Nano bimetallic alloy of Ni-Co obtained from LaCoxNi1-xO3 and its catalytic performance for steam reforming of ethanol

Article abstract of DOI:10.1039/c4ra14131h

A novel preparation method for supported bimetallic alloy nanoparticles was investigated in this work. LaCo_xNi_1-xO₃ with perovskite structure was prepared according to the citrate complexing method. Metal Ni and Co could be obtained by reducing the perovskite precursors, and the characterization results from X-ray diffraction, temperature programmed reduction and transmission electron microscopy indicated that nanoparticles of Ni-Co solid solution alloy with size of around 12 nm were formed. As a lot of metal ions can enter into perovskite lattice, this method should be extended to prepare various bimetallic or polymetallic nanoparticles. The resultant catalysts of Ni-Co alloy supported on La₂O₃ presented high activity and good stability for steam reforming of ethanol. Carbon deposition and sintering of the metal nanoparticles were investigated. After 10 hours' running at the relatively low reaction temperature of 550 °C, Co-rich catalysts showed some deactivation, which was ascribed to carbon deposition. The cobalt sites in the catalysts were inclined to form amorphous carbon, while the nickel sites tended to generate filamentous carbon, and the former led to more severe deactivation. At the relatively high reaction temperature of 700°C, the bimetallic alloy catalysts presented better anti-sintering ability than the monometal catalysts. This journal is

Full text of DOI:10.1039/c4ra14131h

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Nano bimetallic alloy of Ni–Co obtained from
LaCoxNi1ꢀxO3 and its catalytic performance for
steam reforming of ethanol
Cite this: RSC Adv., 2015, 5, 16837
ab
Fen Liu,ab Yinghong Qu,ab Yizhi Yue,ab Guilong Liuab and Yuan Liu
*
A novel preparation method for supported bimetallic alloy nanoparticles was investigated in this work.
LaCoxNi1ꢀxO3 with perovskite structure was prepared according to the citrate complexing method. Metal
Ni and Co could be obtained by reducing the perovskite precursors, and the characterization results
from X-ray diffraction, temperature programmed reduction and transmission electron microscopy
indicated that nanoparticles of Ni–Co solid solution alloy with size of around 12 nm were formed. As a
lot of metal ions can enter into perovskite lattice, this method should be extended to prepare various
bimetallic or polymetallic nanoparticles. The resultant catalysts of Ni–Co alloy supported on La2O3
presented high activity and good stability for steam reforming of ethanol. Carbon deposition and
sintering of the metal nanoparticles were investigated. After 10 hours' running at the relatively low
reaction temperature of 550 ꢁC, Co-rich catalysts showed some deactivation, which was ascribed to
carbon deposition. The cobalt sites in the catalysts were inclined to form amorphous carbon, while the
nickel sites tended to generate filamentous carbon, and the former led to more severe deactivation. At
the relatively high reaction temperature of 700 ꢁC, the bimetallic alloy catalysts presented better anti-
sintering ability than the monometal catalysts.
Received 8th November 2014
Accepted 19th January 2015
DOI: 10.1039/c4ra14131h
www.rsc.org/advances
hydrotalcite-like compound is Mg6Al2(OH)16$4H2O, where M2+
and M3+ are divalent and trivalent metal cations, respectively. It
is required that the sizes of M2+ and M3+ are close to the sizes of
1. Introduction
Bimetallic alloy nanoparticle as catalyst have been drawn much
attention. The addition of the second metal can tailor the
electronic and geometric structures of the nanoparticles and
synergistic effect can be observed, which is a promising
approach to develop new catalysts and to improve the ability on
anti-sintering and/or anti-carbon deposition.1,2
Mg2+ and Al3+, respectively. Anꢀ stands for anions, which can be
OHꢀ, CO32ꢀ, NO3ꢀ, etc. The spinel oxides are generally written
as AB2O4, the metal ions at A and B sites are divalent and
trivalent ions, respectively. Thus, it can be seen that
hydrotalcite-like compounds and spinel oxides are suitable
precursors for preparing supported bimetallic alloy nano-
particles, while a drawback is that the two metal ions are
restricted divalent and trivalent ions and with size restriction.
Perovskite-type oxides (PTOs) are generally written as ABO3,
where A sites are usually occupied by rare-earth, alkaline-earth,
or other large ions; while the B sites are lled with transition-
metal cations. Moreover, A and B sites can be partially
replaced by other metal ions, and it is reported that 90% of
metallic elements in the periodic table can be stabilized in the
lattice of the perovskite structure.9 Studies on PTOs as catalysts
are extensive, such as for oxidation/reduction reactions,10,11 and
PTOs are extensively studied materials for electrolytes, elec-
trodes and so on.12,13
The reported preparation methods for supported bimetallic
alloy nanoparticles mainly include impregnation and precipi-
tation methods,3,4 which lead to non-homogeneous distribution
of the alloy nanoparticles5 and/or the existence of the mono-
metal particles.6 It is an effective way for preparing nano-
bimetallic alloy catalysts by using complex metal oxides as the
precursors, such as hydrotalcites and spinels. In a complex
metal oxide, the metal ions are uniformly mixed at atomic level.
For example, Li et al. obtained small Ni–Cu alloy particles by
reducing a hydrotalcite-like compound containing elements of
Ni, Cu, Mg and Al.7 In another case, bimetallic Cu–Co alloy
supported on alumina was prepared by reducing CuCo2O4 on
the alumina.8 Hydrotalcite-like compounds are generally
3+
written as [M1ꢀx2+Mx
(OH)2]x+(Ax/nnꢀ)$mH2O, the typical
The reports on using PTOs as the precursors to prepare
supported nano metal catalysts are limited to single metal
catalysts. For example, De Lima et al. fabricated Ni/La2O3
catalyst derived from LaNiO3, and compared with traditional
impregnation method, smaller nickel particles were obtained
which exhibited better anti-carbon deposition ability for SRE
aCollaborative Innovation Center of Chemical Science and Engineering (Tianjin),
Tianjin 300072, China. E-mail: yuanliu@tju.edu.cn; Tel: +86 022 87401675
bTianjin Key Laboratory of Applied Catalysis Science and Technology, School of
Chemical Engineering, Tianjin University, Tianjin 300072, China
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 16837–16846 | 16837

View Article Online
RSC Advances
Paper
reaction.14 Previously, our research group prepared LaFe0.7Cu0.3O3 a D/MAX-2500 instrument, a Cu Ka radiation source was used
from citrate complexation method and loaded Co3O4 on it.15 By with an accelerating voltage of 40 kV and intensity of 200 mA.
reducing the resulted Co3O4/LaFe0.7Cu0.3O3, copper ions in the The spectra were collected at scanning speed of 5ꢁ per minute.
perovskite lattice were reduced to metal copper at lower
The BET specic surface areas of the samples were per-
temperature, and the resultant metal copper nanoparticles formed on a Micromeritics ASAP 2010 analyzer according to N2
catalyzed the reduction of Co3O4, which led to that the gener- adsorption at ꢀ196 ꢁC. 150 mg of the sample was used, and
ated metal cobalt was covered on the surface of the copper before eaꢁch test, the samples were degassed under vacuum for 8
particles, forming a core–shell structure of Cu@Co supported h at 300 C to remove the physisorbed substances.
on LaFeO3 with perovskite structure.
Temperature-programmed reduction (TPR) measurements
The reported catalysts used in steam reforming of ethanol were conducted on a Thermo-Finnigan apparatus in 5 vol%
present good activity while the stability of catalysts is needed to be H2/Ar gaseous mixture at a ow rate of 20 ml minꢀ1. 50 mg of the
improved.16,17 Thus many researches on SRE catalysts concen- sample in 60–100 mesh was loaded in a quartz tube reactor.
trated on the study of stability due to carbon deposition and/or Before TPR, the sample was ushed with a N2 ow at a rate of
sintering. Owing to low price and good activity for SRE, base 50 ml minꢀ1 at 120 ꢁC for 1 h to remove adsorbed H2O, and then
metal catalysts have been studied extensively,18–20 of which metal cooled down to room temperature under the N2 atmosphere.
Ni exhibited very good SRE catalytic performance17,19 and Co-based The sample was heated from room temperature to 900 ꢁC with a
ꢀ1
catalyst is another promising candidate.21 Hence, Ni–Co bimetallic heating rate of 10 C min in 5 vol% H2/Ar. The H2 consump-
ꢁ
systems should be of interest as catalysts for SRE reaction.
tion was monitored by a thermal conductivity detector (TCD).
In this work, PTOs of LaCoxNi1ꢀxO3 were used as the
The element detection of Co and Ni for catalysts aer
precursors to prepare bimetal of Ni–Co alloy nanoparticles, and hydrogen pretreatment were operated on an Agilent 7700x
the resulted Ni–Co/La2O3 were investigated as catalysts for SRE. model ICP-MS system. The samples were diluted with dilute
The signicance of this work includes the following two aspects: nitric acid solution to concentrations within the detection range
the rst, this preparation method for bimetallic nanoparticles of instrument.
can be extended to a lot of other kind of bimetals. Namely,
Transmission electron microscopy (TEM) investigations
LaM1ꢀxMx0 O3 can be used as the precursors for preparing were carried out on a FEI TecnaiG2 F20 equipped with energy-
bimetallic nanoparticles of M–M0, where M and M0 are any dispersive X-ray (EDX) system operating at 200 kV. The spec-
transitional metal ions which can be lled in the perovskite imen for TEM were prepared by ultrasonic dispersion of the
lattice. Furthermore, lanthanum ions in LaM1ꢀxMx0 O3 can be powder samples in the absolute ethanol, then the ethanol
replaced by other ions. The second, this method can be applied suspension was dropped and dried on a Cu grid with a layer of
to prepare supported nano bimetallic alloy catalysts. For holey carbon lm.
example, loading LaCoxNi1ꢀxO3 on SiO2 can prepare Ni–Co/SiO2
Thermogravimetric analysis (TG) and differential thermal
with La as a doping agent. As stated above that many metal ions analysis (DTA) were operated on DTG-50/50H to obtain the
can be incorporated in a perovskite lattice, hence this method amount and types of coke deposited on the spent catalysts.
can be applied to prepare a lot of bimetallic catalysts.
During TG-DTA tests, the sample was heated from 20 to 900 ꢁC
with a ramping rate of 10 ꢁC minꢀ1 in air. The carbon formation
rate was calculated by:
2. Experimental
2.1 Preparation of catalysts
w1 ꢀ w2
ꢂ1000 mg gꢀ1
(1)
w2 ꢂ t
The LaCoxNi1ꢀxO3 with perovskite structure were prepared by
the citrate complexing method as described elsewhere.15 Briey,
stoichiometric lanthanum, cobalt and nickel nitrate were added
into deionized water, and the resultant mixed nitrate solution
was stirred for several minutes. Then, citric acid and ethylene
glycol were added into the solution with the molar ratio of the
total metal ion/citric acid ¼ 1 : 1.2 and the molar ratio of citric
acid/ethylene glycol ¼ 1 : 0.5. The mixed solution was stirred
violently for ve minutes at room temperature and then the
transparent solution was stirred slowly at 80 ꢁC until resin solid
was obtained. The resin solid was dried at 120 ꢁC for 24 h to get
sponge solid. The resulted sponge solid was calcined at 300 ꢁC
where w1 is the weight of the reacted catalyst before TG tests, w2
is the weight of the reacted catalyst aer the TG tests, t is the
reaction time.
2.3 Catalytic performance tests
The catalytic performance for SRE was carried out on a
continuous ow xed-bed reactor at atmospheric pressure.
Prior to the activity evaluation, the reactor was purged with pure
N2 for 30 min, then the catalyst precursor was reduced in H2
atmosphere at 650 ꢁC for 2 h to generate the active sites for SRE
reaction, and later the system was cooled down to the reaction
temperature. 75 mg of catalyst with 40–60 mesh grain size was
diluted with the same size of SiO2 (SiO2/catalyst ¼ 8.0 in weight),
and the mixture was sandwiched with quartz wool and placed in
an 8 mm i.d. tubular reactor. The mixture of H2O/C2H5OH ¼ 3
in mole ratio was blended evenly as the feed of reactants. The
for 2 h and thenꢀa1t 750 ꢁC 5 h in air atmosphere, with a heating
ꢁ
rate of 2 C min
.
2.2 Characterization methods
The phase structure analyses of the catalysts were determined feed stock was fed into the reactor by a JMS SP 500 pump at a
by the X-ray diffraction (XRD) technique which was recorded on ow rate of 1.2 ml hꢀ1 and vaporized at 140 ꢁC. Pure N2 (60 ml
16838 | RSC Adv., 2015, 5, 16837–16846
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
structure which was ascribed to rhombohedral unit cell, space
group R3c, and no other imparity diffraction peaks could be
observed. The cobalt-containing samples exhibited similar XRD
pattern to LaNiO3, as the ion radius of Co3+ is similar to that of
3+
Ni3+ (0.61 and 0.6 A respectively), facilitating Co cation
˚
substitution inside the LaNiO3 perovskite structure.22
As shown in Table 1, the PTOs range in the average crystal
sizes from 18 to 20 nm which were calculated according to
Scherrer equation. The BET surface areas (SAs) of the samples
are between 5 m2 gꢀ1 and 7 m2 gꢀ1. The low SAs were ascribed to
the high temperature calcination in the preparation process.
The XRD spectra of the reduced samples are displayed in
Fig. 2. The diffraction peaks corresponding to La2O3 (PDF card:
050602) and metal were detected in the patterns. No any binary
or ternary metal oxide phases could be detected, indicating that
perovskite structure was totally destroyed in the reduction
process. The metal loading of the reduced samples was ca.
25 wt% by ICP, it was indicated that Ni and/or Co could be
completely reduced from perovskite structure which was
consistent with XRD results. The weak peaks (27.3, 27.9ꢁ)
belong to La(OH)3 (PDF card: 361481), as La2O3 would react with
H2O in atmospheric air to produce La(OH)3 during the char-
acterization process.
The diffraction peaks of metal nickel of line (a) in Fig. 2A and B
could be indexed to a face centered cubic (fcc) unit cell (PDF card:
040805). Metal cobalt can be existed in hexagonal closed packed
(hcp) structure besides fcc, and it was pointed that the lattice
structure had an important impact on its catalytic activity.23 While
only fcc of metal cobalt could be detected in the reduced LaCoO3
of this work. To see more clearly, the magnied diffraction
patterns of the reduced samples between 42.5 and 48ꢁ are dis-
played in Fig. 2B, and a peak of lanthanum oxide (46.08ꢁ) is used
as the reference to calibrate the shi of the metal peaks. All of the
samples exhibited only one diffraction peak corresponding to
metal, which located between the pure metal nickel at 2q ¼ 44.51ꢁ
and the pure metal cobalt at 2q ¼ 44.22ꢁ. The parameter of the
unit cell increases linearly with the increase of cobalt content in
the perovskites, as seen in Fig. 2C, which follows Vegart Law,
indicating the formation of solid solution alloys of Ni–Co.24
The average crystallite sizes of the metals estimated from
XRD lines broadening are listed in Table 1. The sizes of the
nano metals including pure nickel, pure cobalt and Ni–Co alloy
are in the range of 10 to 14 nm.
Fig. 1 XRD patterns of the perovskite precursors: (a) LaNiO3, (b)
LaCo0.3Ni0.7O3, (c) LaCo0.5Ni0.5O3, (d) LaCo0.7Ni0.3O3 and (e) LaCoO3.
minꢀ1, supplied by a mass ow meter) was introduced and
mixed with the gaseous reactants at the inlet of the reactor. The
activity was performed froꢀm1 350 ꢁC to 700 ꢁC and the space
velocity was 60 000 ml gcat hꢀ1 if not specied. The catalytic
activity evaluation in each data point was calculated from three
measurements under equilibrium, and the uncertainty analysis
indicated that the error was within 5%. The analyses of the
outlet products were on-line on two gas chromatographs with
TCD detectors and two columns of Porapack Q and TDX-01. The
carbon balances of all samples were within 100 ꢃ 10%. The
catalytic performance was evaluated by C2H5OH conversion
donated as and product distribution represented by Si, they
were described by eqn (2) and (3):
n
ꢀ n
C2H5OH;inlet
C2H5OH;outlet
XC H OH
¼
ꢂ 100%
(2)
2
5
n
C2H5OH;inlet
moles Pi
Si ¼
ꢂ 100%
(3)
n
X
moles Pi
i¼1
where n stands for number of moles and Pi is the molar amount
of different components in the reaction products
3. Results and discussion
3.1 X-ray diffraction (XRD)
The XRD patterns of LaCoxNi1ꢀxO3 (x ¼ 0, 0.3, 0.5, 0.7, 1) are
presented in Fig. 1. All of the samples showed single perovskite
Table 1 Textural characters of the prepared catalysts
Aer reduction
Dpa (nm)
Aer reaction
Dpb (nm)
Fresh catalysts
Catalyst
DPTO (nm)
SA (m2 gꢀ1
)
a
LaNiO3
18.2
19.3
18.7
19.6
19.0
6.3
5.8
6.1
5.9
5.7
12.4
12.0
13.2
11.2
10.1
32.7
24.6
24.0
26.8
34.2
LaCo0.3Ni0.7O3
LaCo0.5Ni0.5O3
LaCo0.7Ni0.3O3
LaCoO3
a
Calculated from XRD results with Scherrer formula: D ¼ Kl/(b cos q); DPTO, crystal size of perovskite; Dp, crystal size of metallic phase of the
samples reduced at 650 ꢁC. b Crystal size of metallic phase of samples reacted at 700 ꢁC for 30 h.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 16837–16846 | 16839

View Article Online
RSC Advances
Paper
Fig. 3 Temperature programmed reduction profiles of (a) LaNiO3, (b)
LaCo0.3Ni0.7O3, (c) LaCo0.5Ni0.5O3, (d) LaCo0.7Ni0.3O3 and (e) LaCoO3.
ꢁ
similar to LaNiO3. The low temperature peak at about 440 C
was assigned to the reduction of Co3+ to Co2+ accompanied with
LaCoO2.5 formation, in which the structure of perovskite was
maintained. The high temperature peak at 610 ꢁC was ascribed
to the reduction of Co2+ in LaCoO2.5 to Co0 and La2O3 forma-
tion.27 Compared with LaNiO3, the two reduction peaks of
LaCoO3 shied toward higher temperature, suggesting the
easier reducibility for LaNiO3.
For LaCoxNi1ꢀxO3 (x ¼ 0.3, 0.5, 0.7), two successive reduction
peaks were observed, which were much similar to that of
LaNiO3 and LaCoO3, corresponding to the reduction of Ni3+ to
Ni2+ and Co3+ to Co2+ for the lower temperature peaks and the
divalent metallic ions to metal for the higher temperature
peaks, respectively. The amounts of hydrogen consumed at the
lower temperature were much close to that of the hydrogen
required for converting Ni3+ to Ni2+ and Co3+ to Co2+, as shown
in Table 2. The hydrogen consumptions for the higher
temperature peaks also supported higher temperature peaks
were attributed to the reduction of Ni2+ to Ni0 and Co2+ to Co0.
The information value given by the TPR results is that the
reduction of the nickel ions and the cobalt ions in the PTOs are
simultaneously, i.e., the reduction from Ni3+ to Ni2+ is accom-
panied with Co3+ to Co2+ and Ni2+ to metal Ni is accompanied
with Co2+ to metal Co. By the way, nickel and cobalt ions in
LaCoxNi1ꢀxO3 are evenly mixed at atomic level. Thus, the
resultant bimetal of Ni–Co tends to be uniformly mixed,
favoring the formation of Ni–Co alloy.
Fig. 2 (A) XRD patterns of LaCoxNi1ꢀxO3 catalysts after reduction: (a)
LaNiO3, (b) LaCo0.3Ni0.7O3, (c) LaCo0.5Ni0.5O3, (d) LaCo0.7Ni0.3O3 and
(e) LaCoO3. (c) La2O3, (#) metal or metallic alloy, (*) La(OH)3; (B) the
magnified XRD spectra at 42.5–48ꢁ of (A); (C) lattice parameters of the
metals (A˚) vs. Co content in the reduced catalysts.
3.2 Temperature-programmed reduction (TPR)
Since the PTOs were used as the precursors, reduction treatment
was handled to obtain metal nanoparticles which were the active
3.3 Transmission electron microscopy (TEM)
phases for SRE reaction. Thus, the reducibility analysis was The TEM images of the representative reduced samples are
carried out using TPR technique, and the H2-TPR proles are showed in Fig. 4. The particle size of metal nickel reduced
presented in Fig. 3. For LaNiO3, two main hydrogen consumption from LaNiO3 is about 10 nm (seen in Fig. 4a), in accordance
peaks were observed, which were located at around 375 and with XRD results of above. In order to obtain the feature of
ꢁ
510 C separately. Many studies about the reduction of LaNiO3 bimetallic catalysts, high resolution TEM micrograph
have been reported,25,26 and according to the literatures the peak (HRTEM) and Fast Fourier Transform (FFT) were investigated.
at the lower temperature was related to the formation of Brown- As for a representative metal particle reduced from LaCo0.3Ni0.7O3,
millerite La2Ni2O5 corresponding to the reduction of Ni3+ to Ni2+. the lattice distance is 1.768 A, which is between the lattice
˚
˚
˚
The peak at the higher temperature was corresponded to the parameter of metal Ni (1.762 A) and metal Co (1.772 A), sup-
reduction of Ni2+ in the Brownmillerite to Ni0 and the formation porting the formation of Ni–Co alloy. The FFT analyses of the
of La2O3, leading to the disappearance of perovskite structure.
inserted picture in Fig. 4b show cubic structure, indicating that
While for LaCoO3, two H2-consumption peaks were observed the metal alloy possessed cubic crystal structure which agrees
which accounted for two successive reduction steps of Co3+
,
with XRD results of above. The crystal lattice spacing of the
16840 | RSC Adv., 2015, 5, 16837–16846
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Table 2 Hydrogen consumptions of LaCoxNi1ꢀxO3
Experiment values
Theoretical values
H2 consumptions (mmol)
TL
TH
Ni3+ / Ni2+
Co3+ / Co2+
Ni2+ / Ni0
Co2+ / Co0
LaNiO3
96.1
197.5
189.7
194.5
187.1
201.5
101.8
71.2
50.9
30.5
—
—
203.6
142.5
101.8
61.0
—
—
LaCo0.3Ni0.7O3
LaCo0.5Ni0.5O3
LaCo0.7Ni0.3O3
LaCoO3
102.9
104.3
107.2
95.1
30.5
50.9
71.2
101.7
61.1
101.8
142.4
203.4
˚
selected metal particle is 2.036 A which corresponded to the
For the sample reduced from LaCo0.5Ni0.5O3, an observed
[111] plane of the alloy solid solution (the inset in Fig. 4b) and metal crystal with size of about 10 nm, see Fig. 4c, the crystal
˚
˚
the lattice distance of 3.02 A was attributed to [101] lattice plane lattice spacing is 2.038 A, which was ascribed to [111] plane of
of La2O3. The line scanning proles (Fig. 4b0) present that the Ni–Co solid solution alloy. The line scanning in Fig. 4c0 shows
element distributions of Co and Ni are simultaneously, and the that the element distribution of Co and Ni is also simulta-
ratio of Ni/Co is much close to 3/7 of the Ni/Co ratio in its neously, and the ratio of Ni/Co is much close to 1, suggesting
precursor, conrming the Ni–Co alloy formation.
the formation of Ni–Co alloy solid solution. The TEM images of
LaCo0.7Ni0.3O3 also support the Ni–Co alloy formation, which is
not shown here.
The STEM and EDS mapping images of the samples
reduced from LaCo0.3Ni0.7O3 and LaCo0.5Ni0.5O3 are shown in
Fig. 5. It can be seen that the region where nickel element
enriches also gathers cobalt element, that is, the element
distribution morphology of Ni and Co is almost the same. This
also supports the Ni–Co solid solution alloy formation. In
addition, La2O3 was formed aer reduction, while the inter-
action between metal active sites and the La2O3 support was
helpful to inhibit sintering of the metal nanoparticles.28 The
uniform dispersion of the three elements of Ni, Co and La
should be attributed to the homogeneous distribution of the
three elements in their precursors of LaCoxNi1ꢀxO3. In the
reaction process, La2O3 would react with CO2 to generate
La2O2CO3 which could eliminate coke,29 hence, the lanthanum
oxide species decorated the metal particles was in favor of
removing the deposited carbon.
In summary, nanoparticles of Ni–Co alloy can be made by
reducing LaCoxNi1ꢀxO3, owing to the evenly distribution of
nickel and cobalt ions in the perovskite lattice and the simul-
taneous reduction of the nickel and cobalt ions.
By the way, let us state a little more for making bimetallic
nanoparticles from reducing a perovskite oxide precursor. To
generate bimetallic nanoparticles of M and M0 (stand for two
Fig. 4 TEM micrographs of (a) LaNiO3 catalyst after reduction, the
scale stands for 100 nm; (b) LaCo0.3Ni0.7O3 catalyst after reduction, the
scale stands for 20 nm; (c) high resolution TEM micrograph of
LaCo0.5Ni0.5O3 catalyst after reduction, the scale stands for 2 nm; (b0)
line scanning of selected area of (b) image; (c0) line scanning of Fig. 5 STEM-EDS mapping images of chemical elements for catalysts
selected area of (c) image.
reduced from (a) LaCo0.3Ni0.7O3 and (b) LaCo0.5Ni0.5O3.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 16837–16846 | 16841

View Article Online
RSC Advances
Paper
transition metal elements) by reducing LaMxM10 ꢀxO3, if the ions methane would react with water to produce hydrogen and CO or
of M and M0 are reduced sequentially other than simulta- CO2; for the latter, acetone would be further polymerized to
neously, nanoparticles with core–shell structure can be made. form coke.32 Opposite to the acetaldehyde variation, the content
And more, if the properties of M and M0 are not t for forming of methane in the exit gases decreases with the increase of
alloy, nanoparticles of metal M being adjacent to nanoparticles cobalt content in the catalysts. This indicated that nickel was
of metal M0 can be made. These are under work in our group. All more active for catalyzing acetaldehyde decomposition to
of these indicate that PTOs are a kind of interesting precursors methane and CO, which agrees with the point view of that Ni is
for making bimetallic nanoparticles.
more active to rupture C–C than metal Co on nature.33 Thus the
methane selectivity over nickel-rich catalysts was higher and the
acetaldehyde selectivity over Co-rich catalysts was higher.
Besides, it may be deduced that more coke composed of
polymers would be generated over the Co-rich catalysts than
over the Ni-rich catalysts, for that the Co-rich catalysts gener-
ated more acetaldehyde and acetaldehyde tended to polymerize
to form coke.
The content of methane in the exit gases decreases with the
reaction temperature increasing, which is due to that steam
reforming of methane was facilitated at the higher temperature.
Selectivity to hydrogen was 70% for Co-rich catalysts at low
temperature, while the value of hydrogen selectivity was about
60% over Ni-rich catalysts at the same temperature, which was
ascribed to the higher methane content for the Ni-rich catalysts,
and methane contains hydrogen atom which reduces the
hydrogen selectivity.
3.4 Catalytic performance test and the states of the used
catalysts
3.4.1 Effect of the reaction temperature. The variation of
ethanol conversions and the reaction products distributions
with reaction temperature are shown in Fig. 6. Ethanol
completely conversed at 400 ꢁC over LaNiO3, as some nickel was
partially replaced by cobalt, the temperature for complete
ethanol conversion increased, and the temperature for
ꢁ
complete ethanol conversion increased to 550 C over LaCoO3.
This shows that nickel is more active than cobalt for SRE.
The product distributions shown in Fig. 6b–f are of interest.
The content of acetaldehyde in the exit gases increases with the
cobalt content increasing in the catalysts, in which acetalde-
hyde came from dehydrogenation of ethanol. The generated
acetaldehyde either decomposed into methane and CO or
polymerized to yield acetone.30,31 For the former, the resultant
The ratio of CO/CO2 in the exit gases is depended on water
gas shi reaction (WGSR) and the reverse WGSR. At lower and
ꢁ
ꢁ
higher temperatures, namely, at 350 C and between 600 C to
ꢁ
700 C, the content of CO is higher; while in the temperature
ꢁ
ꢁ
range of 400 C to 600 C, the content of CO2 is higher. At low
reaction temperature of 350 ꢁC, the WGSR rate was low, so
smaller amount of CO which came from acetaldehyde decom-
ꢁ
position converted to CO2. At high temperature of over 600 C,
thermodynamic restriction limited the conversion of CO to CO2,
for that WGSR is an exothermic reaction.
Fig. 6 Ethanol conversion (a) and variation of products distributions Fig. 7 Variation of ethanol conversion (a) and products distributions
with reaction temperature over catalysts: (b) LaNiO3, (c) LaCo0.3
-
;
with reaction time over LaCoxNi1ꢀxO3: (b) LaCo0.5Ni0.5O3, (c)
Ni0.7O3, (d) LaCo0.5Ni0.5O3, (e) LaCo0.7Ni0.3O3, (f) LaCoO3 (--- SCH CHO
LaCo0.7Ni0.3O3 and (d) LaCoO3 (--- SCH CHO; -<- SCH COCH ; -:-
3
3
3
3
-:- SH ; -;- SCO; -A- SCH ;-◄- SCO ). Reaction condition: GHSV ¼ SH ; -;- SCO; -A- SCH ;-◄- SCO ). Reaction condition: GHSV ¼ 60 000
2
4
2
2
4
2
60 000 ml gcatꢀ1 hꢀ1, the mole ratio of water/ethanol ¼ 3.
ml gcatꢀ1 hꢀ1, the mole ratio of water/ethanol ¼ 3 at 550 ꢁC.
16842 | RSC Adv., 2015, 5, 16837–16846
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
Table 3 Carbon deposition rate over LaCoxNi1ꢀxO3 catalysts after reaction
Amount of cokea (mg gcat hꢀ1
)
Amount of cokeb (mg gcat hꢀ1
)
ꢀ1
ꢀ1
Catalysts
LaNiO3
22.9
21.6
18.2
16.8
14.4
6.3
4.8
4.6
5.3
6.2
LaCo0.3Ni0.7O3
LaCo0.5Ni0.5O3
LaCo0.7Ni0.3O3
LaCoO3
a
The reaction condition: GHSV ¼ 60 000 ml gcatꢀ1 hꢀ1, the mole ratio of water/ethanol ¼ 3 at 550 ꢁC for 10 h. b The reaction condition: GHSV ¼
120 000 ml gcat hꢀ1, the mole ratio of water/ethanol ¼ 3 at 700 ꢁC for 30 h.
ꢀ1
3.4.2 The catalytic performance of LaCoxNi1ꢀxO3 at 550 ꢁC
tubes or nano bers. It is accepted that the amorphous carbon
combusted at lower temperature and the lamentous carbon
and the characterization of the used catalysts. As shown in
Fig. 7, over LaNiO3 and LaCo0.3Ni0.7O3 the ethanol conversion
maintained 100% in 10 hours' running period. The stability of
the catalysts decreased with the increase of Co content in the
catalysts, the conversion of ethanol decreased aer 7 hours' and
3 hours' running over LaCo0.7Ni0.3O3 and LaCoO3, respectively.
The deactivation of the catalysts for SRE is mainly attributed
to coke deposition at relatively low reaction temperature.20,34
Hence, TG-DTA and TEM analyses were conducted to study the
carbon deposited. The carbon deposition rates on catalysts are
listed in Table 3. Increasing the Co content decreased the
amount of carbon deposited, the carbon deposited on the used
LaNiO3 is the highest and the carbon amount on the used
LaCoO3 is the lowest. This suggests that the type of carbon is the
more critical inuence factor as compare to the amount of the
carbon deposited.
combusted at higher temperature.35–37 Thus, it is speculated
ꢁ
that the carbon corresponding to the peak at about 480 C in
DTA is amorphous carbon generated from cobalt sites, and the
ꢁ
carbon corresponding to the peak at 530 C in DTA is lamen-
tous carbon generated from nickel sites. It is known that the
amorphous carbon tends to envelope the metal active sites,
leading to severe deactivation; while the lamentous carbon
tends to spread away from the metal active sites, leading to
lighter deactivation.37,38
TEM images were carried out to shed light on the nature of
carbon deposited on the catalysts aer reaction at 550 ꢁC for
10 h. Fig. 9 shows the TEM images of bimetallic catalyst of
LaCo0.5Ni0.5O3 and monometal catalysts of LaNiO3 and LaCoO3
aer reaction. Many lamentous carbons can be seen from the
used LaNiO3 (seen in Fig. 9a), as the red arrowheads indicated.
For the used LaCo0.5Ni0.5O3, the amount of lamentous carbon
was reduced and amorphous carbon can be observed, as seen in
Fig. 9b where red arrowheads point lamentous carbon and the
yellow arrowheads direct the amorphous carbon. For the used
LaCoO3, the particles of the catalyst are enclosed by the
Fig. 8 shows the TG and DTA proles of the used
LaCoxNi1ꢀxO3. From TG proles, one stage of weight loss was
observed for LaNiO3 and LaCo0.3Ni0.7O3, while two stages of
weight loss were seen for LaCo0.5Ni0.5O3, LaCo0.7Ni0.3O3 and
LaCoO3. Similarly, the DTA proles exhibited one peak centered
ꢁ
at about 530 C for LaNiO3 and LaCo0.3Ni0.7O3, and two peaks
were shown for LaCo0.5Ni0.5O3, LaCo0.7Ni0.3O3 and LaCoO3. The
peaks at about 480 ꢁC appeared only for Co-containing catalysts
and its intensity increased with the increase of cobalt content in
the catalysts; similarly the peaks at around 530 ꢁC were related
to nickel existence.
The deposited carbon are classied into two kinds, one is
amorphous carbon which is composed of polymers and gener-
ally named as coke, the other is lamentous carbon as nano
Fig. 8 TG/DTA curves of the used catalysts: (a) LaNiO3, (b) LaCo0.3
-
Fig. 9 TEM micrographs of (a) LaNiO3 after reaction, the scale stands
Ni0.7O3, (c) LaCo0.5Ni0.5O3, (d) LaCo0.7Ni0.3O3 and (e) LaCoO3. Reac- for 50 nm; (b) LaCo0.5Ni0.5O3 after reaction, the scale stands for 50 nm;
tion condition: GHSV ¼ 60 000 ml gcatꢀ1 hꢀ1, the mole ratio of water/ (c) LaCoO3 after reaction, the scale stands for 100 nm; (c0) magnified
ethanol ¼ 3 at 550 ꢁC for 10 h.
TEM micrograph of (c), the scale stands for 20 nm.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 16837–16846 | 16843

View Article Online
RSC Advances
Paper
amorphous carbon, seen in Fig. 9c and c0. The TEM images the initial 100% to about 70% aer 5.5ꢀh1ours' reaction at 600 ꢁC
at a space velocity of 11 500 ml gcat hꢀ1
.
Mariana et al.
40
agreed with the TG-DTA results stated above.
The results of products' distributions show in Fig. 7b–d, no studied Ni/MgAl2O4 for SRE, the catalyst began to deactivate
acetaldehyde and acetone was detected over LaNiO3 and aer 100 min's reaction and the ethanol conversion signi-
LaCo0.3Ni0.7O3 (similar to Fig. 6, not shown here), while over cantly decreased to 50.5% at 700 min's reaction at the mole ratio
ꢁ
LaCo0.5Ni0.5O3 a trace of acetaldehyde and acetone began to of H2O/C2H5OH ¼ 4.8, at 650 C and under a space velocity of
generate aer 7 hours, and a certain amount of acetaldehyde 16 000 ml gcatꢀ1 hꢀ1 41 Compared with the literatures, catalysts
.
and acetone could be observed over LaCo0.7Ni0.3O3 and LaCoO3, in this work presented very good activity and stability.
which was indicated that over cobalt sites the polymerization The XRD patterns of the used catalysts are presented in
reactions would occur as more acetaldehyde and acetone were Fig. 11A. It can be observed that lanthanum species were mainly
detected and polymers tended to take place, which formed in the state of La(OH)3 and accompanied with a trace of
amorphous carbon on the catalyst surface. Nickel is more active La2O2CO3.
to break the chemical bond between two carbon atoms than
To show the diffraction peaks of the metal clearly, the
cobalt,39 so amorphous carbon formation was comparatively magnied XRD patterns of the used samples between 37.5 and
rare on the active sites of nickel. On the other side, it is known 45.5ꢁ are shown in Fig. 11B, and the diffraction peak of
that metal nickel is an active catalyst for decomposing methane lanthanum hydroxide (39.48ꢁ) was used as the reference to
to form lamentous carbon.24,26 These evidences support the calibrate shis of the metals' peaks. The peaks associated with
view point of above, that is, cobalt sites were inclined to form metals showed just one peak in the bimetallic catalysts, and the
amorphous carbon while nickel sites tended to generate la- crystalline cell parameters manifested a well linear relation to
mentous carbon.
the cobalt content, as shown in Fig. 11C. This indicated that the
3.4.3 The catalytic performance of LaCoxNi1ꢀxO3 at 700 ꢁC Ni–Co alloy was stable under SRE reaction condition even at the
and the characterization of the used catalysts. The stability tests high reaction temperature of 700 ꢁC. No evidence of alloy phase
at 7ꢀ010 ꢁC were carried out at a high space velocity of 120 000 ml
gcat
hꢀ1, shown in Fig. 10. The conversion of ethanol
decreased slightly merely over LaCoO3 while other catalysts
maintained complete ethanol conversion in the 30 hours'
testing period, and the product distribution changed slightly
over LaCoO3 while stable product distributions were observed
over other catalysts (ethanol conversion and product distribu-
tions over LaCo0.5Ni0.5O3 and LaCo0.7Ni0.3O3 were similar to
LaCo0.3Ni0.7O3, not shown here). Carrera Cerritos et al. studied
Ni/Al2O3–La2O3 for SRE, and the results indicated that as La2O3
loading was 12 wt%, the catalyst began to deactivate aer
1.5 hours' running and the ethanol conversion decreased from
Fig. 10 Variation of ethanol conversion and products distributions Fig. 11 (A) XRD patterns for LaCoxNi1ꢀxO3 catalysts after stability tests
with reaction time over LaCoxNi1ꢀxO3: (a) LaNiO3, (b) LaCo0.3Ni0.7O3 at 700 ꢁC for 30 h: (a) LaNiO3, (b) LaCo0.3Ni0.7O3, (c) LaCo0.5Ni0.5O3, (d)
and (c) LaCoO3 (--- XCH CH OH; -:- SH ; -;- SCO; -A- SCH ;-◄- LaCo0.7Ni0.3O3 and (e) LaCoO3. (!) La2O2CO3; (*) La(OH)3; (#) metal or
3
2
2
4
h
ꢀ1, the mole metallic alloy; (Y) C; (B) the magnified XRD spectra at 37.5–45.5ꢁ; (C)
ꢀ1
SCO ). Reaction condition: GHSV ¼ 120 000 ml gcat
2
ratio of water/ethanol ¼ 3 at 700 ꢁC.
lattice parameters of metals (A˚) vs. content of Co in the catalysts.
16844 | RSC Adv., 2015, 5, 16837–16846
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
decomposition into separate metals of cobalt and nickel was
observed aer the long-term test.
Notes and references
Aer the 30 hours' running at 700 ꢁC, the sizes of the metal
nanoparticles increased from 10–13 nm before reaction to
24–35 nm aer stability tests, shown in Table 1. Considering the
severe reaction conditions, the anti-sintering ability of the metal
nanoparticles was good. The average sizes of the bimetallic alloy
nanoparticles are in the range of 24–26 nm, and the sizes of the
monometal of nickel and cobalt are 32.7 and 34.2 nm respec-
tively. The crystallite sizes of the monometal particles are
obviously larger than the sizes of the bimetallic alloy nano-
particles, suggesting that the anti-sintering ability of the
bimetallic alloy nanoparticle is better.
The carbon deposition rates for the reaction at 700 ꢁC are
listed in Table 3. Low carbon deposition rates are observed over
studied catalysts. It was proposed that high temperature favors
the carbon eliminating reactions of the reverse Boudouard
reaction (eqn (4)) and the coke gasication reaction (eqn (5)):14,32
1 L. H. Huang, R. R. Chen, D. Chu and A. T. Hsu, Int.
J. Hydrogen Energy, 2010, 35, 1138.
2 B. Lorenzut, T. Montini, L. De Rogatis, P. Canton,
A. Benedetti and P. Fornasiero, Appl. Catal., B, 2011, 101, 397.
3 W. Q. Xu, R. Si, S. D. Senanayake, J. Llorca, H. Idriss,
D. Stacchiola, J. C. Hanson and J. A. Rodriguez, J. Catal.,
2012, 291, 117.
4 N. Homs, J. Llorca and P. R. de la Piscina, Catal. Today, 2006,
116, 361.
5 A. L. Kustov, A. M. Frey, K. E. Larsen, T. Johannessen,
J. K. Nørskov and C. H. Christensen, Appl. Catal., A, 2007,
320, 98.
6 J. Bian, M. Xiao, S. J. Wang, Y. X. Lu and Y. Z. Meng, Appl.
Surf. Sci., 2009, 255, 7188.
7 D. L. Li, M. Koike, J. H. Chen, Y. Nakagawa and
K. Tomishige, Int. J. Hydrogen Energy, 2014, 39, 10959.
8 J. J. Wang, P. A. Chernavskii, A. Y. Khodakov and Y. Wang,
J. Catal., 2012, 286, 51–61.
9 M. A. Pena and J. L. G. Fierro, Chem. Rev., 2001, 101, 1981.
10 K. J. May, C. E. Carlton, K. A. Stoerzinger, M. Risch,
J. Suntivich, Y. L. Lee, A. Grimaud and Y. Shao-Horn,
J. Phys. Chem. Lett., 2012, 3, 3264.
11 J. Faye, A. Baylet, M. Trentesaux, S. Royer, F. Dumeignil,
D. Duprez, S. Valange and J. M. Tatibouet, Appl. Catal., B,
2012, 126, 134.
C + CO2 / 2CO
(4)
(5)
C + H2O / CO + H2
The amount of carbon deposited on the bimetallic alloy
catalysts was close to the monometal catalysts, and the carbon
deposition was not the main reason for the catalysts deactiva-
tion at the high reaction temperature.
12 D. Lin, Q. Wang, K. Peng and L. L. Shaw, J. Power Sources,
2012, 205, 100.
13 W. Zhou, J. Sunarso, M. W. Zhao, F. L. Liang, T. Klande and
A. Feldhoff, Angew. Chem., Int. Ed., 2013, 52, 14036.
14 S. M. de Lima, A. M. da Silva, L. O. da Costa, J. M. Assaf,
G. Jacobs, B. H. Davis, L. V. Mattos and F. B. Noronha,
Appl. Catal., A, 2010, 377, 181.
4. Conclusions
Bimetallic nanoparticles of Ni–Co could be prepared by
reducing LaCoxNi1ꢀxO3 perovskite precursors, as Ni3+ and
Co3+ were simultaneously reduced from perovskite structure,
nanoparticles of Ni–Co alloy were obtained, which were
conrmed by XRD, TEM and TPR techniques. Considering
that ninety percent of metallic elements in the periodic table
can be doped into the perovskite structure, this method for
preparing bimetallic nanoparticles is potential for extensive
application. The resultant Ni–Co alloy nanoparticles sup-
ported on La2O3 showed much high activity and good stability
for steam reforming of ethanol. Aer 10 hours' running at low
reaction temperature, Co-rich catalysts showed certain deac-
tivation due to the carbon deposition. Over the cobalt sites
amorphous carbon was generated while lamentous carbon
was formed over the nickel sites; and amorphous carbon
would result in more severe deactivation. As the stability tests
were conducted at elevated reaction temperature of 700 ꢁC,
bimetallic alloy catalysts presented better anti-sintering ability
than monometal catalysts.
15 Y. Z. Fang, Y. Liu and L. H. Zhang, Appl. Catal., A, 2011, 397,
183.
16 R. Espinal, E. Taboada, E. Molins, R. J. Chimentao,
F. Medina and J. Llorca, RSC Adv., 2012, 2, 2946–2956.
17 L. O. O. da Costa, A. M. da Silva, F. B. Noronha and
L. V. Mattos, Int. J. Hydrogen Energy, 2012, 37, 5930.
18 I. Rossetti, C. Biffi, C. L. Bianchi, V. Nichele, M. Signoretto,
F. Menegazzo, E. Finocchio, G. Ramis and A. Di Michele,
Appl. Catal., B, 2012, 117, 384.
19 S. R. Li, M. S. Li, C. X. Zhang, S. P. Wang, X. B. Ma and
J. L. Gong, Int. J. Hydrogen Energy, 2012, 37, 2940.
20 I. Rossetti, J. Lasso, E. Finocchio, G. Ramis, V. Nichele,
M. Signoretto and A. Di Michele, Appl. Catal., A, 2014, 477,
42.
21 M. S. Batista, R. K. S. Santos, E. M. Assaf, J. M. Assaf and
E. A. Ticianelli, J. Power Sources, 2003, 124, 99.
22 G. S. Gallego, C. Batiot-Dupeyrat, J. Barrault, E. Florez and
F. Mondragon, Appl. Catal., A, 2008, 334, 251.
23 H. Woo, R. Srinivasan, L. Rice, R. De Angelis and P. Reucro,
J. Mol. Catal., 1990, 59, 83.
Acknowledgements
The nancial support of this work by National Natural Science
Foundation of China (no. 21263011 and 21376170) are grate-
fully acknowledged.
24 K. Takanabe, K. Nagaoka, K. Nariai and K. Aika, J. Catal.,
2005, 232, 268.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 16837–16846 | 16845

View Article Online
RSC Advances
Paper
´
25 G. Valderrama, A. Kiennemann and M. Goldwasser, Catal. 34 A. E. Galetti, M. F. Gomez, L. A. Arrua and M. C. Abello, Appl.
Today, 2008, 133, 142. Catal., A, 2008, 348, 94.
26 M. Rivas, J. Fierro, M. Goldwasser, E. Pietri, M. Perez-Zurita, 35 G. M. Zeng, Q. H. Liu, R. X. Gu, L. H. Zhang and Y. D. Li,
A. Griboval-Constant and G. Leclercq, Appl. Catal., A, 2008,
Catal. Today, 2011, 178, 206.
344, 10.
36 S. M. de Lima, A. M. da Silva, L. O. O. da Costa,
U. M. Graham, G. Jacobs, B. H. Davis, L. V. Mattos and
F. B. Noronha, J. Catal., 2009, 268, 268.
27 S. M. Sun, L. Yang, G. S. Pang and S. H. Feng, Appl. Catal., A,
2011, 401, 199.
28 J. Sun, X. P. Qiu, F. Wu and W. T. Zhu, Int. J. Hydrogen Energy, 37 H. Wang, Y. Liu, L. Wang and Y. N. Qin, Chem. Eng. J., 2008,
2005, 30, 437. 145, 25.
29 H. Q. Chen, H. Yu, F. Peng, G. X. Yang, H. J. Wang, J. Yang 38 A. L. Alberton, M. M. V. M. Souza and M. Schmal, Catal.
and Y. Tang, Chem. Eng. J., 2010, 160, 333. Today, 2007, 123, 257.
30 B. C. Zhang, X. L. Tang, Y. Li, Y. D. Xu and W. J. Shen, Int. 39 P. Bichon, G. Haugom, H. J. Venvik, A. Holmen and
J. Hydrogen Energy, 2007, 32, 2367. E. A. Blekkan, Top. Catal., 2008, 49, 38.
31 A. N. Fatsikostas and X. E. Verykios, J. Catal., 2004, 225, 439. 40 R. C. Cerritos, R. F. Ramırez, A. F. A. Alvarado,
´
˜
´
32 J. Vicente, J. Erena, C. Montero, M. J. Azkoiti, J. Bilbao and
J. M. M. Rosales, T. V. Garcıa and I. R. G. Esquivel, Ind.
A. G. Gayubo, Int. J. Hydrogen Energy, 2014, 39, 18820.
Eng. Chem. Res., 2010, 50, 2576.
´
33 L. Chen, C. K. S. Choong, Z. Zhong, L. Huang, Z. Wang and 41 M. N. Barroso, A. E. Galetti, M. F. Gomez, L. A. Arrua and
J. Lin, Int. J. Hydrogen Energy, 2012, 37, 16321. M. C. Abello, Chem. Eng. J., 2013, 222, 142.
16846 | RSC Adv., 2015, 5, 16837–16846
This journal is © The Royal Society of Chemistry 2015