 mA. Scans were recorded in the 10 ≤ 2θ ≤ 90 interval using a step
◦
◦
2
50 C, reaching a selectivity of 79.8 % with a conversion of 85.7 %
size of 0.01 and a step time of 6 s. The identification of the crystalline
[
38]. Nevertheless, the catalyst exhibited low stability under cycling. It
phases was carried out using the ICSD database.
is worth noting that it is unknown which species of cobalt are more
active and what is the effect of the support on the catalytic properties of
XPS characterization of the catalyst’s surface was performed using a
SPECS, NAP-XPS, spectrophotometer equipped with a PHOIBOS 150 1D-
Co
3
O
4
particles in the partial oxidation reaction of glycerol to lactic acid.
DLD analyzer, using a monochromatic Al-K X-ray source (1486.7 eV)
α
Therefore, in order to determine the potential of these catalysts in the
selective conversion of glycerol to lactic acid, it is necessary to continue
the development and physicochemical characterization.
operated at 100 W and 13 kV. The pass energy was set to 90 eV for the
survey spectra and at 30 eV for high-resolution spectra with steps of 1 eV
and 0.1 eV, respectively. The surface charge compensation was
controlled with a flood gun operated at 3 eV and 20 mA. The analysis of
the spectra was performed using a binding energy (BE) scale calibrated
by adjusting the adventitious carbon C-H to 284.0 eV and for quantifi-
cation, the values of the Relative Sensitivity Factors (RSF), obtained
from the Scofield database were: C 1s (1.0), O 1s (2.93), Co 2p (19.16),
Ce 3d (51.62), Zr 3d (7.04) and Ti 2p (7.81). The XPS spectra were
analyzed using CasaXPS software and all signals were treated using an
Off Shirley background and for analysis of the signals, asymmetric
functions were considered using a line shape SGL(p)T(k).
In this work, spinel-like cobalt oxide particles are supported in three
different supports but with similar textural properties: cerium oxide,
zirconium oxide, and titanium oxide. Co
3 4
O is stable under basic con-
ditions as well as CeO , ZrO and TiO . The aim is to determine the
2
2
2
influence of the support chemical properties in terms of acidity and
redox properties on spinel-like cobalt oxide species and to determine the
activity and stability of these catalysts in glycerol selective conversion to
lactic acid.
2
. Experimental
Characterization of morphology and structural parameters was per-
formed using High Resolution Transmission Electron Microscopy
(HRTEM). The micrographs were obtained using a Tecnai F20 Super
Twin TMP instrument operated at 200 kV. For analysis, samples were
dispersed in ethanol thorough out sonication for 30 min before being
dropped on carbon-coated copper grid.
2
2
.1. Supports preparation
.1.1. CeO
CeO was prepared by hydrothermal treatment of a solution con-
taining 9 g of Ce(NO
2
synthesis
2
3
)
3
⋅6H
2
O dissolved in 240 mL of ultrapure H
2
O and
Temperature programmed reduction analysis was carried out in an
AUTOCHEM 2010 Micromeritics equipment. For the analysis, 200 mg of
◦
1
9 g of NaOH (NaOH:Ce molar ratio of 23) at 100 C for 14 h [38].
◦
ꢀ 1
solid was treated at 400 C under Ar (flow rate of 50 mL min , heating
◦
ꢀ 1
◦
2
.1.2. ZrO
50 mL of a 2.5 M NaOH solution was added dropwise to a solution
containing 9 g of ZrO(NO O dissolved in 150 mL of ultrapure H
⋅XH
39]. The resulting mixture was stirred for 30 min and then it was
2
synthesis
rate of 5 C min ) for 1 h. Then, the powder was cooled to 50 C and
◦
1
finally heated to 900 C under H
rate of 50 mL min-1 and a heating rate of 10 C min
2
(diluted in Ar, 10 vol.%) with a flow
◦
ꢀ 1
3
)
2
2
2
O
.
[
Temperature programmed desorption of ammonia (NH -TPD) was
3
◦
transferred to an autoclave for a hydrothermal synthesis at 170 C for
performed in an AUTOCHEM 2010 Micromeritics instrument. Briefly,
◦
5
h.
200 mg of the solid was treated at 400 C for 1 h (heating rate of
◦
ꢀ 1
ꢀ 1
1
0 C min ) under He (flow rate of 30 mL min ). After cooling down to
◦
2
.1.3. TiO
2
synthesis
150 C, NH
3
was adsorbed by exposing the powder to a flow of NH
3
ꢀ
1
1
0 mL of titanium isopropoxide (TTIP) was mixed with 15 mL of
diluted in He (5 vol.%, flow rate of 30 mL min ) for 2 h. Then, the gas
ꢀ
1
◦
ethanol, then 15 mL of ultrapure H O was added dropwise and the
2
flow was changed to He (flow rate of 30 mL min ) for 1 h at 150 C with
the aim to eliminate the physisorbed NH . Finally, NH was desorbed by
heating from 50 C to 800 C under a He flow (heating rate of
resulting mixture was stirred for 2 h at room temperature [40], then the
3
3
◦
◦
◦
mixture was transferred into an autoclave and heated at 80 C for 4 h.
◦
◦
ꢀ 1
ꢀ 1
Finally, the recovered solid was calcined at 400 C for 4 h.
10 C min , flow rate of 30 mL min ). The signal was followed by TCD
and MS detectors.
2
.2. Catalysts preparation
Raman spectra were obtained with a Raman confocal instrument,
Horiba Jobin Yvon (Labram HR model), using excitation laser with a
wavelength of 632.81 nm, spectra were recorded from 100 to 2000
3
g of support was dispersed in 10 mL of ultrapure H
2
O and then a
ꢀ 1
solution of Co(NO O containing the required amount for the
3
)
2
⋅6H
2
cm ; using a D 0.3 filter, a slit of 600 m, acquisition time of 20 s.
μ
desired cobalt loading (20 wt%) was added dropwise. The resulting
mixture was stirred for 1 h at room temperature, followed by a slow
Thermogravimetric analysis (TGA) was performed on a TA In-
struments SDT-Q600 equipment. 20 mg of the powder was deposited in
◦
◦
evaporation of the solvent at 50 C under stirring. Then, the solid was
an Al
ꢀ 1
rate of 5 C min , under an air rate flow of 100 mL min .
2
O
3
capsule and heated from room temperature to 900 C, heating
◦
◦
ꢀ 1
further dried at 50 C in a vacuum oven and finally calcined in air at
◦
◦
ꢀ 1
4
00 C during 4 h, with a heating rate of 2 C min .
2
.4. Catalytic tests
Activity of catalysts in glycerol selective conversion to lactic acid was
2
.3. Characterization of catalysts
The experimental content of cobalt in the catalysts was determined
evaluated in a batch Parr reactor with a capacity of 250 mL. The cata-
◦
by atomic emission spectroscopy using a 4200 MP-AES Agilent tech-
nologies spectrophotometer, equipped with a CCD detector, wavelength
of 340.512 nm for cobalt. For analysis, 3 mg of powder was dissolved in
lytic reactions were performed at 250 C, since at this temperature the
conversion of glycerol in presence of cobalt oxide catalysts is favored
[38]. The reactor was charged with a solution containing 5 wt% glycerol
concentration with a NaOH:glycerol molar ratio of 1, and 0.6 g of
3
aqua regia (HCl : HNO volume ratio 3 : 1) and HF.
Nitrogen adsorption-desorption isotherms were obtained in
a
catalyst. The reactor was sealed, purged with N
2
and then heated under
◦
Micromeritics ASAP 2020 instrument. Before analysis, the samples were
autogeneous pressure to 250 C. After reaching the reaction tempera-
◦
degassed at 250 C for 12 h. The specific surface area was determined
2
ture, it was pressurized with 32 bar of N (the total pressure was 70 bar)
2