Chelating agent effects in the synthesis of supported Ni nanoparticles as catalysts for hydrogen production

Article abstract of DOI:10.1016/j.apcata.2021.118219

A series of Ni supported over MgAl₂O₄–CeO₂ catalysts have been prepared by the wet impregnation method as catalysts for steam ethanol reforming. The synthesis was performed using nitrilotriacetic acid (NTA) or citric acid (CA) as chelating agents (L) with different L/Ni molar ratios. The features of catalytic solids were determined by DR–UV–Vis–NIR, FTIR, BET, TGA, XRD, H₂–TPR, XPS, Raman, HRTEM and SEM. Changes in the crystallite size of NiO and CeO₂ were evidenced with the use of NTA, while the addition of CA allowed to reduce the crystallite size of NiO and Ni with a slight effect in CeO₂ size. The use of chelating agents induced changes in the Ni-CeO₂ interactions and an increase in the Ce/Ni⁰ surface ratio. Catalytic systems prepared using an L/Ni ratio leading to the most stable complex formation (NTA/Ni=1 and CA/Ni=2) exhibited the best performances in the reforming reaction under the operation conditions studied.

Full text of DOI:10.1016/j.apcata.2021.118219

Applied Catalysis A, General 622 (2021) 118219
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Chelating agent effects in the synthesis of supported Ni nanoparticles as
catalysts for hydrogen production
a,
a
a
b
Alejandra C. Villagranꢀ Olivares *, Mariana N. Barroso , Carlos A. L o´ pez , Jordi Llorca ,
a
María C. Abello
a
Instituto de Investigaciones en Tecnología Química (INTEQUIꢀ CONICETꢀ UNSL), Facultad de Química, Bioquímica y Farmacia. Almirante Brown 1455, 5700, San
Luis, Argentina
b
Instituto de T ´e cnicas Energ ´e ticas, Departamento de Ingeniería Química, Universitat Polit `e cnica de Catalunya, EEBE, 08019, Barcelona, Spain
A R T I C L E I N F O
A B S T R A C T
Keywords:
A series of Ni supported over MgAl O –CeO catalysts have been prepared by the wet impregnation method as
2
4
2
Nanoꢀ Ni catalysts
Chelating agent
Complexing method
Hydrogen
catalysts for steam ethanol reforming. The synthesis was performed using nitrilotriacetic acid (NTA) or citric acid
(
CA) as chelating agents (L) with different L/Ni molar ratios. The features of catalytic solids were determined by
DR–UV–Vis–NIR, FTIR, BET, TGA, XRD, H –TPR, XPS, Raman, HRTEM and SEM. Changes in the crystallite size
of NiO and CeO were evidenced with the use of NTA, while the addition of CA allowed to reduce the crystallite
size of NiO and Ni with a slight effect in CeO size. The use of chelating agents induced changes in the Ni-CeO
2
2
Ethanol steam reforming
2
2
interactions and an increase in the Ce/Ni0 surface ratio. Catalytic systems prepared using an L/Ni ratio leading to
the most stable complex formation (NTA/Ni=1 and CA/Ni=2) exhibited the best performances in the reforming
reaction under the operation conditions studied.
1
. Introduction
C
2
H
4
O(g) + H
2
O(g)→2 CO(g) + 3H
2
(g) ΔH 09 23K = + 186.8 kJ m olꢀ 1
(
3)
Ethanol steam reforming (ESR) is a potential sustainable hydrogen
C
2
2
H
5
5
OH(g)→C
2
H
4
(g) + H
2
O(g) ΔH 02 = + 45.5 kJm olꢀ 1
(4)
production process, and it has been extensively studied over non-noble
transition metal catalysts because they are less expensive than the noble
metal ones. In particular, Ni catalysts are preferred because Ni dissoci-
98K
C
H
OH(g)→CO(g) + CH
4
(g) + H
2
(g) ΔH0 = + 49.6 kJ m olꢀ 1
298K
–
ates C C bonds more efficiently than Co [1–3]. ESR is an endothermic
(5)
(6)
process, which produces 6 mol of hydrogen and 2 mol of carbon dioxide
if only the reforming reaction occurs (Eq. 1). However, other reactions
usually occur such as ethanol dehydrogenation, acetaldehyde steam
reforming, ethanol dehydration, ethanol decomposition, acetaldehyde
decomposition, water gas shift reaction, ethylene decomposition,
methane decomposition and Boudouard reaction (Eqs. 2–10) [1]:
C
2
H
4
O(g)→CH
4
(g) + CO(g) ΔH0
=
ꢀ 19.3 kJ m olꢀ 1
298K
CO(g) + H
2
O(g) ↔ CO
2
(g) + H
2
(g) ΔH0
=
ꢀ 41.4 kJ m olꢀ 1
(7)
(8)
298K
C
2
H
4
(g)→2 C(s) + 2H
2
(g) ΔH0 = ꢀ 52.3 kJ m olꢀ 1
298K
0
2
ꢀ 1
C
2
H
5
OH(g) + 3H
2
O(g)→2 CO
2
(g) + 6H
2
(g) ΔH
ΔH
= + 173.5 kJ m ol
= + 206.4 kJ m ol
(g) + C(s) ΔH02
= + 74.9 kJm olꢀ 1
98K
23K
CH
4
(g)→2H
2
(9)
98K
0
9
ꢀ 1
2 CO(g) ↔ CO (g) + C(s) ΔH0 = ꢀ 172.6 kJm olꢀ 1
(
1)
(10)
2
298K
0
= + 68.9 kJm olꢀ 1
The occurrence of these reactions mainly depends on the properties
of catalyst and operation conditions. One of the disadvantages of the
secondary reactions over Ni catalysts is the deactivation by carbon
deposition, sintering and active phase oxidation [4]. Ethylene
C
2
H
5
OH(g)→C
2
H
4
O(g) + H
(g) ΔH
2
(2)
2
98K
*
Corresponding author.
E-mail address: avillagran@unsl.edu.ar (A.C. Villagranꢀ Olivares).
https://doi.org/10.1016/j.apcata.2021.118219
Received 11 February 2021; Received in revised form 6 May 2021; Accepted 15 May 2021
Available online 19 May 2021
0
926-860X/© 2021 Elsevier B.V. All rights reserved.

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
decomposition, Boudouard reaction and methane decomposition are the
main routes to carbon deposition [1–3,5]. However, there is no linear
relationship between the decrease in catalyst activity and coke content
resistance to carbon deposition [30]. In addition, the effect of EDTA/Ni
molar ratio was examined in the synthesis of Ni catalysts [31]. The ca-
pacity of EDTA to form stable complexes with nickel and cerium pro-
[
5]. These phenomena of deactivation are the main technological ob-
2
motes a better interaction between Ni and CeO and decreases the size of
stacles in today’s reformers. The development of a Ni stable catalyst
could be achieved by metallic particles below a critical size. As in the
methane reforming the control of metal assembly size in the ESR has the
same objective: to increase dispersion and decrease carbon deposition
by inhibiting its formation reactions. Small particles decrease carbon
deposition rate [6], however, they are more susceptible to sintering and
oxidation. The method variation of the active phase incorporation in a
catalyst could result in different physicochemical properties, allowing
low carbon amounts and increasing the stability of Ni catalysts [7,8]. In
addition, the characteristics of the precursor species of the active phase
often affect the catalyst surface properties and its catalytic behaviour.
The choice of a precursor is a very important step in catalyst prepara-
tion. Aqueous solutions containing chelating agents in the wet impreg-
nation method has been reported as a strategy to promote smaller active
phase particles [9]. Van Dillen et al. [9] have pointed out that the use of
aqueous solutions of chelating agents in the conventional wet impreg-
nation method is very attractive, since it retains the simplicity of the
technique and it allows to obtain uniform distributions of the active
phase and, in general, greater dispersions. A characteristic of these so-
lutions is that, during the solvent remotion, the viscosity of metal
complex solutions increases substantially due to a gelling process,
affecting the drying stage, inhibiting the redistribution of surface species
and favoring the formation of poorly crystallized compounds with high
dispersion. In literature, the use of chelating agent has been reported
mainly in the synthesis of hydrodesulfurization catalysts [10–12], cat-
alysts for Fischer-Tropsch synthesis [13,14] and recently, for reforming
catalysts [15–17].
NiO and CeO . A catalyst prepared using the stoichiometric ratio for the
2
complete complexation of nickel (EDTA/Ni molar ratio = 1) showed the
best performance in ethanol steam reforming reaction under mild
operation conditions and presented small crystallite size, and higher
3
+
4+
◦
Ce /Ce and Ce/Ni ratios.
In this work, the use of different amounts of NTA or CA as chelating
agent in the synthesis of supported-Ni catalysts and their influence in
ESR reaction are discussed.
2. Experimental
2.1. Catalyst preparation
Nickel catalyst were synthetized following the procedure reported in
[30]. Briefly, MgAl
2
O
4
was prepared using citrate method and it was
(5 wt.%, nominal loading) by wet impregnation
.6H O (Aldrich, 99 %). Nickel (8
wt.%, nominal loading) was incorporated over the modified support,
MgAl (in its dried form), using aqueous solution of Ni
ꢀ CeO
(NO O (Merck, 99 %) in presence of chelating agent (L), NTA
⋅6H
modified with CeO
2
with an aqueous solution of Ce(NO
3
)
3
2
2
O
4
2
3
)
2
2
(Sigma-Aldrich, 99 %) or CA (Sigma-Aldrich, 99.5 %), in different L/Ni
molar ratios (L/Ni = 0, 0.5, 1 and 2). First, the chelating agent was
dissolved using a NH
3
solution and then Ni(NO
3
)
2
⋅6H
2
O solution was
added. The pH of these impregnation solutions was adjusted at a pH
). The pH should
between 6 ꢀ 7 (lower than the PZC of MgAl
2
O
4
ꢀ CeO
2
be lower than the PZC to ensure a strong interaction between the
complex with negative charge and the surface of support with a positive
charge [32,33].These solutions were aged under stirring at room tem-
The incorporation of chelating agents into a transition metal solution
generates metal complexes that are charged. In the catalysts preparation
by using these impregnation solutions it is necessary to take into account
factors such as: chelating agent/metal ratio, pH of impregnation solution
and the point of zero charge (PZC) of support [18]. Some published
works on ESR have reported important improvements in catalytic
properties by using citric acid (CA) [19,20]. Nitrilotriacetic acid, NTA,
and CA are chelating agents with the same amount of carbon atoms. NTA
have carboxylates and nitrogen groups, while CA has carboxylate groups
and a hydroxyl group. The donor capacity of a chelating agent is given
by the amount of binding sites presents in its structure, both functional
groups and coordination sites, as well as the pair of free electrons of the
nitrogen atom in NTA or oxygen atom in CA. The greater the
electron-giving capacity of a chelating agent, the greater the bond
strength with the central ion [21] and consequently more stable the
complex formed will be. NTA and CA acids have the same amount of
functional groups and coordination sites. However, NTA forms com-
plexes more stable with nickel and cerium than CA [22,23], then it is
interesting to compare the behaviour of both acids.
perature for 2 h. Then, 1 g of the MgAl
2
O
4
ꢀ CeO
2
was added and stirred
for 1 h. The resulting suspensions were carried to an ultrasound bath at
room temperature for 10 min and then the solvent was slowly removed
◦
under vacuum at 80 C for 1 h. The dried samples were calcined in static
◦
◦
ꢀ 1
air from room temperature to 650 C at 10 C min . The catalysts were
labeled as NiL(x), where x corresponds to the L/Ni molar ratio, being L =
NTA or CA. NiL(0) corresponds to the nickel catalyst free of chelating
agent.
2.2. Characterization of catalysts
2.2.1. Chemical composition
The solid samples were brought into solution by an alkali fusion with
KHSO4 and subsequent dissolution with aqueous HCl. Ni chemical
composition was measured by atomic absorption spectroscopy using
Varian AA 50 equipment. Cerium chemical composition was determined
by inductively coupled plasma-atomic emission spectroscopy (ICP). A
sequential ICP spectrometer Baird ICP 2070 (BEDFORD) with a Czerny
Turner monochromator (1 m optical path) was employed.
On the other hand, in the literature it has been reported the modi-
fication of acid/base properties of supports by addition of modifiers such
as alkaline metals (Na, K), alkaline earth metals (Mg, Ca) or the addition
2.2.2. Adsortionꢀ desorption N2 isotherms
of rare earth oxides (Ce, La, Pr) [3,8,24]. CeO
2
has been used as a sup-
The specific surface area was determined by the BET method from
port or promoter because its particular properties of oxygen storage and
release that allow carbon deposits to be gasified and the presence of the
the adsorption-desorption isotherms of N at the boiling temperature of
2
◦
liquid nitrogen (ꢀ 195.8 C), using a Micromeritics Gemini 5 equipment.
3
+
4+
◦
Ce /Ce couple [25–27]. Other approach to enhance the properties of
CeO is doping ceria with other elements, such as Zr or Pr [28]. The
Previously, samples were degassed at 250 C for 16 h in N flow. From
2
2
the desorption branch, pore size distribution values were obtained using
the BJH model (Barreto, Joyner and Halenda) suitable for mesoporosity
determination.
incorporation of other cation with different size into the ceria structure
to make oxygen release easier [26]. Elmhamdi et al. have also pointed
out the importance of having CeO
properties [29]. In a previous work, the decrease of Ni and CeO
talline sizes on Ni/MgAl catalysts due to the use of chelating
2
at a nanometric level to enhance its
2
crys-
2.2.3. Temperature programmed reduction (TPR)
2
O
4
ꢀ CeO
2
The redox properties of the samples were examined using conven-
tional equipment. Before to the TPR experiments, the samples were
agents has been reported [30]. The system prepared in the presence of
ꢀ
1
◦
nitrilotriacetic acid (NTA), with a molar ratio of NTA/Ni=1 showed a
treated in O flow (17 %)/He (36 mL min ) at 300 C for 1 h and then in
2
ꢀ 1
good performance in ESR, obtaining high selectivity to H
2
and high
He flow (30 mL min ) for 30 min. After cooling in He samples were
2

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
(50 mL min 1) from room temper-
ꢀ
and a resolution of 0.19 nm was used for transmission electron micro-
scopy. The samples were deposited in lacey-carbon Cu grids.
reduced in flow of H
2
(5 vol.%)/N
2
◦
◦
ꢀ 1
ature up to 950 C at 10 C min , keeping this temperature for 1 h.
2
.2.4. Fourier transformed infrared spectroscopy (FTIR)
2.2.9. Thermal gravimetric analyses (TGA)
The samples were analyzed by infrared spectroscopy using an FTIR
The decomposition of the precursors of dried catalytic systems was
performed in oxidizing atmosphere in a TGA 51 Shimadzu unit, with an
ꢀ 1
Nicolet Prot ´e g ´e ꢀ 460 spectrophotometer, in a range of 4000–250 cm
ꢀ 1
ꢀ 1
◦
◦
with a resolution of 4 cm and 64 scans. The samples were mixed with
KBr at approximately 1 wt.%, crushing the mixture into an agate mortar.
air flow of 50 mL min from room temperature up to 800 C at 5
ꢀ 1
min . The amount of carbon deposited on catalysts after the ESR re-
action was studied in oxidizing atmosphere under an air flow of 50 mL
ꢀ 2
Then, the mixture was compressed at a pressure of 300 Kg m and room
temperature.
ꢀ
1
◦
◦
min , from room temperature to 900 C with a heating rate of 10 C
ꢀ 1
min . The samples were placed in a platinum capsule. The percentage
of carbon deposited during the reaction for catalysts was calculated
using the following equation:
2
.2.5. UV–vis and UV–vis–NIR spectroscopy
The UV–vis spectra of the different impregnation solutions were
obtained using an Agilent 8454 spectrophotometer with diode network
detector using the 8453 UV Visible ChemStation Rev.A.10.01 software,
associated with the instrument. Catalytic samples were packed using
(W
initial ꢀ Wfinal)∗100
%C =
W
final
BaSO
4
as blank and placed on the integrative sphere, ISRꢀ 3100. The
(Winitial ꢀ Wfinal) is the amount of carbon on the catalyst and Wfinal is
Shimadzu UVꢀ 3600 was operated in diffuse reflectance mode in the
the final weight of the sample after TGA analysis.
UV–Vis–NIR (DR UV–Vis–NIR) range. The spectra for all samples were
ꢀ 1
taken between 2500–200 cm with a slit of 5 nm.
2.2.10. Raman spectroscopy
Renishaw’s inVia Qontor Raman microscope equipment was used.
The spectra were taken with the visible laser of 532 nm with 15 accu-
2
.2.6. Xꢀ ray diffraction (XRD)
ꢀ 1
The diffraction patterns of the samples were obtained in a RIGAKU
mulations (3 s), in the range between 600–1800 cm
.
(
ULTIMA IV) diffractometer operated at 30 kV and 20 mA using CuKa
radiation (λ = 1.5418 Å) with Ni filter and horizontal goniometric. The
2.2.11. Scanning electron microscopy and energy dispersive Xꢀ ray
spectroscopy (SEMꢀ EDAX)
◦
ꢀ 1
◦
data were acquired at 3 min , for values of 2θ between 10 and 70 . The
powder samples were placed without any prior treatment in a flat glass
sample holder. The identification of the crystalline phases was per-
formed using reference patterns (JCPDS files of the ICDD). To study the
crystallite size, the patterns were recorded in a step mode (2θ = 10 ꢀ
For qualitative determination of the local composition and texture of
catalyst surfaces, a scanning electron microscope, LEO 1450 VP equip-
ped with an EDAX Genesis 2000 microchannelizer with Si (Li) detector,
was used. The samples were metalized with gold to prevent them from
loading.
◦
◦
9
0 ; step = 0.02 and counting time: 5 s) and refined by the Rietveld
method using the Fullprof program [34,35]. Fig. S1 shows Rietveld re-
finements. For this purpose, the identified phases were considered with
2
.3. Catalytic experiments
Catalytic tests of the ESR reaction were carried out at 650 C and
2 4 2
the following structural models: MgAl O (cubic, Fd 3 m), CeO (cubic,
Fm 3 m), NiO (cubic, Fm 3 m) and Ni (cubic, Fm 3 m). The shape of the
peaks was modeled using the Thompson-Coxꢀ Hastings pseudo-Voigt
function [36]. In some samples the crystallite size was obtained by
applying the Scherrer equation [37]:
◦
atmospheric pressure, in a quartz tubular reactor (4 mm internal
diameter), for 7 h. A previous inꢀ situ reduction procedure was per-
◦
formed with a flow of H
2
(5%) in N
2
for 45 min at 650 C. The catalyst
Kλ
β =
weight was 0.05 g with a particle size range between 0.3–0.4 mm (35–50
mesh) without dilution. The reactor feed consisted of liquid mixture of
d cosθ
ꢀ
1
◦
ethanol and water (0.15 mL min ), which was evaporated at 130 C in a
He flow free of O . The steam to carbon ratio, S/C, was equal to 2.5, with
a composition of 9.4 % ethanol (yC2H5OH) in the feed at an ethanol flow
β is the width of the peak that corresponds to a given value of 2θ, λ
corresponds to the wavelength of the X ꢀ rays, K is a constant that is a
function of the crystalline shape but it is generally taken equal to 0.9 for
spherical particles and d corresponds to the crystallite size. The d value
was manually obtained using the width of the following planes: (200)
2
ꢀ 3
ꢀ 1
ꢀ 1
of 1.02 10 mol min . The space-time was 0.018 gcat h gC2H5OH (49 gcat
ꢀ 1
min molC2H5OH). This reduced space-time allowed early detection of any
deactivation effect [5].
2
and (111) for NiO and CeO , respectively. The sizes of reduced samples
Reagents and reaction products were analyzed on-line by two gas
chromatographs. Catalytic activity experiments were performed in
duplicate to verify their reproducibility and the percent of carbon
deposition. The repeatability of data was in the range ± 5%. The cata-
were also obtained from Rietveld refinements where the size of each
phase was automatically calculated by the program from all reflections
and the Scherrer equation. In both cases the instrumental broadening
was considered in the calculus.
2
lytic experiments for NiNTA(0.5) and NiCA(1) were only run once. H ,
CH
while CO (previous passage through a methanator), C
OH without reacting and other compounds containing carbon, with
the FID.
Ethanol conversion was calculated as:
4
and CO
2
were determined with TCD using N
2
as internal standard;
2
.2.7. Xꢀ ray photoelectron spectroscopy (XPS)
2
4 2 4
H O, C H and
XPS analysis was performed on a SPECS Multitechnical equipment
2 5
C H
equipped with an Mg/Al dual X ꢀ ray source and a PHOIBOS 150
hemiespheric analyzer in Fixed Analyzer Transmission (FAT) mode. The
spectra were obtained with a pass-through energy of 30 eV and Al anode
operated at 200 W. The pressure in the analysis chamber during the
in
out
C2H5OH = F
ꢀ F
C2H5OH
C2H5OH
ꢀ 8
X
∗100
in
measurement was less than 2.10 mbar. To reduce the load effect on
the samples, a "flood gun" electron cannon was used. Before acquisition
F
C2H5OH
FC2H5OH is the molar ethanol flow; in and out correspond to inlet and
of the spectra, samples were subjected to an in-situ reduction in H
2
/Ar
◦
outlet flow of the reactor.
flow for 10 min at 400 C, and then evacuated to a pressure less than 2.8
ꢀ 7
Selectivity to carbon products was calculated as:
×
10 mbar for at least two hours before readings.
ni F io ut
S
i
=
ꢀ
) ∗100
2
.2.8. High resolution transmission electron microscopy (HRTEM)
in
out
2
F
ꢀ F
C2H5OH
C2H5OH
A TECNAI F20 FEI equipment with field emission source at 200 kV
3

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
out
i
F
is the outlet flow of each hydrocarbon and ni is the number of carbon
atoms in each product i.
Hydrogen yield was calculated as:
out
F
H2
Y
H2 = Fin
C2H5OH
F
oH u2 t is the outlet flow of hydrogen of the reactor.
3
. Results and discussion
3
.1. Impregnation solutions
Thermodynamic speciation diagrams of nickel were obtained using
Medusa Software [38]. They were calculated considering the ionic
strength and the concentrations of each species present in the solution.
Table 1 shows the predominant nickel species in impregnation solutions
used in the synthesis of catalysts. The synthesis of catalysts was made by
impregnation of support using a solution with a pH below to the PZC of
the support (PZC = 9.6 [30]) to ensure a strong interaction between the
complex with negative charge and the surface of support with a positive
charge. At pH of impregnation solution, the thermodynamic diagrams
◦
predict the formation of different nickel species at 25 C. Only one
ꢀ
4ꢀ
complex species in solution, [Ni(NTA)(H
2
O)
2
]
and [Ni(CA)
2
]
, was
predicted by employing an L/Ni ratio leading to the most stable complex
formation (NTA/Ni = 1 and CA/Ni = 2). Different nickel species in so-
lution were predicted for the other L/Ni ratios and they could promote
different interactions with the support affecting its catalytic properties.
UV–vis spectra of the impregnation solutions allow to identify the
nickel coordination in the solution, Fig. 1A. The figure shows the UV–Vis
spectra of the impregnation solution containing NTA and CA with
different L/Ni molar ratio. In the UV zone it is observed at 300 nm the
ꢀ
π
ꢀ
π
* transition of the NO
3
. The visible zone of the spectra shows the d-
2
+
Fig. 1. UV–vis spectra of impregnation solutions (A) and DR UV–vis–NIR
d transition of Ni in an octahedral environment (Table S1 [39,40]),
spectra of dried catalysts (B).
3
3
3
3
allowed transitions: λ
1
: A2g (F) → T1g (P); λ
2
: A2g (F) → T1g (F); λ
3
:
3
3
3
1
A
2g (F) → T2g (F); and forbidden spin transition: λfs: A2g (F) → E
g
(D).
Fig. 1B shows the DR UV–Vis–NIR spectra of dried systems. Between
The blue shift of the nickel bands using NTA or CA with different L/Ni
ratios reveals the complex formation under the experimental synthesis
conditions. The chelating agent, NTA or CA, produces a slight blue shift,
2
ꢀ
4+
2
00–300 nm it is observed a band associated with O → Ce transition
3
+
and/or 4f→ 5d of Ce transition [42] or associated with charge trans-
ference between O2 → Ni [43]. In all spectra a band at 300 nm
ꢀ
2+
2
+
indicating the complex formation by water substitution in [Ni(H
41].
O) ]
2 6
ꢀ
associated with the
πꢀ
π* transition of NO
3
without coordination is
[
observed, interacting only by physical adsorption with the support [42,
4]. Table S1 shows the transitions in dried systems compared to
impregnation solutions. For all dried systems, in the Vis ꢀ NIR zone, two
4
3
.2. Dried systems
The nickel coordination on the surface of dried catalysts was studied
bands are observed, λ
2
and λ
3
, between 500 and 1500 nm, while the λ
1
transition could be overlapped with the band at 300 nm. In the NiL(0)
dried sample a slight blue shift is noticed respect to the bands in its
impregnation solution. This indicates that the nickel aquocomplex, [Ni
by DR UV–Vis–NIR spectroscopy. From a comparison between impreg-
nation solution spectra and dried catalysts spectra it is possible to
identify if the complex in solution is maintained in the dried system.
2
+
(H O) ] , is interacting as such with the support. The positions of the
2
6
bands in the NiNTA(x>0) dried samples have experimented a slight red
Table 1
shift respect to bands of impregnation solutions. The same occurs in the
2
+
Nickel species at pH of impregnation solutions.
NiCA(0.5) and NiCA(1) dried samples. Unexpectedly, the Ni bands are
of low intensity in the NiCA(2) system. The red shift or towards higher
wavelength values indicate that the nickel complex on the surface is a
lower strength field. This red shift is related to a bonding weakening
System
pH
I*
Complex in solution
2
+
Solution L-Ni(0)
Solution NTA-Ni
5.9
6.6
0.4
1.3
[Ni(H
[Ni(NTA)(H
2 6
O) ]
ꢀ
ꢀ
ꢀ
2+
; [Ni(NH
2
2
2
O)
2
2
2
]
; [Ni(H
2
O)
6
]
3
)
among the chelating and Ni2 by formation of hydrogen bonding be-
tween OH groups of the support and the oxygen atoms of the chelating
agent. During drying step the chelated metal precursor forms a gel-like
phase and this gel would sufficiently interact with the support [9].
Ortega-Domínguez et al. [45] have informed similar results using a
Ni-EDTA impregnation solution over SBAꢀ 15. Koizumi et al. [46] have
+
2
+
(
0.5)
Solution NTA-Ni
1)
Solution NTA-Ni
2)
Solution CA-Ni
0.5)
2 5
(H O) ]
6.7
6.6
6.5
6.5
6.6
1.9
2.3
1.6
1.6
2.4
[Ni(NTA)(H
[Ni(NTA)(H
O)
O)
]
(
4ꢀ
] ; [Ni(NTA)
2
]
(
ꢀ
4ꢀ
[Ni(CA)(H
2
O)
2
]
; [Ni(CA)
2
]
3
; [Ni(NH )
2
+
(
(H
2
O)
5
]
4
ꢀ
ꢀ
2+
Solution CA-Ni(1)
[Ni(CA)
2
2
]
; [Ni(CA)(H
2
O)
2
] ; [Ni(NH
3
)
also reported the same interaction between NTA and Co in the solu-
+
4
(
H
2
O)
5
]
tion and claimed that it was maintained on the dried surface of Co
catalysts synthesized with NTA. NIR zone, between 800–2500 nm,
shows a band around 1570 nm corresponding to the first overtone of
ꢀ
Solution CA-Ni(2)
*
[Ni(CA)
2
]
∑
1
2
i
I = Ionic strength. I was calculated as I = 2
ciz . c
i
and z
i
are the con-
i
ν
(NH) vibration [47]. Between 1500 and 2500 nm vibrations related to
centrations of each ion and its charge, respectively.
4

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
◦
chelating structures are observed (they are not present in the NiL(0)
dried spectrum). NiL(0) dried sample only shows a band at 1425 nm that
corresponds to the first overtone of OH neutral species. In the NiNTA
precursor species. NiNTA(0.5) shows significant weight loss from 230 C
◦
to constant weight at 400 C. The maximum decomposition temperature
◦
at 245 C is associated to the decomposition of the nickel aquocomplex
(
x>0) or NiCA(x>0) major differences with the increase of chelating
and the complex with NTA. The maximum decomposition temperatures
◦
agent are not observed, indicating that the nickel complex on the surface
of the dried systems could have the same nature.
are detected at 231, 294 and 360 C for NiNTA(1), while two weight loss
◦
◦
events are observed at 245 C and around 375 C for NiNTA(2). The
latter event is related to decomposition of free NTA. In the NiNTA(x>0)
systems are not possible to dismiss the Ce-NTA complex formation. NiCA
In order to identify the species on the surface in the dried systems,
samples were examined by FT-IR analyses (Fig. 2). All the systems
ꢀ 1
◦
prepared with NTA or CA show, between 2600 and 3700 cm , a band
(0.5) only shows a weight loss at 245 C, while in NiCA(1) are observed
ꢀ 1
◦
associated with OH group vibrations (3470 cm ) and bands related to
three weight loss events at 233, 278 and 358 C. The main weight loss is
ꢀ
1
◦
◦
symmetric and asymmetric stretching of CH
2
(3140–2800 cm ) [48]. In
O) appears [49]. The
shown at 243 C and the smallest one at 393 C for NiCA(2). The pres-
ꢀ
1
◦
1
630 cm the band of H
2
O deformation (δH
2
ence of a weak peak between 300 and 500 C is only observed in NiCA(1)
characteristic sharp bands of nitrate are observed in agreement with the
sample. Probably the interaction between CA and Ce is negligible, its
stability constant is lower in comparison with NTA. Wang et al. [48]
ꢀ
NO
3
πꢀ π* transition assigned in UV–Vis spectroscopy. In particular, the
ꢀ 1
transitions of nitrate ion at 826 and 1765 cm are observed in all the
have studied the decomposition in air of Na
2
[Ni(CA)]. These authors
ꢀ 1
◦
dried samples and the transitions at 1393 and 1630 cm
are less
have identified two combustion events: a weight loss between 30–340 C
notorious in the systems prepared with chelating agents. The bands
attributed to coordinated water removal and CO
2
release, and the latter
ꢀ 1
◦
detected between 1700–1200 cm correspond to the contribution of
carboxylates group bonded to a metal and the presence of nitrate ion
without coordination [50]. Furthermore, it is not observed an additional
in the range of 340–520 C to the combustion of the organic component.
◦
They did not observe mass changes at temperatures below 300 C, and it
is considered that some of the H
2
O were incorporated into the structure
].
ꢀ 1
band to values higher than 1630 cm indicating the absence of car-
boxylates without coordination [51,52]. The major contribution for all
the dried systems prepared with NTA or CA correspond to the symmetric
of the complex like [Ni(CA)(H O)
2
3
ꢀ
ꢀ
3.3. Fresh and reduced catalysts
stretching of COO
(
ν
s
) and asymmetric stretching of COO
(νas) [53]
and it indicates that the metal is coordinated by carboxylate groups
belonging to chelating agents regardless of the L/Ni molar ratio used.
Infrared spectra of fresh systems (Fig. S2) show symmetric and
Table 2 shows the values of chemical composition obtained for the
catalysts. Nickel and cerium content were lower than the nominal values
probably due to the precursor salts that are easy to hydrate. For NiNTA
ꢀ
asymmetric bands of COO groups. This carbonaceous species corre-
(
2) and NiCA(2) the Ni values are higher and this fact will contribute to
sponds to wastes of the synthesis of the support. They remain after the
its catalytic performance. N adsorption-desorption isotherms are shown
2
◦
combustion of citric acid used in MgAl
2
O
4
preparation at 650 C for 3 h
in Fig. S4A. The isotherms are IV type with H2 hysteresis for all the
catalysts in this work, this is typical of mesoporous solids with cylindric
porous or aggregates or agglomerates of spherical particles with non-
uniform size and shape [55]. The SBET values are presented in Table 2.
[
31]. It is not observed differences in the spectra of calcined samples
when NTA or CA is used in the impregnation solutions. This indicates
that the vibrational nature of oxide species is similar in all catalysts
independently of chelating agent and the quantity used.
A decrease in all the catalysts respect to the support, MgAl
2
O
4
ꢀ CeO
2
,
Thermal decomposition of precursor was studied under an oxidant
atmosphere. Fig. S3 shows the thermograms and dweight/dt vs. tem-
perature of dried catalysts where different weight loss events are
observed. The theoretical weight losses considering that precursor
decomposition leads to NiO are shown in Table S2. The experimental
values are lower than the theoretical ones, except for the NiCA(0.5)
sample. Similar results were observed with catalysts prepared with
EDTA [31], suggesting that the decomposition began in the dried step. In
ꢀ 2
SBET = 102 g m is observed. This decrease could be related with the
thermal treatment after the impregnation step or the porous blockage by
NiO phase impregnated [56]. The SBET for the catalysts is not affected by
the nature of the chelating agent or for the L/Ni molar ratio. Pore size
distribution (Fig. S4B) is similar for all the catalysts independently of the
precursor used in the synthesis. The catalysts have larger porous respect
to MgAl
2
O
4
ꢀ CeO
2
, with a high contribution of porous between 2 and 8
nm and an increase in the porous size associated with decomposition of
precursor during the calcination step [57].
◦
the sample free of chelating agent, NiL(0), the weight loss up to 100 C
could be assigned to the elimination of physisorbed water, and that at
X-ray patterns of fresh catalysts are shown in Fig. S5. MgAl
and NiO phases are observed in all the catalysts. Table 3 shows the
crystallite size of NiO and CeO phases determined by Scherrer equation
2 4 2
O , CeO
◦
3
00 C corresponds to the combustion of the precursor to produce NiO
[
54]. For the NiNTA(x>0) systems differences are observed in the
2
number of weight loss events associated to decomposition of different
◦
◦
(
between brackets) using the peaks in 2θ = 43 and 28 , respectively.
◦
The broadening of the peaks in 2θ = 43 of the NiO phase shows that the
use of NTA promotes the formation of smaller NiO crystallites regardless
of the amount of used NTA respect to the NiL(0). In addition, NiNTA(1)
◦
and NiNTA(2) samples show a broadening of peak in 2θ = 28 corre-
2
sponding to CeO , this effect is less notorious in the NiNTA(0.5) system.
Table 2
Some properties of catalysts synthesized with NTA and CA.
2
Sample
Ni wt.
%
CeO
%
2
wt.
S
BET (m
d
pore
Vpore (mL
ꢀ 1
g )
ꢀ
1
g
)
(nm)
NiL(0)
5.7
6.5
6.3
8.2
6.8
5.0
9.0
ꢀ ꢀ
4.4
4.7
5.4
5.6
5.2
4.5
4.5
n.d.
88
95
92
99
91
80
94
4.9
4.7
5.8
5.0
5.0
5.4
4.9
4.4
0.14
0.14
0.17
0.15
0.13
0.14
0.15
0.08
NiNTA(0.5)
NiNTA(1)
NiNTA(2)
NiCA(0.5)
NiCA(1)
NiCA(2)
MgAl
2
O
4
ꢀ CeO
2
102
Fig. 2. Infrared spectra of dried catalysts.
n.d.= not determined.
5

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
Table 3
crystallite size follows the order: NiCA(2) < NiCA(0.5) < NiCA(1) < NiL
Crystallite size of reduced catalysts from Rietveld refinement.
(0). The CeO
2
crystallite size follows the order: NiCA(2) < NiCA(0.5) ~
XRD
Ni
XRD
NiO
XRD
CeO2
XRD
NiCA(1) < NiL(0). Li et al. [60] have reported a decrease in the size of
Sample
d
(nm)
d
(nm)*
d
(nm)*
d
(nm)
MgAl2O4
◦
◦
NiO crystallite and Ni in reduced Ni/SiO catalysts calcinated at 500 C
2
NiL(0)
4.0
2.2
4.0
2.0
2.6
3.3
2.1
13.0 (13.1)
5.9 (7.1)
6.6 (9.3)
6.6 (9.5)
4.3 (8.8)
7.4 (12.4)
6.5 (9.3)
6.0 (4.6)
6.1 (8.9)
2.7 (~ 4)
1.9 (5.5)
4.4 (6.0)
4.4 (6.3)
3.2 (7.5)
5.3
5.2
5.0
5.0
4.9
4.9
4.9
and prepared in the presence of citric acid. These authors have reported
that the smallest crystallite size was obtained using a molar CA/Ni ratio
of 1.5.
NiNTA(0.5)
NiNTA(1)
NiNTA(2)
NiCA(0.5)
NiCA(1)
The DR UV–Vis–NIR spectra of fresh catalysts, Fig. S6, shows that the
phase NiO is the predominant one on the catalysts and that it has not
been influenced by the chelating agent or the employed amount in the
synthesis. This behaviour is related to the stability of the support not
allowing the nickel spinel formation. The most intense band at 263 nm
NiCA(2)
*
Values between brackets correspond to crystallite size of fresh catalysts
obtained with Scherrer equation.
2
ꢀ
2+
would be associated to a process of charge transference between O
→
2
+
MgAl
2
O
4
ꢀ CeO
2
support is used without calcination, and it has been
dissolution occurs during impregnation step [31].
Ni [43]. Bands around 720 and 1190 nm would correspond to Ni
transitions with an octahedral environment. The bands at 583 and 624
nm associated to the presence of NiAl phase were not detected [61,
verified that a CeO
2
Similar results were reported by Santolalla-Vargas et al. [58] in the
preparation of W catalysts. The authors have noted a redispersion of W
on the surface when the impregnation solution containing CyDTA as a
2 4
O
2
+
62]. This indicates that all the catalytic systems exhibit Ni species with
a similar coordination environment regardless the addition of NTA or
CA.
chelating agent is mixed with dried γꢀ Al
2 3
O support promoting the
formation of different W species. 0.5 NTA mol per Ni mol would promote
From TPR profiles it is possible to identify the influence of amount of
chelating agent used in the synthesis on the reducibility of different
species, Fig. 4. All the TPR profiles are broader indicating the presence of
different species. From a comparison with some results reported in
literature on reducing systems containing Ni, Ce, Mg and/or Al, Table 4,
the complexation only of Ni2 ion while a higher amount of NTA (NTA:
+
3
+
Ni = 1 or 2) could also afford an interaction with Ce in the solution.
The redispersion of ceria on the NiNTA(1) and NiNTA(2) samples allows
its redeposition as amorphous particles or smaller crystallites. CeO
2
crystallite size reduction induces changes in its properties allowing an
effective interaction with nickel phases on the surface. Moreover, this
interaction between Ni and Ce promoted by the use of chelating agent
has started in the impregnation solution [30,59]. The use of the CA also
promotes the NiO crystallite size decrease and a slight effect in those of
the peaks in the TPR profiles could be assigned to the following species:
◦
α
species (between 200–400 C): CeO
2
surface reduction and/or NiO
amorphous species reduction;
◦
2+
β species (between 400–600 C): reduction of Ni species interacting
+
with CeO
γ species (between 600–800 C): reduction of Ni
interaction with MgAl
δ species: reduction of CeO
The γ species were only assigned to Ni ꢀ MgAl
to the absence of DR UV–Vis–NIR bands associated to NiAl
2
and/or Ni2 species slightly interacting with MgAl
2
O ;
4
2
+
◦
2+
with a high
CeO
and not with Ce3 , in agreement with that observed by TG analysis
Fig. S3).
X-ray diffractograms of reduced catalysts are shown in Fig. 3.
MgAl , CeO , Ni and NiO phases are identified in all diffraction pat-
2
, Table 3. This allows to infer that the CA rather interacts with Ni
+
2 4
O ;
◦
2
bulk above 750 C.
(
+
2
2
O
4
interactions due
, Fig. S6.
2
O
4
2
2 4
O
terns. The crystallite sizes of the phases are shown in Table 3.
The presence of NiO phase reveals that reduction during 45 min at
The reduction profile of NiNTA(2) is similar to NiL(0), although the γ
species reduce to higher temperatures. The same is observed in NiNTA
(0.5) sample. The sharp peak on the TPR profiles of systems prepared
◦
6
50 C under H
2
/N
2
flow was incomplete. The intensity of this peak in
◦
2
θ = 43 is high in the NiL(0), NiNTA(2) and NiCA(1) samples related to
with NTA is associated with β species indicating that this chelating agent
2
+
2+
the presence of major amount of Ni
species that are hard to be
promotes an intimate contact between Ni and CeO
2
. NiNTA(1) profile
◦
◦
◦
reduced. Between 50 and 53 a Ni peak is observed. Ni crystallite size
of catalysts prepared with NTA follows the order: NiNTA(2) ~ NiNTA
suggests a major contribution of β species in comparison with the other
NTA-systems. This redox behaviour was verified by performing a com-
◦
(
0.5) < NiL(0) ~ NiNTA(1). The CeO
2
crystallite size follows the order:
plementary TPR experiment up to a final temperature of 700 C on an
NiNTA(2) ~ NiNTA(1) < NiL(0) ~ NiNTA(0.5). The use of the molar
automatic Micromeritics AutoChem II equipment. The profile showed
the same behaviour, Fig. S7.
ratio NTA/Ni = 2 generates smaller Ni◦ and CeO
crystallites. XRD of
2
NiNTA(2) also shows a high intensity of NiO peak, probably by an excess
lso, NiCA(0.5) and NiCA(1) have a notorious presence of species with
◦
◦
◦
of nickel content in this sample. The use of CA leads to smaller Ni
intermediate reduction temperatures between 400 and 600 C. TPR
crystallite size respect to NiL(0), Table 3. In NiCA(x) systems, the Ni
profile of NiCA(2) is similar to NiL(0), but it is evidenced a major
Fig. 3. XRD patterns of reduced catalysts.
Fig. 4. TPR profiles of fresh catalysts.
6

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
Table 4
Table 4 (continued)
2
+
4+
Reduction of Ni and/or Ce species in the literature.
Catalytic system
TPR
Reduction events
Ref.
Catalytic system
Hydrotalcites
TPR
Reduction events
Ref.
conditions
conditions
The low temperature peak is
attributed to a solid solution
between Ni and Ce or small
NiO particles.
ꢀ 1
ꢀ
1
◦
33 mL min
10 C min
The incorporation of:
x Y
CeNi O
–
Ce produces a reduction of
2
H (5 %)/Ar
nickel spinel at lower
temperatures.
◦
ꢀ 1
10 C min
Peaks between:
2
+
350ꢀ 500 ◦C attributed to
2
–Ni into the CeO network
forms a solid solution allowing
reduction to lower
–
ꢀ
1
8
0 mL min
(5 %)/N
[49]
Ni-Mg-Al-Ce
amorphous NiO.
H
2
2
–500-520 ◦C attributed to
temperatures that overlap with
Ni(>9.1 wt.
ꢀ 1
crystal NiO.
4
0 mL min
(5%)/Ar
[70]
2
+
◦
the less interacted Ni
reduction peaks.
%)/γꢀ Al
2
O
3
–520-700 C attributed to
H
2
x y
NiAl O (a deficient-Ni-state
ꢀ 1
The reduction starts in the Ni-
compound).
◦
1
0
C min
◦
◦
–650-800 C attributed to
NiAl O .
2 4
Ce/MgAl
O
2 4
system at 200 C.
The surface reduction appears
5 ◦C min
ꢀ
1
◦
The peak at:
at temperatures above 350
with the increase of Ce.
The peak at:
C
–351 ◦C associated with
surface lattice oxygen removal
Ni(10 wt.%)ꢀ Ce (2.5
2+
◦
in CeO2.
–
550 C is attributed to Ni
species in interaction with
CeO
[59]
[63]
[64]
wt.%)/MgAl
2
O
4
Ni(2 wt.%)
ꢀ 447 ◦C associated to
H
2
(10 %)/Ar
ꢀ 1
@
SiO
2
@CeO
2
30 mL min
(3.2%)/N
2
[71]
reductions of Ni species that
have weak Ni-SiO2 interaction.
ꢀ 730 ◦C related to reduction
of Ni species that have strong
2
(
27.6%wt.%)
H
2
◦
–
780 C is attributed to
reduction of complex NiO
species that interact strongly
with MgAl
O
2 4
.
2
Ni-SiO interaction.
◦
ꢀ 1
1
0
C min
The peak at:
◦
–
450 C is attributed to
Ni(7 wt.
)/CeO
ꢀ 1
5
0 mL min
(5 %)/Ar
superficial reduction of CeO
2
.
contribution of β species. The use of NTA or CA produces the reduction
of γ species at higher temperature and promotes a major contribution of
β species that reduce at a lower temperature respect to NiL(0) regardless
the amount of chelating agent used.
%
2
ꢀ MgO
◦
H
2
ꢀ 315 C is attributed to
reduction of bulk NiO.
α
species assigned to the
reduction of bulk NiO or free
NiO without direct interaction
with support or species with a
very weak interaction.
Ni/MgO (2.5, 6.25, y
2.5 wt.%)ꢀ Al
ꢀ 1
◦
1
0
C min
The influence of the amount of NTA or CA on redox properties is also
studied by X-ray photoelectron spectroscopy. XPS spectra are shown in
1
2 3
O
◦
Fig. 5. Fresh catalysts were reduced ex situ at 650 C for 45 min in
β1 species are NiO species in a
Ni-riched mixed with greater
H
2
(5%)/N
2
flow. Before recording XPS spectra, the samples were in situ
◦
reduced at 400 C for 10 min in the pre-chamber of the equipment, so
interaction than
α
species.
Ni from Ni(II)ꢀ
ethylenediamine
en/Ni = 2)
ꢀ 1
◦
4
0 mL min
(5 %)/Ar
β2 species attributed to NiO in
a Al-riched mixed phase.
γ species assigned to the Ni
spinel strongly associated with
the spinel structure.
that the contribution by Ni oxidation in contact with the atmosphere
H
2
2+
can be considered negligible. The presence of Ni
in the reduced
(
samples is related to the used reduction conditions, in agreement with
what was observed in the XRD patterns, Fig. 3. Binding energy (BE)
values of the different peaks belonging to Ce 3d and Ni 2p are presented
The broad peak between 300
◦
ꢀ 1
◦
2+
◦
1
0
C min
and 400 C is attributed to
in Table S3. The BE values of Ni and Ni can be assigned near to 855
and 852 eV, respectively. A slight decrease in BE values of Ni 2p is
observed with the use of NTA and CA, Table S3. This decrease is an
reduction of NiO.
The small shoulder at high
temperatures corresponds to
the reduction of the aluminate
phase of Ni
Ni(1.5, 4.5, y 10 wt.
)/MgAl
[
65]
%
2
O
4
ꢀ 1
4
0 mL min
(10 %)/N
2
H
2
The peak at low temperatures
of large NiO particles without
interaction with the support.
◦
ꢀ 1
◦
Hydrotalcites
10 C min
The peak around 800 C is
ꢀ
1
attributed to the reduction of
NiO with strong interaction
3
0 mL min
(5 %)/N
[66]
Ni(>9 wt.%)/Mg/Al
Ni/xMg-ATP
x = 0,
H
2
2
2 3
with MgO and/or Al O .
◦
ꢀ 1
10 C min
Peaks between:
◦
–
300-500 C attributed to NiO
with a slight interaction with
γꢀ Al
2 3
O .
◦
ꢀ
1
–500 and 700 C attributed to
4
0 mL min
[67]
5
, 10, 20 wt.%
a solid nickel-aluminate
H
2
(10 %)/Ar
solution.
◦
–
700 and 900 C attributed to
ATP: attapulgite
NiO species with strong
interaction with γꢀ Al
2
O
3
.
◦
The broad peak at 265 C is
attributed to the reduction of
NiO that interacts weakly with
the spinel.
ꢀ
0 mL min
1
2
3
Ni(8 wt.%)/ MgAl
Mixed oxides
O
2 4
[68]
[69]
2
H (10 %)/N
◦
The peak at 710 C to small
NiO particles with strong
interaction with the support.
◦
ꢀ 1
10 C min
Fig. 5. XPS spectra of reduced catalysts.
7

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
indication that the addition of chelating agents increases the electronic
density of Ni species. A "memory effect" of Ni species seems to be present
generating different environments in the catalysts prepared from
chelating agents [44]. It has been reported that the decrease in BE of Ni
observed. As already mentioned, all catalysts were in situ reduced before
an experimental run. For catalysts prepared with NTA, the ethanol
conversions were higher than 80 %. In particular, the NiL(0) and NiNTA
(1) catalysts were the most active, presenting average conversions of 98
and 97 %, respectively, with a slight loss of activity with operating time.
The NiNTA(0.5) and NiNTA(2) samples were less active with average
conversions of 86 and 91 % and a loss of activity of about 11 % after 7 h
2
2p3/2 is related to higher interaction between Ni and CeO [62,72],
which it has been also identified with the increase in the atomic ratio
◦
Ce/Ni for catalysts prepared with chelating agents. In the Ce 3d region,
4
+
3+
peaks belonging to Ce and Ce are also observed. The presence of the
in operation. The main products were H
of CH , C O and C . Selectivity to H
yield, expressed as H
mol per ethanol ꢀ fed mol, follows the order:
NiNTA(0.5) (4.3) < NiNTA(2) (4.6) < NiL(0) (5) < NiNTA(1) (5.3 mol
2
, CO and CO
2
and less amount
3
+
4+
Ce /Ce redox couple participates in carbon removal reactions and its
presence increases tolerance to carbon deposition, Table 5. The values of
4
2
H
4
2
H
4
2
exceeded 70 %. Its average
2
3
+
4+
the Ce /Ce ratio are higher in the NiNTA(1) and NiNTA(2) systems,
being equal to 1.5, while for the sample prepared without chelating
ꢀ 1
H mol C H OH ). The NiL(0) system had the lowest selectivity to CO ,
2
2
5
2
agent or with 0.5 mol of NTA, the Ce3 /Ce ratio is 0.7. This increase in
the degree of surface reduction of ceria is associated with an improve-
ment in the oxygen mobility [73], due to the nanometric size of the ceria
+
4+
around 43 %, while those systems prepared with NTA reached values of
50 %. CO selectivity was higher for the chelating ꢀ free system, being
lower for NiNTA(0.5) and NiNTA(2) systems. For NiNTA(0) and NiNTA
and its better interaction with nickel. The higher Ce3 /Ce ratio values
+
4+
(1), the selectivity to minority products: CH
below 6 %; while in NiNTA(0.5) and NiNTA(2), the formation of these
products was higher with significant selectivities to CH and to C O.
In all systems, selectivity to C was lower than 1.5 %. The higher
selectivity of methane, of course, affects selectivity to H [74]. CH can
be formed by the decomposition of ethanol (C + CO + H ),
OH → CH
and/or by the methanation of CO and CO [75,76]. The catalytic results
of the NiNTA(2) system are unexpected, taking into account the small
4 2 4 2 4
, C H O and C H , were
correspond to systems with smaller crystallite sizes of CeO
2
, NiNTA(1)
3
+
4+
and NiNTA(2), Table 3. For the NiCA(1) sample, the Ce /Ce ratio is
4
2 4
H
+
4+
1
.0. As it was already mentioned, the presence of the Ce3 /Ce redox
2 4
H
couple participates in carbon removal reactions and its presence is key
for increasing carbon deposition tolerance.
2
4
2
H
5
4
2
◦
◦
The Ni /Ni surface ratio, considered as a measure of Ni dispersion,
increases with the increase of NTA in NiNTA(x) systems, Table 5.
However, it does not show a clear dependence with the amount of CA.
The obtained values for NiCA(x) are very similar to each other, in
agreement with the smaller differences in crystallite sizes determined by
XRD. It can be inferred that the addition of CA in the preparation pro-
2
◦
3+ 4+
size of the Ni and CeO and the high Ce /Ce ratio. The presence of a
2
higher amount of Ni in its oxidized form, at the beginning of the reaction
and after reduction (XRD of the reduced system, Fig. 3), possibly affects
the different stages in the reaction mechanism, promoting secondary
◦
◦
duces slight changes in Ni dispersion. The Ce/Ni surface ratio,
considered as a measure of the Ni-CeO interaction, changes with the
NTA/Ni molar ratio and follows the order of NiL(0) (1.26) < NiNTA(0.5)
1.70) < NiNTA(2) (1.82) < NiNTA(1) (2.80). The highest value is ob-
4 2 4 2 4
reactions of formation of CH , C H O and C H . The best catalytic
2
behaviour was observed for the NiNTA(1) catalyst, which corresponds to
the stoichiometric ratio of the most stable complex between nickel and
nitrilotriacetic acid. Although, in NiNTA(1) was evidenced a slight
deactivation that could be related to the type of carbon on the surface.
For all catalysts prepared with CA, the ethanol conversions were
higher than 90 %. The distribution of the products showed slight vari-
(
tained for the NiNTA(1), catalyst that presented a TPR sharp peak
related to the reduction of Ni2 species interacting with CeO
+
◦
(β spe-
2
cies). In the same way, the Ce/Ni surface ratio clearly increases by using
CA in preparation, following the order of NiL(0) (1.26) < NiCA(0.5)
ations. The major products were H
higher than 80 % in all the cases. The NiCA(2) system exhibited a
complete ethanol conversion during 7 h, while the other three CA-
systems showed a slight decrease in activity. Its H average yield fol-
lows the order: NiL(0) (5.0) < NiCA(0.5) (5.2) < NiCA(1) (5.5) < NiCA
2 2
, CO and CO, being the selectivity to
(
2.21) < NiCA(1) (2.27) < NiCA(2) (2.30).
H
2
Representative HRTEM images of NiL(0), NiNTA(1) and NiCA(2)
samples are shown in Fig. 6. From these images it is possible to identify
2
◦
different phases. These systems revealed the presence of Ni , NiO, CeO
2
ꢀ 1
and MgAl
different zones of the reduced catalysts show spots at 0.25–0.26 nm and
.21 nm corresponding to (1 1 1) and to (2 0 0) crystallographic planes
of NiO, respectively; spots at 0.25 ꢀ 0.26 nm and 0.20–0.21 nm corre-
sponding to (3 1 1) and to (4 0 0) planes of MgAl ; spots at 0.26 nm
2
O
4
crystalline phases. The Fourier Transform images in
(2) (5.6 mol H
the quantity of secondary products, such as CH
use of CA leads to a sightly increase in the amount of CH
were also observed in small quantities on NiCA(0.5) and NiCA(1).
For NiCA(1) a slight decay in CH production with an increase in C
was detected with time on stream, perhaps by a loss of activity in the
O decomposition reaction [8]. From the analysis of these results,
the NiCA(2) catalyst showed the best catalytic behaviour. This sample
2
2
mol C H
5
OH ). The most important differences were on
, C O and C . The
. C O and
4
2
H
4
2 4
H
0
4
2 4
H
2 4
C H
2
O
4
4
2 4
H O
and 0.30–0.33 nm corresponding to (2 0 0) and (1 1 1) planes of CeO
2
◦
and spots at 0.19–0.21 nm of (1 1 1) plane of Ni . In the images CeO
2
2 4
C H
particles with sizes less than 10 nm are observed in concordance with
the crystallite size obtained from XRD, Table 3.
◦
◦
presented the smallest Ni crystallite size and a high Ce/Ni ratio. Be-
sides, the CA:Ni molar ratio corresponded to the stoichiometric ratio to
form the most stable complex between nickel and citric acid.
3
.4. Ethanol steam reforming
A direct comparison of catalytic behaviour between Ni catalysts
tested in ESR is difficult due to not only catalyst features (metal per-
centage, support, modifiers, etc.) and synthesis conditions (different
precursor salts and preparation methods) but also to significant changes
in the reforming conditions (steam to carbon ratio, temperature, catalyst
weight, degree of bed dilution, percentage of ethanol in the feed and
activation conditions). Bepari et al. [49] have studied a series of
cerium-promoted Ni-Mg-Al hydrotalcite catalysts in the ESR. The best
All catalysts were tested in ethanol steam reforming reaction, Fig. 7.
Under the operation conditions (without bed dilution and yC2H5OH = 9.2
) used in this study, an early deactivation on the catalysts could be
%
Table 5
Surface atomic ratio from XPS of reduced catalysts.
3
+/ 4+
0
2+
0
0
Sample
Ce Ce
Ni /Ni
Ce/Ni
Ni /Ni
Ni/Ʃ*
Ce/Ʃ*
◦
performance was obtained at 540 C (operation conditions: 3.0 gcat, S/C
NiL(0)
0.7
0.7
1.5
1.5
0.6
1.0
0.4
0.5
0.3
0.5
0.4
0.4
0.5
0.3
1.26
1.70
2.80
1.82
2.21
2.27
2.30
0.33
0.25
0.34
0.42
0.29
0.32
0.24
0.09
0.12
0.07
0.13
0.09
0.06
0.11
0.04
0.05
0.06
0.09
0.06
0.04
0.06
ꢀ 1
ratio = 4.5 and space-time of 22.04 kgcath kmol of ethanol fed) for the
NiNTA(0.5)
NiNTA(1)
NiNTA(2)
NiCA(0.5)
NiCA(1)
5
.5 % Ni and 10.5 % Ce catalyst. Under this condition, about 97 % of
ethanol conversion was obtained with a H
2
yield of 4.3 mol H
2
.mol
ꢀ 1
C
2
H
5
OH , also were detected CO
2
, CO and CH
4
as products. Słowik
et al. [19] prepared a Ni/CeO
2
catalyst using an aqueous solution of
NiCA(2)
◦
nickel nitrate with citric acid CA (Ni/CA = 1), calcined at 420 C and
*
◦
Ʃ= Ni + Ce + Al + Mg.
reduced with hydrogen at 420 C for 1 h. This catalyst was 100 % active
8

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
Fig. 6. HRTEM images of reduced catalysts.
in the ESR (operation conditions: 0.1 gcat with dilution and S/C ratio =
small weight loss of about 2%, in agreement with those observed by
XRD, Table 6. The other NTA catalysts show 32 and 35 carbon % for the
NTA/Ni ratios of 0.5 and 2, respectively. The NiNTA(1) system showed
◦
6
) at 420 C for 30 h, the main products were H
2
(68–75%), CO
2 4
(41 to 35 %) with traces of CO and C H O.
2
(
58–60%) and CH
4
2
larger Ni crystallite sizes than the other ones but small CeO crystallite
3
.5. Spent catalysts
size, which enhances its oxygen release and storage properties [29] and
◦
this would allow a cooperative work with Ni . In addition, it has a high
3
+
4+
◦
The diffraction patterns of used catalysts are shown in Fig. 8.
Ce /Ce surface ratio and the highest Ce/Ni ratio, Table 5. The
properties of ceria nanoparticles on this catalyst could explain the high
removal carbon capacity during ESR. The smaller crystallite size of ceria
promotes a high oxygen vacancy and mobility that would directly
0
MgAl
2 4 2
O , CeO , NiO and Ni phases are detected. The spinel and ceria
are stable under the reaction conditions. The characteristic peak at 2θ =
5
1.8 belonging to Ni0 is clearly detected, showing that the systems
◦
3
+
4+
remained active during the reaction time. This peak remains broad
indicating that the nickel crystallites are nano-sized even under reaction
conditions. However, it is not possible to dismiss a slight sintering of the
active phase. The presence of residual NiO is also observed in 2θ = 43.1◦
mainly in NiNTA(0.5), NiNTA(2) and NiCA(1) systems as a consequence
of incomplete reduction, although no reoxidation of the active phase is
ruled out. The intensity of this peak is lower than that observed in fresh
samples (Fig. S5), suggesting that the reduction has continued under the
reactive atmosphere. NiNTA(1), NiCA(0.5) and NiCA(2) reveal less re-
sidual amount of NiO, which would indicate that these systems would
contribute with the gasification of carbon. The high Ce /Ce ratio
would be related with more oxygen vacancy. Wang et al. indicated that
small ceria nanoparticles promote a better performance in gasification of
◦
formaldehyde at 250 C [77] and this was attributed to the presence of
more oxygen vacancy in the ceria nanoparticles, with a crystalline size
less than 5 nm. In the NiNTA(0.5) system, the amount of carbon can be
attributed to differences in the degree of interaction between phases Ni
2
and CeO , and partially related to the metal particle size. The reduced
NiNTA(0.5) system presented the largest crystallite size of CeO
2
(Table 3) and low Ce3 /Ce and Ce/Ni . In addition, the XRD for the
NiNTA(0.5) and NiNTA(2) systems reveal the presence of residual NiO,
which would affect their performance in the reaction under the opera-
tion conditions used in this work. Maximum combustion temperatures of
carbon deposits are lower in NTA-prepared catalysts, Table 6. Lower
combustion temperatures are advantageous when considering catalyst
regeneration procedures. For the NiNTA(1) system, the temperature
+
4+
◦
have Ni2 species easier to be reduced. The intensity of the characteristic
+
◦
graphitic carbon peak at 2θ = 26 is higher for the NiL(0), NiNTA(0.5),
NiNTA(2) and for NiCA(x>0) systems and it is almost undetectable to
the NiNTA(1) system.
The thermograms of the catalysts used are shown in Fig. S8. The
weight loss of the samples under an oxidizing atmosphere mostly cor-
responds to the combustion of carbon deposits formed during the
reforming reaction. It can be observed that the NiNTA(1) system has a
◦
decrease is higher than 100 C.
In the CA ꢀ prepared systems the percentage of carbon deposited is
9

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
Fig. 7. Ethanol conversion and product distribution in ethanol steam reforming during 7 h. T = 650 ◦C, S/C = 2.5, yC2H5OH = 9.2 %.
1
0

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
using Lorentzian curves for G, D1, D2 and D4 contributions and using
ꢀ 1
Gaussian curve for D3 contribution. The G band around 1580 cm is
2
associated with the E2g optical mode of graphite or sp carbon materials
[
80]. The D bands are characteristic of disordered graphite and their
intensity relative to the G band increase with a rising degree of disorder
ꢀ 1
in the graphitic structure [79]. The D1 band around 1340 cm corre-
sponds to a graphitic lattice vibration mode with A1g symmetry [79].
The ID1/I
G
ratio increases with the amount of disorder [81]. The band
intensity
positions and Raman ratios are shown in Table 7. The ID1/I
G
ratio were higher for the NiNTA(x) and NiCA(x) systems. Then, it could
be inferred that the use of chelating agents (NTA or CA) increases the
amount of disorder on the carbonaceous deposits, and it is also related to
decrease in the combustion temperature, Table 6, indicating that the
carbon with a major disorder burns at less temperature. A shift in G band
positions is observed when NTA or CA were used in the catalyst prep-
aration, suggesting smaller crystallite sizes of the formed graphite-like
carbon [82].
Fig. 8. XRD patterns of spent catalysts after 7 h of ethanol steam reforming.
The carbon deposition was also examined by SEM. All SEM images of
the catalysts show two zones, Fig. S9: (i) one with almost negligible
amount of carbon deposits and (ii) other with an abundant deposition of
carbon [83–85]. This clearly reveals that carbon deposition is not ho-
mogeneous regardless of the precursor used in the synthesis of catalysts.
SEM images of NiNTA(0.5) and NiNTA(2) show zones with an abundant
amount of filaments of different sizes and carbon free regions. For
NiNTA(1), the carbon deposits appear to be amorphous carbon type, in
Table 6
Carbon amount and maximum combustion temperature from TGA-TPO of cat-
alysts after 7 h in reaction.
◦
ꢀ 1 ꢀ 1
mgC gcat
Sample
T
max ( C)
%C
h
NiL(0)
702
633
537
630
644
639
653
22
32
~ 2
35
20
29
28
30.6
45.6
~ 1.4
47.3
28.6
37.4
37.7
NiNTA(0.5)
NiNTA(1)
NiNTA(2)
NiCA(0.5)
NiCA(1)
G G
agreement with the high ID3/I ratio, Table 7 [79]. The ID3/I ratios for
NiCA(x>0) systems are also high, however only filamentous-type car-
bon deposits were observed by SEM. For NiCA(1) catalyst, the carbon
filaments seem to be longer. The carbon type is also related to the
presence of secondary products.
NiCA(2)
Nickel signal is clearly detected by EDX in the carbon regions for all
catalysts, which allows to infer that the metal particle could be at the
end of the filaments, which is very common in nickel catalysts. The
presence of areas with and without carbon makes it possible to assume
that deposition is associated with the presence of metal particles and
that the bare support does not mostly contribute to the deposition
mechanism or the occurrence of secondary reactions.
similar in all the samples. The NiCA(x>0) samples have similar prop-
3
+
erties, the main differences are related to the XPS surface ratios (Ce
/
4
+
◦
Ce and Ce/Ni ratio), Table 5. However, differences in combustion
temperature are observed, associated with graphitization degree of
carbon deposits. This would allow the use of a lower temperature (in the
◦
order of 50 C less) when implementing a regeneration strategy.
The Raman spectra of NiNTA(x) and NiCA(x) systems are shown in
Fig. 9. The Raman spectra of the first-order region correspond to
carbonaceous zone on the catalysts [78]. These spectra were deconvo-
lutioned in five contributions according to what has been reported by
Sadezky et al. [79]. The deconvolution of the spectra was performed
4
. Conclusions
The catalytic performance of a series of Ni/MgAl
2
O
4
ꢀ CeO
2
catalysts
were tested in ethanol steam reforming reaction. The catalysts were
prepared by the wet impregnation method using L: NTA or CA as
chelating agents. The influence of different L/Ni molar ratios was
studied.
In all dried samples the presence of some complex species L-Ni on the
2
+
support surface was verified. However, in the calcined samples, Ni
species were in an octahedral environment similar to NiO without a
significant amount of carbon residues from the chelating agent. The
changes in surface properties of catalysts were associated to a memory
effect.
2 2
Improved interactions among NiO-CeO and Ni-CeO were obtained
by the use of chelating agents. These interactions were revealed by
Table 7
Raman band positions and ID1/IG ratio for used catalysts.
ꢀ
1
Raman shift (cm
)
Sample
I
D1/I
G
ID3/I
G
G
D1
D2
D3
D4
NiL(0)
1576
1580
1581
1590
1582
1587
1585
1340
1339
1338
1347
1338
1346
1339
1603
1604
1604
1614
1604
1613
1606
1508
1522
1528
1530
1529
1530
1529
1200
1191
1194
1217
1209
1218
1212
1.63
1.70
1.96
1.76
1.90
1.90
1.85
0.53
0.80
1.29
0.84
1.25
0.34
0.58
NiNTA(0.5)
NiNTA(1)
NiNTA(2)
NiCA(0.5)
NiCA(1)
NiCA(2)
Fig. 9. Raman spectra of used catalysts after 7 h in ethanol steam reforming.
1
1

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
HRTEM images, the behaviour under reduction conditions and the in-
[5] C. Montero, A. Remiro, B. Valle, L. Oarꢀ Arteta, J. Bilbao, A.G. Gayubo, Ind. Eng.
Chem. Res. 58 (2019) 14736–14751, https://doi.org/10.1021/acs.iecr.9b02880.
0
crease in the Ce/Ni surface ratio.
[
6] J.R. Rostrupꢀ Nielsen, J. Sehested, J. Nørskov, Adv. Catal. 47 (2002) 65–139,
https://doi.org/10.1016/S0360ꢀ 0564(02)47006ꢀ X.
The NTA addition in the preparation of catalytic systems allowed to
decrease the NiO crystallite sizes. After reduction, the Ni0 crystallite
sizes were smaller for NiNTA(0.5) and NiNTA(2) systems. Due to the
redissolution of the ceria during impregnation with solutions containing
[7] Y.C. Sharma, A. Kumar, R. Prasad, S.N. Upadhyay, Renew. Sust. Energy Rev. 74
(
2017) 89–103, https://doi.org/10.1016/j.rser.2017.02.049.
8] L.V. Mattos, G. Jacobs, B.H. Davis, F.B. Noronha, Chem. Rev. 112 (2012)
094–4123, https://doi.org/10.1021/cr2000114.
[9] A.J. van Dillen, R. Terorde, D. Lensveld, J. Geus, K. de Jong, J. Catal. 216 (2003)
57–264, https://doi.org/10.1016/S0021ꢀ 9517(02)00130ꢀ 6.
10] U.C. Abubakar, K.R. Alhooshani, T.A. Saleh, J. Environ. Chem. Eng. 8 (2020),
03811, https://doi.org/10.1016/j.jece.2020.103811.
[
4
2
+
2
Ni and NTA, the crystallite size of CeO was also affected, mainly with
2
a molar ratio NTA/Ni =1 and 2.
[
The CA addition in the preparation of Ni catalysts promoted the
1
0
formation of smaller NiO and Ni crystallite sizes, with a slight effect on
[11] J. Escobar, M.C. Barrera, J.A. de los Reyes, J.A. Toledo, V. Santes, J.A. Colin,
J. Mol. Catal. A Chem. 287 (2008) 33–40, https://doi.org/10.1016/j.
molcata.2008.02.022.
2
the size of CeO particles.
The catalysts prepared with a molar ratio NTA/Ni=1 and CA/Ni=2
[
[
[
12] T. Shimizu, K. Hiroshima, T. Honma, T. Mochizuki, M. Yamada, Catal. Today 45
(1998) 271–276, https://doi.org/10.1016/S0920ꢀ 5861(98)00227ꢀ 2.
13] T. Mochizuki, T. Hara, N. Koizumi, M. Yamada, Appl. Catal. A Gen. 317 (2007)
97–104, https://doi.org/10.1016/j.apcata.2006.10.005.
exhibited better catalytic performances in the ethanol reforming reac-
tion under the operation conditions used. The NiNTA(1) and NiCA(2)
system presented high Ce/Ni0 surface ratios. The NiNTA(1) showed a
14] J.ꢀ S. Girardon, E. Quinet, A. Gribovalꢀ Constant, P.A. Chernavskii, L. Gengembre,
A.Y. Khodakov, J. Catal. 248 (2007) 143–157, https://doi.org/10.1016/j.
jcat.2007.03.002.
high tolerance to carbon deposition while no improvement in tolerance
to carbon deposition was observed for NiCA(2). The use of chelating
agents in catalysts preparation induced changes in the type of carbon.
The carbon removal through oxidation in catalysts prepared using NTA
or CA requires lower temperatures, which is beneficial for implementing
a regeneration strategy. Although, all the NiL(x) systems showed a good
catalytic performance in ethanol steam reforming, under the experi-
mental conditions used, the best performance in terms of hydrogen
production were obtained with NiNTA(1) and NiCA(2) systems. These
systems were prepared using the stoichiometric molar ratios to form the
most stable complex in solution between nickel and chelating agent.
[
15] S. Karnjanakom, G. Guan, B. Asep, X. Dua, X. Hao, C. Samart, A. Abudula, Energy
Conv. Manag. 98 (2015) 359–368, https://doi.org/10.1016/j.
enconman.2015.04.007.
[16] X. Lu, J.ꢀ F. Chen, Y. Tan, Y. Zhang, Catal. Comm. 20 (2012) 6–11, https://doi.org/
1
0.1016/j.catcom.2012.01.002.
[
17] Y. Bang, S. Park, S.ꢀ J. Han, J. Yoo, J.ꢀ H. Song, J.H. Choi, K.H. Kang, I.K. Song,
Appl. Catal. B: Environ. 180 (2016) 179–188, https://doi.org/10.1016/j.
apcatb.2015.06.023.
[
[
[
[
[
[
18] K. Alꢀ Dalama, B. Aravind, A. Stanislaus, Appl. Catal. A Gen. 296 (2005) 49–53,
https://doi.org/10.1016/j.apcata.2005.07.041.
19] G. Słowik, M. Greluk, M. Rotko, A. Machocki, Appl. Catal. B: Environ. 221 (2018)
490–509, https://doi.org/10.1016/j.apcatb.2017.09.052.
20] C. Wu, P.T. Williams, Appl. Catal. B: Environ. 102 (2011) 251–259, https://doi.
org/10.1016/j.apcatb.2010.12.005.
CRediT authorship contribution statement
21] J. Ryczkowski, W. Grzegorczyk, D. Nazimek, Appl. Catal. A Gen. 126 (1995)
3
41–349, https://doi.org/10.1016/0926ꢀ 860X(95)00032ꢀ 1.
22] G. Anderegg, Pure Appl. Chem. 54 (1982) 2693–2758, https://doi.org/10.1351/
pac198254122693.
Alejandra C. Villagran¡Olivares: Investigation, Formal analysis,
Writing - original draft, Visualization. Mariana N. Barroso: Conceptu-
alization, Validation. Carlos A. L o´ pez: Formal analysis, Validation.
Jordi Llorca: Validation. María C. Abello: Conceptualization, Valida-
tion, Visualization, Writing - review & editing.
23] W. Hummel, G. Anderegg, L. Rao, I. Puigdom `e nech, O. Tochiyama, Chemical
Thermodynamics of Compounds and Complexes of U, Np, Pu, Am, Tc, Se, Ni and Zr
With Selected Organic Ligands, Elsevier Science, USA, 2005.
[
[
[
24] A.E. Galetti, M.F. Gomez, L.A. Arrúa, M.C. Abello, Appl. Catal. A: General 348
(
2008) 94–102, https://doi.org/10.1016/j.apcata.2008.06.039.
25] A. Trovarelli, J. Llorca, ACS Catal. 7 (2017) 4716–4735, https://doi.org/10.1021/
acscatal.7b01246.
Declaration of Competing Interest
26] L. Soler, A. Casanovas, J. Ryan, I. Angurell, C. Escudero, V. P ´e rezꢀ Dieste, J. Llorca,
ACS Catal. 9 (2019) 3641–3647, https://doi.org/10.1021/acscatal.9b00463.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
[27] T. Montini, M. Melchionna, M. Monai, P. Fornasiero, Chem. Rev. 116 (2016)
5987–6041, https://doi.org/10.1021/acs.chemrev.5b00603.
[
28] F. Wang, Y. Wang, L. Zhang, J. Zhu, B. Han, W. Fan, L. Xu, H. Yu, W. Cai, Z. Li,
Z. Deng, W. Shi, Catal. Today 355 (2020) 502–511, https://doi.org/10.1016/j.
cattod.2019.06.067.
[29] A. Elmhamdi, R. Casta
n˜ eda, A. Kubacka, K. Nahdi, A. Martínez, Appl. Catal. B:
Acknowledgments
Environ. 188 (2016) 292–304, https://doi.org/10.1016/j.apcatb.2016.02.011.
[
30] A.C. Villagr a´ nꢀ Olivares, M.F. Gomez, M.N. Barroso, M.C. Abello, Mol. Catal. 481
Financial supports are acknowledged to CONICET (PIP 00327/CO
and PIP 00227/CO) and Universidad Nacional de San Luis (PROICO
(
2020), 110164, https://doi.org/10.1016/j.mcat.2018.08.006.
[31] A.C. Villagr ´a nꢀ Olivares, M.F. Gomez, C. L o´ pez, M.N. Barroso, M.C. Abello, Appl.
Catal. B: Environ. 264 (2020), 118510, https://doi.org/10.1016/j.
apcatb.2019.118510.
32] B.A.T. Mehrabadi, S. Eskandari, U. Khan, R.D. White, J.R. Regalbuto, Adv. Catal.
61 (2017) 1–35, https://doi.org/10.1016/bs.acat.2017.10.001.
33] F. Wang, K. Han, L. Xu, H. Yu, W. Shi, Ind. Eng. Chem. Res. 60 (2021) 3324–3333,
https://doi.org/10.1021/acs.iecr.0c06020.
2
ꢀ 0116). The authors are grateful to ANPCyT for Grant PME 8ꢀ 2003 to
finance the purchase of the UHV Multi Analysis System. JL is a Serra
Húnter fellow and is grateful to ICREA Academia program and proj-
ectsMICINN/FEDER RTI2018-093996-B-C31 and GC 2017 SGR 128. The
authors would like to thank to Eng. Manuel F. Gomez for his assistance
in ESR experiments.
[
[
[
[
[
[
34] H.M. Rietveld, J. Appl. Crystallogr. 2 (1969) 65–71, https://doi.org/10.1107/
S0021889869006558.
35] J. Rodríguezꢀ Carvajal, Phys. B Condens. Matter 192 (1993) 55–69, https://doi.
org/10.1016/0921ꢀ 4526(93)90108ꢀ I.
Appendix A. Supplementary data
36] P. Thompson, D.E. Cox, J.B. Hastings, J. Appl. Crystallogr. 20 (1987) 79–83,
https://doi.org/10.1107/S0021889887087090.
37] J.R.H. Ross, Chapter 5 ꢀ catalyst characterization, in: J.R.H. Ross (Ed.),
Contemporary Catalysis, Elsevier, 2019, pp. 121–132, https://doi.org/10.1016/
b978ꢀ 0ꢀ 444ꢀ 63474ꢀ 0.00005ꢀ 9. ISBN 9780444634740.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.apcata.2021.118219.
[
[
[
[
38] I. Puigdomenech, Chem. Equilibr. Diagrams 2 (2013) https://www.kth.se/en/che/
medusa/downloadsꢀ 1.386254.
References
39] L. Bonneviot, O. Legendre, M. Kermarec, D. Olivier, M. Che, J. Colloid Interf. Sci.
1
34 (1990) 534–547, https://doi.org/10.1016/0021ꢀ 9797(90)90161ꢀ G.
40] G. George, S. Anandhan, RSC Adv. 4 (2014) 62009–62020, https://doi.org/
0.1039/C4RA11083H.
41] V.A. Su ´a rezꢀ Toriello, C.E. Santolallaꢀ Vargas, J.A. de los Reyes, A. V a´ zquez-
Zavala, M. Vrinat, C. Geantet, J. Mol. Catal. A Chem. 404–405 (2015) 36–46,
https://doi.org/10.1016/j.molcata.2015.04.005.
42] J. Calvacheꢀ Mu n˜ oz, F.A. Prado, J.E. Rodríguezꢀ P ´a ez, Colloids Surf. A
Physicochem. Eng. Asp. 529 (2017) 146–159, https://doi.org/10.1016/j.
colsurfa.2017.05.059.
[
[
[
1] S. Ogo, Y. Sekine, Fuel Process. Technol. 199 (2020), 106238, https://doi.org/
0.1016/j.fuproc.2019.106238.
2] Q. Zhang, K. Long, J. Wang, T. Zhang, Z. Song, Q. Lin, Int. J. Hydrog. Energy 42
1
1
(
2017) 14103–14114, https://doi.org/10.1016/j.ijhydene.2017.04.090.
3] J.L. Contreras, J. Salmones, J.A. Colıínꢀ Luna, L. Nu n˜ o, B. Quintana, I. C o´ rdova,
B. Zeifert, C. Tapia, G.A. Fuentes, Int. J. Hydrog. Energy 39 (2014), 18835, https://
doi.org/10.1016/j.ijhydene.2014.08.072.
[
[
4] S. Bepari, D. Kuila, Int. J. of Hydrog. Energy 45 (2020) 18090–18113, https://doi.
org/10.1016/j.ijhydene.2019.08.003.
1
2

A.C. Villagranꢀ Olivares et al.
Applied Catalysis A, General 622 (2021) 118219
[
[
[
43] Z. Yu, X. Hu, P. Jia, Z. Zhang, D. Dong, G. Hu, S. Hu, Y. Wang, J. Xiang, Appl. Catal.
B: Environ. 237 (2018) 538–553, https://doi.org/10.1016/j.apcatb.2018.06.020.
44] K.ꢀ Q. Sun, E. Marceau, M. Che, Phys. Chem. Chem. Phys. 8 (2006) 1731–1738,
https://doi.org/10.1039/B513319J.
45] R.A. Ortegaꢀ Domínguez, H. Vargasꢀ Villagr ´a n, C. Pe n˜ alozaꢀ Orta,
K. Saavedraꢀ Rubio, X. Bokhimi, T.E. Klimova, Fuel 198 (2017) 110–122, https://
doi.org/10.1016/j.fuel.2016.12.037.
[65] A. Di Michele, A. Dell’Angelo, A. Tripodi, E. Bahadori, F. S ´a nchez, D. Motta,
N. Dimitratos, I. Rossetti, G. Ramis, Int. J. Hydrog. Energy 44 (2019) 952–964,
https://doi.org/10.1016/j.ijhydene.2018.11.048.
[66] M. Li, X. Wang, S. Li, S. Wang, X. Ma, Int. J. Hydrog. Energy 35 (2010) 6699–6708,
https://doi.org/10.1016/j.ijhydene.2010.04.105.
[67] M. Chen, Y. Wang, Z. Yang, T. Liang, S. Liu, Z. Zhou, X. Li, Fuel 220 (2018) 32–46,
https://doi.org/10.1016/j.fuel.2018.02.013.
[
46] N. Koizumi, Y. Ibi, D. Hongo, Y. Hamabe, S. Suzuki, Y. Hayasaka, T. Shindo,
M. Yamada, J. Catal. 289 (2012) 151–163, https://doi.org/10.1016/j.
jcat.2012.02.003.
[68] S.C.M. Mizuno, A.H. Braga, C.E. Hori, J.B.O. Santos, J.M.C. Bueno, Catal. Today
296 (2017) 144–153, https://doi.org/10.1016/j.cattod.2017.04.023.
[69] L. Jalowieckiꢀ Duhamel, C. Pirez, M. Capron, F. Dumeignil, E. Payen, Int. J.
Hydrog. Energy 35 (2010) 12741–12750, https://doi.org/10.1016/j.
ijhydene.2009.08.080.
[70] R. Yang, X. Li, J. Wu, X. Zhang, Z. Zhang, Y. Cheng, J. Guo, Appl. Catal. A Gen. 368
(2009) 105–112, https://doi.org/10.1016/j.apcata.2009.08.021.
[71] K. Han, W. Yu, L. Xu, Z. Deng, H. Yu, F. Wang, Fuel 291 (2021), 120182, https://
doi.org/10.1016/j.fuel.2021.120182.
[72] A. Iriondo, V.L. Barrio, J.F. Cambra, P.L. Arias, M.B. Guemez, M.
C. Sanchezꢀ Sanchez, R.M. Navarro, J.L.G. Fierro, Int. J. Hydrog. Energy 35 (2010)
11622–11633, https://doi.org/10.1016/j.ijhydene.2010.05.105.
[73] E. Varga, Z. Ferencz, A. Oszk o´ , A. Erd o˝ helyi, J. Kiss, J. Mol. Catal. A Chem. 397
(2015) 127–133, https://doi.org/10.1016/j.molcata.2014.11.010.
[74] J.W.C. Liberatori, R.U. Ribeiro, D. Zanchet, F.B. Noronha, J.M.C. Bueno, Appl.
Catal. A Gen. 327 (2007) 197–204, https://doi.org/10.1016/j.apcata.2007.05.010.
[75] J. Chen, D. Xu, Int. J. Sustain. Green Energy 6 (2017) 28–38, https://doi.org/
10.11648/j.ijrse.20170603.11.
[
[
[
[
[
[
[
[
[
[
47] L. Espinosaꢀ Alonso, K.P. de Jong, B.M. Weckhuysen, J. Phys. Chem. C 112 (2008)
7
201–7209, https://doi.org/10.1021/jp710676v.
48] L.Y. Wang, G.Q. Wu, D.G. Evans, Mater. Chem. Phys. 104 (2007) 133–140, https://
doi.org/10.1016/j.matchemphys.2007.02.098.
49] S. Bepari, S. Basu, N. Pradhan, A.K. Dalai, Catal. Today 291 (2017) 47–57, https://
doi.org/10.1016/j.cattod.2017.01.027.
50] G. S o´ crates, Infrared and Raman Characteristic Group Frequencies: Tables and
Charts, 3rd edition, J. Wiley and Sons, 2001. ISBN: 0ꢀ 471ꢀ 85298ꢀ 8.
51] O. Gyliene, J. Aikaite, O. Nivinskiene, J. Hazard. Mater. 109 (2004) 105–111,
https://doi.org/10.1016/j.jhazmat.2004.03.008.
52] K.A. Tarasov, D. O’Hare, V.P. Isupov, Inorg. Chem. 42 (2003) 1919–1927, https://
doi.org/10.1021/ic0203926.
53] F. Wypych, G.G. Carbajal Arízaga, J.E. Ferreira da Costa Gardolinski, J. Colloid
Interf. Sci. 283 (2005) 130–138, https://doi.org/10.1016/j.jcis.2004.08.125.
54] W. Brockner, C. Ehrhardt, M. Gjikaj, Thermochim. Acta 456 (2007) 64–68, https://
doi.org/10.1016/j.tca.2007.01.031.
[76] F. Haga, T. Nakajima, H. Miya, S. Mishima, Catal. Lett. 48 (1997) 223–227,
https://doi.org/10.1023/A:1019039407126.
55] G. Leofanti, M. Padovan, G. Tozzola, B. Venturelli, Catal. Today 41 (1998)
2
07–219, https://doi.org/10.1016/S0920ꢀ 5861(98)00050ꢀ 9.
[77] Y. Wang, F. Wang, K. Han, W. Shi, H. Yu, Catal. Sci. Technol. 10 (2020)
6701–6712, https://doi.org/10.1039/D0CY00988A.
56] S.A. Theofanidis, V.V. Galvita, H. Poelman, R. Batchu, L.C. Buelens, C. Detavernier,
G.B. Marin, Appl. Catal. B: Environ. 239 (2018) 502–512, https://doi.org/
[78] M. Pawlyta, J.ꢀ N. Rouzaud, S. Duber, Carbon 84 (2015) 479–490, https://doi.org/
1
0.1016/j.apcatb.2018.08.042.
10.1016/j.carbon.2014.12.030.
[
[
57] M. Zarei, F. Meshkani, M. Rezaei, Adv. Powder Technol. 27 (2016) 1963–1970,
https://doi.org/10.1016/j.apt.2016.06.028.
58] C.E. Santolallaꢀ Vargas, V.A. Su ´a rez Toriello, J.A. de los Reyes, D.K. Cromwell,
B. Pawelec, J.L.G. Fierro, Mater. Chem. Phys. 166 (2015) 105–115, https://doi.
org/10.1016/j.matchemphys.2015.09.033.
[79] A. Sadezky, H. Muckenhuber, H. Grothe, R. Niessner, U. P o¨ schl, Carbon 43 (2005)
1731–1742, https://doi.org/10.1016/j.carbon.2005.02.018.
[80] J. McGregor, Z. Huang, E.P.J. Parrott, J.A. Zeitler, K.L. Nguyen, J.M. Rawson,
A. Carley, T.W. Hansen, J.ꢀ P. Tessonnier, D.S. Su, D. Teschner, E.M. Vass,
A. Knopꢀ Gericke, R. Schl o¨ gl, L.F. Gladden, J. Catal. 269 (2010) 329–339, https://
doi.org/10.1016/j.jcat.2009.11.016.
[
[
[
[
59] K.Y. Koo, S.ꢀ h. Lee, U.H. Jung, H.ꢀ S. Roh, W.L. Yoon, Fuel Process. Technol. 119
(
2014) 151–157, https://doi.org/10.1016/j.fuproc.2013.11.005.
[81] L. Bokobza, J.ꢀ L. Bruneel, M. Couzi, Carbon 1 (2015) 77–94, https://doi.org/
60] B. Li, X. Qian, X. Wang, Int. J. Hydrog. Energy 40 (2015) 8081–8092, https://doi.
org/10.1016/j.ijhydene.2015.04.104.
10.3390/c1010077.
[82] H.ꢀ Z. Wang, L.ꢀ L. Sun, Z.ꢀ J. Sui, Y.ꢀ A. Zhu, G.ꢀ H. Ye, D. Chen, X.ꢀ G. Zhou,
W.ꢀ K. Yuan, Ind. Eng. Chem. Res. 57 (2018) 8647–8654, https://doi.org/
10.1021/acs.iecr.8b01313.
[83] A.C. Villagr ´a n Olivares, M.N. Barroso, M.F. Gomez, M.C. Abello, Int. J. Ind. Chem.
9 (2018) 61–73, https://doi.org/10.1007/s40090ꢀ 018ꢀ 0135ꢀ 6.
[84] M.N. Barroso, M.F. Gomez, L.A. Arrúa, M.C. Abello, Int. J. Hydrog. Energy 39
(2014) 8712–8719, https://doi.org/10.1016/j.ijhydene.2013.12.043.
[85] A.E. Galetti, M.F. Gomez, L.A. Arrúa, M.C. Abello, Appl. Catal. A Gen. 408 (2011)
78–86, https://doi.org/10.1016/j.apcata.2011.09.006.
61] R. Jalali, B. Nematollahi, M. Rezaei, M. Baghalha, Int. J. Hydrog. Energy 44 (2019)
1
0427–10442, https://doi.org/10.1016/j.ijhydene.2018.12.186.
62] S. Damyanova, B. Pawelec, R. Palcheva, Y. Karakirova, M.C. Capel Sanchez,
G. Tyuliev, E. Gaigneaux, J.L.G. Fierro, Appl. Catal. B: Environ. 225 (2018)
3
40–353, https://doi.org/10.1016/j.apcatb.2017.12.002.
[
[
63] J.A. Santander, G.M. Tonetto, M.N. Pedernera, E. L o´ pez, Int. J. Hydrog. Energy 42
(
2017) 9482–9492, https://doi.org/10.1016/j.ijhydene.2017.03.169.
64] A. Ebadi, S. Tourani, F. Khorasheh, Ind. Eng. Chem. Res. 57 (2018) 12700–12714,
https://doi.org/10.1021/acs.iecr.8b01878.
1
3