Nickel-supported on La2Sn2O7 and La2Zr2O7 Pyrochlores for methane steam reforming: Insight into the difference between tin and zirconium in the b site of the compound

Article abstract of DOI:10.1002/cctc.201402551

La₂Sn₂O₇ and La₂Zr₂O₇, two pyrochlore compounds with different B-site cations, were prepared and used as supports for Ni in methane steam reforming. Compared with Ni/γ-Al₂O₃

Full text of DOI:10.1002/cctc.201402551

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201402551
Nickel-Supported on La2Sn2O7 and La2Zr2O7 Pyrochlores for
Methane Steam Reforming: Insight into the Difference
between Tin and Zirconium in the B Site of the Compound
Youhe Ma,[a] Xiang Wang,*[a] Xiaojuan You,[a] Jianjun Liu,[a] Jinshu Tian,[a] Xianglan Xu,[a]
Honggen Peng,[a] Wenming Liu,*[a] Changqing Li,[b] Wufeng Zhou,[b] Ping Yuan,[b] and
Xiaohong Chen[b]
La2Sn2O7 and La2Zr2O7, two pyrochlore compounds with differ-
ent B-site cations, were prepared and used as supports for Ni
in methane steam reforming. Compared with Ni/g-Al2O3, both
Ni/La2Zr2O7 and Ni/La2Sn2O7 show very stable reaction per-
formance. Whereas Ni/La2Zr2O7 also displays reasonably high
activity for the reaction, the activity of Ni/La2Sn2O7 is extremely
low. It was found that severe coking occurred with Ni/g-Al2O3,
but no coke formation was observed on the two pyrochlore
catalysts. On the reduced and spent Ni/La2Sn2O7 catalyst, the
Ni3Sn2 and Ni3Sn alloys were detected; these alloys suppressed
coke formation but also decreased the activity of the catalyst.
In comparison, a large amount of La2O2CO3 was formed on the
used Ni/La2Zr2O7 catalyst; these species reacted with the
carbon deposits formed on the Ni particles and continuously
restored the Ni sites. Thus, coking was effectively suppressed
and the initial high activity of the catalyst was maintained.
Thus, Ni/La2Zr2O7 is a superior catalyst having the potential for
industrial use.
Introduction
CO þ H2O ¼ CO2 þ H2 DHꢀ ¼ ꢁ41:2 kJ molꢁ1
ð2Þ
298
With energy shortage and environmental pollution becoming
serious global problems, the search for novel green energy is
currently an urgent and pressing issue. Hydrogen, as a clean
energy source, offers an alternative to fossil fuels.[1] Although
other methods such as photocatalytic splitting of water and
biomass conversion are under intensive study by many chem-
ists, these processes are not mature enough for industrial hy-
drogen production for the present needs. Multistep processing
of hydrocarbon fuels is still the most practical approach to effi-
cient hydrogen production, and steam reforming of hydrocar-
bons has been widely employed in the chemical industry for
this purpose. In particular, the reforming of natural gas with
water steam is the leading method for the production of hy-
drogen around the world[2–5] owing to the abundant storage of
natural gas.
It is clear that this process is highly endothermic and it is
usually performed at approximately 8008C and 1.5–3.0 MPa.[6]
Although noble metal catalysts have been reported to be
more active and less sensitive to carbon deposition, their use
in reforming is confined by their high cost and limited availa-
bility.[7–9] In contrast, owing to abundant sources and their low
price, Ni-based catalysts are still the most appropriate catalysts
for this reaction.[10,11] However, some potential problems can
arise from the high temperature required for the reaction, in-
cluding sintering of the Ni active sites and supports. Therefore,
thermal stability of the catalyst is a critical issue, because low
thermal stability will not only induce deactivation of the cata-
lysts, but it will also crush the catalyst pellets, increase the
pressure drop, and eventually block the reactor.[12] Moreover,
a good support with strong thermal resistance and reasonable
surface area can stabilize the size of Ni even at high tempera-
ture, which will effectively inhibit carbon deposition and main-
tain the initial activity. Al2O3 and SiO2 are two traditional sup-
ports for Ni catalysts for various reforming reactions. Coke-re-
sistant Ni-based catalysts have attracted great interest. Recent-
ly, various strategies involving confinement effects and crystal
facet effects have been developed to suppress coking and to
maintain the high activity of the catalyst.[13–16] To obtain cata-
lysts with more potent coke resistance and higher efficiency,
some other composite oxide supports with a stable structure,
such as perovskite and spinel, have been investigated.[17–19]
Pyrochlores have the empirical formula A2B2O7, in which “A”
is typically a rare-earth element with a +3 valence state and
“B” is generally a transition element with a +4 valence state.
The major reactions of methane steam reforming are
described in Equations (1) and (2):
CH4 þ H2O ¼ CO þ 3 H2 DHꢀ ¼ 206:2 kJ molꢁ1
ð1Þ
298
[a] Y. Ma, Prof. Dr. X. Wang, X. You, J. Liu, J. Tian, Dr. X. Xu, Dr. H. Peng,
Dr. W. Liu
College of Chemistry, Nanchang University
Nanchang, Jiangxi 330031 (P.R. China)
E-mail: xwang23@ncu.edu.cn
wmliu@ncu.edu.cn
[b] C. Li, W. Zhou, P. Yuan, X. Chen
Jiangxi Golden Century Advanced Materials Co. Ltd.
Nanchang, Jiangxi 330013 (P.R. China)
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201402551.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3366

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Pyrochlores exhibit high chemical stability and catalytic activity
at high temperatures in the oxidative coupling of meth-
ane.[20,21] The application of pyrochlores as combustion cata-
lysts was also recently suggested, as reasonable surface area
and catalytic activity were observed despite the high tempera-
ture required for the formation of the pyrochlore structure.[22]
Over the past five years, pyrochlore compounds have also
been investigated as catalysts for different reforming reactions.
Spivey and co-workers studied La2Zr2O7 in which part of the B-
site Zr was substituted by Ru, Rh, Pt, or Ni for methane dry re-
ing. For comparison, Ni/g-Al2O3, a traditional industrial reform-
ing catalyst, was also prepared by using the impregnation
method with commercial g-Al2O3 as the support. Compared
with Ni/g-Al2O3, both Ni/La2Zr2O7 and Ni/La2Sn2O7 showed
much more potent coking resistance. However, high and
stable activity for the reforming reaction was only found for
Ni/La2Zr2O7. With different characterization methods, reasons
leading to these differences in performance were elucidated.
forming, and they found that the substitution led to the for- Results and Discussion
mation of more mobile oxygen species.[23–29] As a result, the re-
Activity evaluation of the catalysts
forming activity of the catalysts was improved and carbon
deposition was suppressed.[30,31] Catalysts substituted by a suit-
able amount of Ru or Rh display performance that is improved
relative to that of unsubstituted La2Zr2O7, and thus they have
the potential to be used in industrial applications.[23–29] Upon
studying Ru–pyrochlore catalysts for the autothermal reform-
ing of volatile organic compounds, Peppley and co-workers
found that Ru-substituted La2Zr2O7 was a very active and
stable catalyst for the target reactions. The reason was as-
cribed to the uniform dispersion of Ru into the La2Zr2O7 pyro-
chlore matrix.[32] Bussi and co-workers investigated Ni-, Co-,
and Cu-modified La2Zr2O7 for ethanol steam reforming. They
found that both Ni- and Co-modified catalysts showed very
high reforming activity. However, the addition of Cu into Ni-
modified La2Zr2O7 decreased the activity of the catalyst. Upon
preparing Ni-modified La2Zr2O7 with both co-precipitation and
impregnation methods, the highest H2 yield and stability were
obtained with the former method.[33–35] For ethanol steam re-
forming on La2Ce2ꢁxRuxO7 pyrochlore solid solution catalysts,
Lee and co-workers observed metal–metal interaction between
La and Ru, which thus resulted in catalysts with high activity
and stability.[36]
The reaction performance of the catalysts for methane steam
reforming at different temperatures is shown in Figure 1. Nota-
bly, all the data shown herein were collected after stabilization
for only 30 min at the corresponding temperature; therefore,
the data reflect the initial activity of the catalysts. On all of the
catalysts, both CH4 conversion and H2 yield increased with an
increase in the reaction temperature. As a result of the strong
endothermic effect of the reaction, the increase in the conver-
sion of methane was very evident, as shown in Figure 1a. In
comparison, the increase in the yield of H2 was much slower,
which could be due to the inhibition of the water–gas shift re-
action, a mild exothermic reaction, at higher temperatures, as
shown in Figure 1b. Among all of the catalysts, Ni/g-Al2O3 dis-
played the highest initial CH4 conversion and the highest H2
yield, but these values dropped very quickly because of the
deactivation of the catalyst, as shown in Figure 2. In contrast,
Ni/La2Sn2O7 showed an unexpected low conversion of CH4 and
a low yield of H2; consequently, it may not be applicable to
real industrial use. In contrast, Ni/La2Zr2O7 had a much higher
initial conversion of CH4 and a higher initial yield of H2 than
Ni/La2Sn2O7, but these values were lower than those of Ni/g-
Al2O3.
Apparently, these reports mainly focused on using substitut-
ed pyrochlores as catalysts directly for reforming reactions.
Given that pyrochlores generally have good chemical and ther-
mal stabilities and reasonable surface areas, they could also be
good candidates as supports for Ni or other metals to prepare
reforming catalysts. However, only a few publications can be
found on this topic. For instance, Hu and co-workers revealed
that NiO supported on La2Zr2O7 pyrochlore was an active and
stable catalyst for methane dry reforming as a result of the
fine and uniform dispersion of NiO on the pyrochlore support
surface.[37] In comparison, by using X-ray photoelectron spec-
troscopy (XPS) Spivey and co-workers found the presence of
NiO and Ni2O3 on the surface of Ni-substituted La2Zr2O7 pyro-
chlore.[23] They observed that the Ni-substituted catalyst deacti-
vated severely because of fast coking. In spite of this discrep-
ancy, to the best of our knowledge, Ni supported on a pyro-
chlore other than La2Zr2O7 as a catalyst for reforming reactions
has not been reported. Therefore, with the expectation to
obtain a catalyst with potent coke resistance, high activity, and
applicable stability for methane steam reforming, in this study,
both La2Sn2O7 and La2Zr2O7 pyrochlores were prepared by the
co-precipitation method, and they were subsequently used as
supports for Ni to prepare catalysts for methane steam reform-
For the purpose of H2 production by steam reforming, ach-
ieving a high H2/CO ratio is of great importance. Therefore, the
H2/CO ratios of the catalysts at different temperatures are com-
pared in Figure 1c. To a certain extent, the H2/CO ratios are
controlled by the second step in the water–gas shift reaction
involved in the steam reforming reaction. Increasing the oper-
ating temperature clearly reduced the H2/CO ratios gradually
for all of the catalysts, because of the exothermic properties of
the second step in the water–gas shift reaction. Among the
three catalysts, the highest H2/CO ratio was achieved on the
Ni/La2Sn2O7 catalyst, and the lowest ratio was achieved on the
Ni/La2Zr2O7 catalyst. However, the difference in the H2/CO ratio
between the Ni/La2Zr2O7 and Ni/g-Al2O3 catalysts is negligible.
Although Ni/La2Sn2O7 displayed the highest H2/CO ratio, its use
could be limited by its low activity, as mentioned above.
For the steam reforming reaction, it is always a big challenge
and a major objective to find a stable catalyst having potent
resistance to coke formation and Ni crystallite size growth
during the high-temperature process. Therefore, to study the
properties of the catalysts further, they were subjected to
long-term stability tests at 8008C, and the results are shown in
Figure 2. As shown in Figure 2a, although Ni/g-Al2O3 exhibited
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3367

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 2. Long-term stability test: a) methane conversion, b) H2 yield, and
c) H2/CO ratio. Reaction conditions: H2O/CH4 =2:1, T=8008C,
Figure 1. Reaction performance of Ni/g-Al2O3, Ni/La2Sn2O7, and Ni/La2Zr2O7
for methane steam reforming: a) methane conversion, b) H2 yield, and c) H2/
ꢁ1
ꢁ1
ꢁ1
ꢁ1
WHSV=18000 mLh
g
.
cat
CO ratio. Reaction conditions: H2O/CH4 =2:1, WHSV=18000 mLh
g
.
cat
the highest initial conversion of CH4 among the three catalysts,
over the first 80h, the conversion of CH4 decreased from 94 to
78%, which indicates that the catalyst deactivates quickly.
However, the yield of H2 remained stable. On the contrary, the
conversion of CH4 on Ni/La2Zr2O7 increased over the first induc-
ing period of 120h and eventually stabilized at 98% after
180h. Furthermore, after an inducing period of 80h, this cata-
lyst exhibited a yield of H2 that was similar to that exhibited by
Ni/g-Al2O3. Similar to Ni/La2Zr2O7, Ni/La2Sn2O7 also showed an
inducing period of approximately 20h, after which the conver-
sion of CH4 remained stable at approximately 30%. Apparently,
the stable activity of Ni/La2Sn2O7 and the yield of H2 on this
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3368

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
catalyst are much lower than that observed on Ni/La2Zr2O7 and
Ni/g-Al2O3. As shown in Figure 2c, Ni/La2Zr2O7 and Ni/g-Al2O3
have the same stable H2/CO ratio, which is slightly lower than
that of Ni/La2Sn2O7. Taking into account all of the aspects of
the three studied catalysts, it is clear that Ni/La2Zr2O7 is the
most promising catalyst for the target reaction in this study,
which has the potential for industrial use.
curred during the steam reforming process, which could be
one of the major causes for the quick decrease in activity. Be-
sides this, a minute weight loss stage with approximately 1.8%
weight loss between 300 and 4008C that was accompanied by
a small exothermic peak at 3188C was also observed for Ni/g-
Al2O3; this could be attributed to the combustion of a small
amount of a very active carbon deposit, such as absorbed
atomic carbon or Ni carbide species.[39,40] Interestingly, for both
the Ni/La2Sn2O7 and Ni/La2Zr2O7 catalysts, no evident weight
loss was observed, which indicates the lack of deposited
carbon species. On the contrary, after 3008C, a small amount
of weight gain was detected, which was due to the reoxidation
of the reduced metallic Ni back into NiO species on the two
catalysts. The two small exothermic peaks at 285 and 3938C in
the DSC profile of Ni/La2Zr2O7 and the small exothermic peak
at 4888C in the DSC profile of Ni/La2Sn2O7 further confirm this.
Clearly, by using the two pyrochlore compounds as the sup-
port for Ni, coke formation was effectively suppressed. After
all, we believe this explains the stable reaction performance of
these two pyrochlore-supported Ni catalysts.
Thermogravimetric analysis and differential scanning
calorimetry of coke formation on the spent catalysts after
long-term stability tests
To clarify the reasons leading to the reaction performance dif-
ference, thermogravimetric analysis and differential scanning
calorimetry (TGA-DSC) were used to analyze the catalysts after
long-term stability tests; the results are shown in Figure 3. For
Scanning electron microscopy analysis of coke formation on
the spent catalysts after long-term stability tests
Scanning electron microscopy (SEM) was used to investigate
the morphology and surface changes of the catalysts before
and after long-term stability tests. For convenient comparison,
the SEM images of freshly reduced Ni/g-Al2O3,Ni/La2Sn2O7 and
Ni/La2Zr2O7 are shown in Figure 4a–c, whereas those after the
long-term stability tests are shown in Figure 4a’–c’ in se-
quence. As presented in Figure 4a, the surface of the freshly
reduced Ni/g-Al2O3 catalyst is very clean and consists mainly of
fine fibrous g-Al2O3. After a reaction time of 80h at 8008C,
whisker-like carbon fiber deposits with an average diameter of
50 nm can be clearly observed, as shown by Figure 4a’. This is
in line with the TGA-DSC results, which indicated that severe
coking occurred to the catalyst during the reaction. The SEM
images of the freshly reduced Ni/La2Sn2O7 and Ni/La2Zr2O7 cat-
alysts show that the two samples are made up of small round
particles that are 25–30 nm in size. After long-term reaction,
no significant change occurred to Ni/La2Sn2O7, but a slight ag-
gregation was observed for Ni/La2Zr2O7. However, the clean
surfaces of both spent catalysts testify to the absence of any
carbon deposit, and this indicates that coking can be com-
pletely suppressed by using the two pyrochlores as supports
for Ni.
Figure 3. a) TGA and b) DSC profiles of long-term used Ni/g-Al2O3, Ni/
La2Sn2O7, and Ni/La2Zr2O7 catalysts
In summary, the SEM images of the catalysts further testify
to the fact that on Ni/g-Al2O3, severe coking takes place during
the methane steam reforming reaction, and this could be one
of the major reasons why the catalyst deactivates so quickly.
However, on the Ni/La2Sn2O7 and Ni/La2Zr2O7 catalysts, no coke
formation was observed. As a consequence, both catalysts
show stable reaction performance for methane steam
reforming.
Ni/g-Al2O3, a 21.5% weight loss stage was observed between
400 to 7008C in the TGA curve, which was accompanied by
two strong exothermic peaks at 554 and 5768C in the DSC
profile; this indicates the presence of two types of carbon de-
posits with different chemical environments, perhaps graphitic
carbon atoms.[38] Apparently, severe coking of this catalyst oc-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3369

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 4. SEM images of freshly reduced a) Ni/g-Al2O3, b) Ni/La2Sn2O7, and
c) Ni/La2Zr2O7 and those of spent a’) Ni/g-Al2O3, b’) Ni/La2Sn2O7, and c’) Ni/
La2Zr2O7.
N2 adsorption–desorption study on the freshly calcined
supports and the corresponding Ni catalysts
To elucidate the reasons leading to the different reaction per-
formance, the texture properties of the freshly calcined sup-
ports and the corresponding Ni catalysts before and after use
were measured by N2 adsorption–desorption, and the results
are compared in Figure 5 and Table 1. As exhibited in Fig-
ure 5a, all the supports show a typical type IV adsorption fea-
ture, which indicates the presence of mesopores that are gen-
erally larger than 5 nm. The presence of H3 hysteresis loops in
the nitrogen adsorption–desorption isotherms of all the sam-
ples further confirm this. In comparison with the supports, all
Table 1. Physicochemical properties of the freshly calcined and spent
Figure 5. N2 adsorption–desorption isotherms of a) the supports, b) the
catalysts measured by N2 adsorption–desorption.
freshly calcined Ni catalysts, and c) the spent Ni catalysts.
Catalysts
BET surface
area [m2 gꢁ1
Pore volume
[m2 gꢁ1
]
Pore size
[nm]
]
g-Al2O3
84.4
22.3
21.8
55.6
8.9
4.4
40.3
7.0
0.513
0.281
0.116
0.366
0.092
0.057
0.279
0.063
0.020
13.2
36.9
14.1
12.6
41.6
13.6
11.2
50.9
32.1
of the catalysts with 12% supported Ni display no large differ-
ence in their adsorption–desorption behavior. However, after
long-term stability tests, the hysteresis loops became smaller
than those of the fresh catalysts, and this indicates that the
structure changed during the reaction. It is clear, however, that
the spent catalysts still retain their mesoporous structures.
For comparison, the BET surface areas, pore volumes, and
pore sizes of the samples are also listed in Table 1. g-Al2O3 has
the largest BET surface area and pore volume among the three
La2Sn2O7
La2Zr2O7
Ni/g-Al2O3
Ni/La2Sn2O7
Ni/La2Zr2O7
Ni/g-Al2O3
Ni/La2Sn2O7
Ni/La2Zr2O7
[a]
[a]
[b]
[a]
[b]
[b]
2.1
[a] Freshly calcined catalyst. [b] Spent catalyst.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3370

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
supports, as a result of its well-known porous structure. Al-
though La2Sn2O7 and La2Zr2O7 have similar surface areas, the
pore volume and pore size of the latter is much smaller.
La2Sn2O7 has the largest pore size among all of the supports.
Considering the results shown in Figure 4b,b’, this suggests
that some mesopores could be formed among the particles.
After Ni loading, the surface areas and pore volumes of all the
samples decreased, and this indicates possible blockage of
part of the pores by NiO. In addition, all of the supported Ni
catalysts were recalcined in air at 8008C for another 6h after
Ni loading, which could be another reason for their lower sur-
face areas and pore volumes. After long-term reaction at
8008C for more than 50h, the surface areas as well as the pore
volumes of all the samples decreased, and this indicates the
impact of high temperature on the pore structure.
X-Ray diffraction analysis of phase composition and texture
properties of the freshly calcined supports and the
corresponding Ni catalysts
The X-ray diffraction (XRD) patterns of the freshly calcined sup-
ports and the corresponding Ni catalysts are displayed in
Figure 6. As shown in Figure 6a, commercial Al2O3 used in this
study shows diffraction peaks at 2q=37.5, 46.8, and 66.78,
which is typical for the g-Al2O3 phase. Both La2Sn2O7 and
La2Zr2O7 display the typical intense diffraction peaks of the py-
rochlore structure at 2q=28.74, 33.31, 47.83, and 56.768,
which confirms that the solid-state reaction between La2O3
and SnO2 or ZrO2 took place during calcination at 8008C thus
to produce the La2Sn2O7 and La2Zr2O7 pyrochlore compounds.
However, in the XRD pattern of La2Sn2O7, a minute peak at
2q=26.68, which is attributed to the strongest diffraction peak
of the tetragonal rutile SnO2 phase, can still be detected. Al-
though a stoichiometric La/Sn ratio of 1:1 was intentionally se-
lected for sample preparation, it seems that the reaction be-
tween La2O3 and SnO2 was not complete, perhaps as a result
of limitations imposed by the reaction equilibrium. Therefore,
in this catalyst, part of the free SnO2 is still present.
Figure 6. XRD patterns of a) the freshly calcined supports and b) the
corresponding Ni catalysts.
H2-TPR study on the freshly calcined supports and the
corresponding Ni catalysts
After Ni loading and recalcination, all of the catalysts exhibit-
ed diffraction features that were essentially the same as those
of the corresponding supports, except that all the diffraction
peaks become more intense. This better crystallization is in line
with the lower surface areas of the supported Ni catalysts. Nev-
ertheless, NiO microcrystallites can be evidently observed for
both Ni/La2Sn2O7 and Ni/La2Zr2O7, as indicated by the specific
diffraction peaks of NiO spinel at (111) 37.688 and (20 0)
43.668. In contrast, no diffraction peaks related to Ni species
can be detected for Ni/g-Al2O3, because of the reaction be-
The H2-TPR profiles of the freshly calcined supports and corre-
sponding Ni catalysts are presented in Figure 7. Both the g-
Al2O3 and La2Zr2O7 supports display no reduction below
9008C; this indicates that these two supports are nonreducible.
However, in the profile of the La2Sn2O7 support shown in Fig-
ure 7a, a reduction peak at 6188C, which is assigned to the re-
duction of SnO2 in the support, is clearly observed.[43] This is in
agreement with the XRD results in Figure 6 and further con-
firms the presence of a small amount of free SnO2 in this
sample. The quantification results of the amount of H2 con-
sumed by the samples are listed in Table 2. The La2Sn2O7 sup-
port has a H2 consumption of 0.8mmolgꢁ1, which is equal to
an O/Sn atomic ratio of 0.25. If all of the Sn4+ in the sample is
reduced to metallic Sn, the H2 consumption amount should be
6.4mmolgꢁ1. Therefore, it is reasonable to deduce that only
approximately 12.5% Sn4+ is present in the sample as free
SnO2. The remaining 87.5% Sn4+ is present in the crystal
matrix of La2Sn2O7 pyrochlore. Owing to the stable structure of
tween NiO and g-Al2O3 to form a new compound, spinel
[41,42]
NiAl2O4.
Indeed, our H2 temperature-programmed reduc-
tion (H2-TPR) results in the following section also provide
strong evidence to testify to the formation of spinel NiAl2O4
compounds in this catalyst, which escapes detection by XRD
as a result of its low crystallization. Notably, the diffraction
peaks of the SnO2 phase become more intense in the pattern
of Ni/La2Sn2O7, which further confirms the presence of free
SnO2 in this catalyst.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3371

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
to achieve more complete information. All of the profiles show
a reduction peak at approximately 6108C, which is assigned to
the reduction of the free SnO2 to metallic Sn. Apparently, with
the reduction stopped at 7008C, even with an extended time,
no further reduction took place, as shown by Figure S1b. How-
ever, if the reduction temperature was improved to 900 and
10008C, a high temperature reduction peak, which is attribut-
ed to the reduction of Sn4+ in the pyrochlore La2Sn2O7 phase,
can be clearly observed.
To gain a deeper understanding of the reduction process,
the phase composition of the reduced samples after reduction
at each ending temperature were also analyzed by XRD. As
shown by Figure S2, after reduction at 7008C, the original
12.5% of free SnO2 in the calcined sample disappeared, where-
as metallic Sn was observed. The pyrochlore phase remained
intact. For clarification, the part of the XRD patterns between
43 and 478 are enlarged and are plotted beside the major
figure to show the metallic Sn and La2O3 signals more clearly.
By increasing the ending temperature to 900 and 10008C, the
intensity of the metallic Sn peaks became larger, and the La2O3
phase started to show at 9008C. On the contrary, the intensity
of the peaks of pyrochlore La2Sn2O7 decreased and eventually
disappeared at 10008C, which is indicative of the collapse of
the pyrochlore structure. These results demonstrate that Sn4+
in the lattice of pyrochlore La2Sn2O7 can only be reduced at
temperatures higher than 7008C and that the reduction is
complete at 10008C. In summary, these results strongly prove
the high chemical stability of the pyrochlore structure, which
impedes the reduction of the Sn4+ cations in the crystal
matrix.
The H2-TPR profiles of the supported Ni catalysts are listed in
Figure 7b. Ni/g-Al2O3 has a reduction peak at 8068C, which is
attributed to the reduction of spinel NiAl2O4.[41,42] This indicates
the presence of a strong interaction between NiO and Al2O3,
which leads to the solid-state reaction between the two
during the high-temperature calcination. The quantified O/Ni
atomic ratio in Table 2 is 0.8, which is lower than the theoreti-
cal value of 1.0. Apparently, owing to the formation of the
NiAl2O4 structure, part of the Ni species becomes more difficult
to reduce. However, it was reported that a mixture of Ni cat-
ions in the matrix of spinel NiAl2O4 can hinder growth of the
Ni particles during prereduction of the catalysts, which has
a positive effect on the activity and stability of the catalysts.[44]
The profile of Ni/La2Sn2O7 shows two reduction peaks at 400
and 5008C. The 4008C peak is assigned to the reduction of
NiO to metallic Ni, which was confirmed by the quantified O/
Ni atomic ratio of 1.0 in Table 2. The 5008C peak is assigned to
the reduction of free SnO2 in the sample. Compared with the
La2Sn2O7 support, the reduction temperature is 1008C lower
because of the H2 spillover effect induced by metallic Ni from
the reduction of NiO in this sample. The deconvolution and in-
tegration results show that the O/Sn atomic ratio based on
this peak is 0.24. If all the Sn4+ in the sample is reduced to
metallic Sn, the theoretical O/Sn ratio should be 2.0. Therefore,
in this Ni-containing sample 12% Sn4+ is present as free SnO2.
Within experimental error, this amount is the same as the
12.5% free SnO2 achieved on the La2Sn2O7 support without Ni.
Figure 7. H2-TPR profiles of a) the freshly calcined supports and b) the
corresponding supported Ni catalysts.
Table 2. Quantification results of H2-TPR for the freshly calcined catalysts.
Catalysts
H2 consumption
O/M atomic ratio
O/Sn O/Ni
[mmolgꢁ1
]
g-Al2O3
–
–
–
La2Sn2O7
La2Zr2O7
Ni/g-Al2O3
Ni/La2Sn2O7
Ni/La2Zr2O7
0.8
–
1.7
2.8
2.0
0.25
–
–
0.24
–
–
–
0.80
1.00
1.00
[a]
[a] The reduction of NiAl2O4.
pyrochlore compounds, a much higher temperature is needed
to reduce this part of Sn4+ into metallic Sn.
To clarify this, the La2Sn2O7 sample was subjected to H2-TPR
experiments with different ending temperatures; the results
are shown in Figure S1 (Supporting Information). For clarifica-
tion, Figure S1a plots the H2 consumption versus the reduction
temperatures, whereas Figure S1b plots the H2 consumption
versus the reduction time. Notably, for each experiment, the
sample was always held at the ending temperature for 30min
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3372

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Ni/La2Zr2O7 also displays two major reduction peaks at 343
and 4618C. As mentioned above, the La2Zr2O7 support itself
shows no reduction in the studied temperature region. There-
fore, the reduction peaks can be safely assigned to the reduc-
tion of Ni oxides. As listed in Table 2, the overall H2 consump-
tion gives an O/Ni atomic ratio of 1.0; this implies that the Ni
species are present solely as NiO in the sample, but with differ-
ent chemical environments. The NiO species with the higher
reduction temperature clearly have a stronger interaction with
the La2Zr2O7 support.[27] After all, the presence of multiple NiO
species with different properties could be an important reason
for the superior reaction performance of the Ni/La2Zr2O7
catalyst.
Ni/La2Sn2O7 catalyst, not metallic Ni but the Ni3Sn2 alloy was
observed, which indicates that NiO reacts with the 12% free
SnO2 on the La2Sn2O7 support during the reduction process.
For elucidation, the 2q values of the typical diffraction peaks of
the Ni3Sn2 alloy are listed in Table S1. Notably, several minute
diffraction peaks of La2O3 were detected for the reduced Ni/
La2Sn2O7 catalyst, and this testifies to the fact that the La2Sn2O7
pyrochlore phase could start to be reduced at this temperature
to release La2O3 species originally in the pyrochlore structure.
After long-term stability tests, there was no significant change
in the Ni/g-Al2O3 and Ni/La2Zr2O7 catalysts, except that a new
graphite diffraction peak was observed in the pattern of Ni/g-
Al2O3, which provides evidence in addition to the TGA-DSC
and SEM data that severe coke formation occurs on this cata-
lyst during the reaction. Interestingly, after a reaction time of
60h, the original Ni3Sn2 alloy phase in Ni/La2Sn2O7 disappeared,
and a new Ni3Sn alloy was formed. Former studies have
proved that the presence of a Ni–Sn alloy in the Ni-based cata-
lysts can suppress coke formation but at the expense of re-
forming activity.[42,45–47]
XRD analysis of phase composition and texture properties
for the freshly reduced and spent catalysts
The XRD patterns of the reduced and spent catalysts are
shown in Figure 8. As shown in Figure 8a, metallic Ni was de-
tected in both the Ni/g-Al2O3 and Ni/La2Zr2O7 catalysts after re-
duction at 8008C for 2h in 10% H2/Ar flow. However, for the
The crystallite sizes of the pyrochlore supports, metallic Ni,
and Ni–Sn alloys are compared in Table 3. After the reaction,
Table 3. Crystallite sizes of the reduced and spent catalysts measured by
XRD.
Catalyst
Pyrochlore crystallites
size [nm]
Metal crystallite
size [nm]
La2Sn2O7
La2Zr2O7
Ni
Ni–Sn alloy
reduced[a]
spent[b]
–
–
–
–
–
–
30.2
34.5
24.5
33.6
–
–
12.3
16.7
–
–
14.4
10.7
–
Ni/g-Al2O3
Ni/La2Sn2O7
Ni/La2Zr2O7
reduced[a] 33.8
spent[b]
reduced[a]
spent[b]
31.7
–
–
–
[a] Reduced at 8008C for 2 h. [b] Ni/Al2O3 obtained after 70 h, Ni/La2Sn2O7
obtained after 60 h, and Ni/La2Zr2O7 obtained after 250 h.
the crystallite size of La2Sn2O7 decreased by approximately
2 nm, but that of La2Zr2O7 increased by approximately 4 nm.
The SEM image in Figure 4c also demonstrates that the
La2Zr2O7 support has a slight aggregation after 250h. This
small change indicates that both of the pyrochlore supports
have high chemical and thermal stability. For both the Ni/g-
Al2O3 and Ni/La2Zr2O7 catalysts, the Ni crystallite size increased
after long-term reaction. However, the Ni size growth rate of
Ni/g-Al2O3 is clearly larger than that of Ni/La2Zr2O7, which could
be another reason for the quick deactivation of Ni/g-Al2O3. This
could be one of the reasons why Ni/La2Zr2O7 has a high reac-
tion stability. Former studies indeed indicate that coking on Ni
is closely related to its size. Generally, the larger the Ni size, the
easier it is to form coke.[48,49] Notably, the relatively smaller Ni
crystallite size on Ni/La2Zr2O7 could be an important reason for
its high activity. It is not very meaningful to compare the crys-
tallite sizes of the two Ni–Sn alloys, because of phase and
chemical composition changes from Ni3Sn2 to Ni3Sn during the
Figure 8. XRD patterns of a) the freshly reduced catalysts and b) the spent
catalysts.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3373

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
reaction. However, the crystallite sizes of both Ni–Sn alloys are
below 15 nm.
Upon studying carbon formation for the reforming reaction
over Ni-based catalysts, Nikolla and co-workers found that the
formation of a Ni–Sn alloy can increase coke resistance of the
catalysts at the expense of activity.[45–47] On the basis of DFT
calculations and scanning transmission electron microscopy
(STEM) experiments, they found that Sn displaced some Ni
from the carbon nucleation sites by alloy formation, which
lowered the binding energy of the surface carbon. Conse-
quently, the overall rate of oxidation of surface carbon was im-
proved significantly, and it was much greater than the rate of
CꢁC bond formation; thus, coke deposition was suppressed ef-
fectively. On the other hand, the addition of Sn to form a Ni–
Sn alloy increases the activation energy for the CH4 dry reform-
ing reaction, which indicates that the presence of Sn has a neg-
ative effect on the activity. The inherent reason was ascribed
to blockage of part of the low-coordinate Ni sites on the Ni
particles by Sn, which are actually the active sites for CꢁH acti-
vation, the rate-determining step for CH4 reforming. The same
phenomenon was observed by us for methane dry reforming
on Sn-modified Ni/Al2O3 catalysts.[42]
In the present study, the formation of Ni3Sn2 and Ni3Sn inter-
metallic compounds on the Ni/La2Sn2O7 catalyst was directly
observed by XRD analysis. Apparently, the presence of Ni3Sn2
and Ni3Sn alloys should be a major reason for the low reform-
ing activity and strong coke resistance of the catalyst.
FTIR spectroscopy study on the surface properties and
adsorbates of the reduced and spent catalysts
Figure 9. FTIR spectra of the reduced and used catalysts: a) Ni/La2Sn2O7 and
b) Ni/La2Zr2O7.
FTIR spectroscopy was used to monitor the surface composi-
tion of both the reduced and spent catalysts. The FTIR spectra
of the reduced and spent Ni/La2Sn2O7 catalysts are compared
in Figure 9a. Both spectra show a band at n˜ =570 cmꢁ1, which
is assigned to vibration of the SnꢁO bonds.[50] Compared with
the reduced sample (Figure 9a, a), several new IR bands ap-
peared at n˜ =916, 992, 1459, and 1504 cmꢁ1 for the spent
sample. According to the literature, the band at n˜ =1504 cmꢁ1
can be assigned to formate species, whereas that at n˜ =
1459 cmꢁ1 can be assigned to La2O2CO3 species, which could
come from the reaction between La2O3 and CO2 in the reaction
flow.[51,52] The FTIR spectra of the reduced and spent Ni/
La2Zr2O7 catalysts are compared in Figure 9b. After long-term
reaction, formate and lanthanum oxycarbonate La2O2CO3 spe-
cies can also be observed, as evidenced by the bands at n˜ =
1512 and 1467 cmꢁ1, respectively. Compared with spent Ni/
La2Sn2O7, the signal for La2O2CO3 in this sample is much stron-
ger, which indicates the formation of more lanthanum oxy-
carbonate.
deactivated by coking can be clean up with restoration of the
activity. In this way, coke formation of the catalysts is impeded,
and thus the catalysts exhibit good stability and activity. In this
study, the La2O2CO3 species was formed on both the Ni/
La2Sn2O7 and Ni/La2Zr2O7 catalysts, which clearly improved
their coking resistance and benefitted their reaction per-
formance. As shown in Figure 2, these two Ni–pyrochlore cata-
lysts showed very stable reforming activity during the long-
term test. Indeed, the presence of La2O2CO3 species could be
the inherent reason for this. Notably, the FTIR results demon-
strate the formation of much more La2O2CO3 species on Ni/
La2Zr2O7 than on Ni/La2Sn2O7, which explains the reason why
Ni/La2Zr2O7 has much higher reforming activity than Ni/
La2Sn2O7. In addition, as we mentioned above, the formation
of Ni–Sn alloys on Ni/La2Sn2O7 can decrease the activity of the
Ni sites, and as a consequence, coke formation on this catalyst
is suppressed at the expense of reforming activity.
Upon studying dry reforming of natural gas on Ni/La2O3,
Shirsat and Verykios and also found the formation of La2O2CO3
species on the catalysts.[51–53] In their opinion, this La2O2CO3
species plays a critical role in maintaining the activity and sta-
bility of the Ni/La2O3 catalysts. In detail, the La2O2CO3 species
can react with carbon deposits on the Ni particles at the inter-
face between Ni and La2O2CO3. Therefore, the initial Ni sites
In conclusion, all of these reasons make Ni/La2Zr2O7 a superi-
or catalyst for methane steam reforming; it shows much better
reaction performance than Ni/g-Al2O3 and Ni/La2Sn2O7. With
further optimization, Ni/La2Zr2O7 could be developed and used
in industrial applications.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3374

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Catalyst characterization
Conclusions
The specific surface areas of the samples were measured by nitro-
gen adsorption–desorption at 77 K with an ASAP2020 instrument.
Specific surface areas were calculated by using the Brunauer–
Emmett–Teller (BET) method in the relative pressure (P/P0) range of
0.05–0.25. The pore size distribution of the samples was calculated
with the Barrett–Joyner–Halenda (BJH) method, and the average
pore sizes were obtained from the peak positions of the distribu-
tion curves. Total pore volume was accumulated at relative pres-
sure P/P0 =0.99.
In this study, La2Sn2O7 and La2Zr2O7, two pyrochlore com-
pounds with different B-site cations, were prepared by the co-
precipitation method and were used as supports for Ni to pre-
pare catalysts for methane steam reforming. Compared with
the traditional Ni/g-Al2O3 catalyst, both Ni/La2Zr2O7 and Ni/
La2Sn2O7 showed very stable reaction performance. Whereas
Ni/La2Zr2O7 also displayed high activity for the reaction, the ac-
tivity of Ni/La2Sn2O7 was extremely low. TGA-DSC and SEM re-
sults revealed that severe coking occurred on Ni/g-Al2O3, but
no carbon deposition was found on either of the pyrochlore
catalysts, which indicates that coking can be effectively sup-
pressed with the two pyrochlores as supports for Ni. XRD anal-
ysis of the reduced and spent Ni/La2Sn2O7 catalysts provided
direct evidence to demonstrate the formation of Ni3Sn2 and
Ni3Sn intermetallic compounds, respectively, which can deacti-
vate the Ni active sites. As a consequence, coke formation on
the catalyst was significantly inhibited but at the expense of
reforming activity. In comparison, FTIR spectroscopy indicated
that a large amount of La2O2CO3 species was formed on the
Ni/La2Zr2O7 catalyst, and it can react with the carbon deposits
formed on the Ni particles to restore the Ni active sites contin-
uously; this thus effectively suppresses coking and maintains
the initial high activity of the catalyst. In addition, Ni supported
on La2Zr2O7 has much smaller crystallite sizes than g-Al2O3,
which is also believed to be beneficial to the activity and sta-
bility of the Ni/La2Zr2O7 catalyst. In conclusion, with further op-
timization, the Ni/La2Zr2O7 catalyst could be developed for
industrial use.
The powder X-ray diffraction (XRD) patterns were recorded with
a Bruker AXS D8 Focus diffractometer operating at 40 kV and
30mA, with CuKa irradiation (l=1.5405 ꢁ). Scans were taken in the
2q range from 10 to 908 with a step of 28minꢁ1. To keep the data
comparable, all of the samples were tested continuously. The
mean crystallite sizes of the samples were calculated with the
Scherrer equation on the basis of the three strongest peaks of the
pyrochlores, La2O3, and Ni.
Hydrogen temperature-programmed reduction (H2-TPR) experi-
ments were performed with a FINESORB 3010C instrument in
a 10% H2/Ar gas mixture flow with a rate of 30mLminꢁ1. The tem-
perature was increased from room temperature to 10008C at
a ramp rate of 108Cminꢁ1. Generally, 50mg of the catalyst was
used for the test. A thermal conductivity detector was employed
to monitor H2 consumption. For H2 consumption quantification,
CuO (99.99%) was used as the calibration standard sample.
Scanning electron microscopy (SEM) images were taken with a Hita-
chi S-4800 field-emission scanning electron microscope. Transmis-
sion electron microscopy (TEM) images were taken with a Tecnai
F30 transmission electron microscope.
Thermogravimetric analysis differential scanning calorimetry (TGA-
DSC) was used to monitor the amount of coke deposited on the
spent catalysts. The experiments were performed with approxi-
mately 10mg of the spent catalyst on a TAQ600 instrument at
a ramping rate of 108Cminꢁ1 from 25 to 8008C in an air flow of
100mLminꢁ1
.
Experimental Section
FTIR spectra were collected with a Nicolet Avatar 360 spectrometer.
The spectra were recorded by accumulating 64 scans at a spectral
resolution of 2 cmꢁ1. The sample was pressed into a self-supported
wafer (18mm diameter, ꢂ0.05g) and introduced into a cell
equipped with KBr windows.
Catalyst preparation
The La2Sn2O7 and La2Zr2O7 pyrochlore supports were prepared by
the co-precipitation method from the corresponding chemicals.
Typically, the calculated amount of La(NO3)3·6H2O (99.9%) and
SnCl4·5H2O (99.9%) was mixed into a uniform aqueous solution,
and then NH3·H2O was added dropwise under constant stirring
until the pH value of the mixture was adjusted to approximately
10. Afterwards, the mixture was separated by centrifuging. The
achieved solid was then washed completely with distilled deion-
ized water until the total dissolved solids of the outflow liquid was
less than 20 ppm. After thoroughly drying, the solid was calcined
at 8008C for 4h in air to obtain the final La2Sn2O7 pyrochlore sup-
port. The above method and procedure was also used to prepare
the La2Zr2O7 support by using La(NO3)3·6H2O (99.9%) and
ZrOCl2·8H2O (99.9%) as precursors.
Activity evaluation
Evaluation of the activity of the catalysts was performed in a fixed-
bed microreactor made of a quartz tube with an inner diameter of
8 mm. The catalyst (400mg) was reduced in 10% H2/Ar at 8008C
for 2h in situ. Deionized water with a flow rate of 0.064mLminꢁ1
,
which gave a steam/methane ratio of 2:1, was fed through
a pump into a heated chamber controlled at 2008C to evaporate
the liquid water completely. Afterwards, steam and CH4 were
mixed in a gas mixer at a stoichiometric ratio of 2:1. Notably, the
flow rate of pure CH4 (99.99%) was 40mLminꢁ1 and that for water
Ni(NO3)3 (99.9%) solution was impregnated onto the pyrochlore
supports with the incipient wetness method. After impregnation,
the samples were place under ambient conditions for 24h, then
dried at 808C, and calcined at 8008C for 4h in air to obtain the Ni/
La2Sn2O7 and Ni/La2Zr2O7 catalysts. For comparison, Ni/g-Al2O3, a tra-
ditional reforming catalyst, was also prepared by using the impreg-
nation method with commercial g-Al2O3 as the support. The Ni
loading for all the prepared catalysts was 12 wt%.
steam was 80mLminꢁ1, which corresponds to a weight hourly
ꢁ1
space velocity (WHSP) of 18000mLh
g
ꢁ1. The outlet gas passed
cat
through a cold water trap to condense the excess amount of
water in the reaction flow, and it was then analyzed on-line with
a GC9310 gas chromatograph equipped with a TDX-01 column
and a thermal conductivity detector by using Ar as the carrier gas
to monitor hydrogen, carbon monoxide, methane, and carbon di-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3375

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[25] D. Pakhare, D. Haynes, D. Shekhawat, J. Spivey, Appl. Petrochem. Res.
2012, 2, 27–35.
[26] D. Pakhare, C. Shaw, D. Haynes, D. Shekhawat, J. Spivey, J. CO2 Utiliza-
tion 2013, 1, 37–42.
oxide. After the reaction, the spent catalyst was analyzed by TGA-
DTA in an air flow to measure the amount of carbon deposition.
[27] V. Abdelsayed, D. Shekhawat, J. A. Poston, J. J. Spivey, Catal. Today
2013, 207, 65–73.
Acknowledgements
[28] D. Pakhare, H. Wu, S. Narendra, V. Abdelsayed, D. Haynes, D. Shekhawat,
D. Berry, J. Spivey, Appl. Petrochem. Res. 2013, 3, 117–129.
[29] D. Pakhare, V. Schwartz, V. Abdelsayed, D. Haynes, D. Shekhawat, J.
Poston, J. Spivey, J. Catal. 2014, 316, 78–92.
[30] S. Q. Chen, H. Wang, Y. Liu, Int. J. Hydrogen Energy 2009, 34, 7995–8005.
[31] R. B. Duarte, M. Olea, E. Iro, T. Sasaki, K. Itako, J. A. V. Bokhoven, Chem-
CatChem 2014, 6, 2898–2903.
[32] A. Qi, C. Thurgood, B. Peppley, Energy Procedia 2012, 29, 503–512.
[33] J. Bussi, N. Bespalko, S. Veiga, A. Amaya, R. Faccio, M. C. Abello, Catal.
Commun. 2008, 10, 33–38.
[34] N. Bespalko, A. C. Roger, J. Bussi, Appl. Catal. A 2011, 407, 204–210.
[35] J. Bussi, M. Musso, S. Veiga, N. Bespalko, R. Faccio, A. C. Roger, Catal.
Today 2013, 213, 42–49.
[36] S. Weng, Y. Wang, C. Lee, Appl. Catal. B 2013, 134–135, 359–366.
[37] H. Meng, R. S. Hu, X. F. Zhang, G. Q. Lu, H. Q. Su, J. Chin. Rare Earth Soc.
2006, 24, 62–67.
[38] J. A. Torres, J. Llorca, A. Casanovas, M. Domꢃnguez, J. Salvadꢄ, D. Mon-
tanꢅ, J. Power Sources 2007, 169, 158–166.
[39] J. J. Guo, H. Lou, H. Zhao, D. F. Chai, X. M. Zheng, Appl. Catal. A 2004,
273, 75–82.
[40] D. P. Liu, R. Lau, A. Borgna, Y. H. Yang, Appl. Catal. A 2009, 358, 110–
118.
[41] J. K. Xu, W. Zhou, Z. J. Li, J. H. Wang, J. X. Ma, Int. J. Hydrogen Energy
2009, 34, 6646–6654.
[42] J. J. Liu, H. G. Peng, W. M. Liu, X. L. Xu, X. Wang, C. Q. Li, W. F. Zhou, P.
Yuan, X. H. Chen, W. G. Zhang, H. B. Zhan, ChemCatChem 2014, 6,
2095–2104.
[43] X. L. Xu, R. B. Zhang, X. R. Zeng, X. Han, Y. C. Li, Y. Liu, X. Wang, Chem-
CatChem 2013, 5, 2025–2036.
[44] N. Muradov, F. Smith, A. T-Raissi, Int. J. Hydrogen Energy 2008, 33, 2023–
2035.
[45] E. Nikolla, J. W. Schwank, S. Linic, Catal. Today 2008, 136, 243–248.
[46] E. Nikolla, J. Schwank, S. Linic, J. Am. Chem. Soc. 2009, 131, 2747–2754.
[47] E. Nikolla, J. Schwank, S. Linic, J. Catal. 2009, 263, 220–227.
[48] X. L. Zhu, P. P. Huo, Y. P. Zhang, D. G. Cheng, C. J. Liu, Appl. Catal. B
2008, 81, 132–140.
[49] C. J. Liu, J. Ye, J. Jiang, Y. Pan, ChemCatChem 2011, 3, 529–541.
[50] K. Li, T. Zhang, H. Wang, H. Yan, J. Solid State Chem. 2006, 179, 1029–
1034.
[51] A. N. Shirsat, M. Ali, K. N. G. Kaimal, S. R. Bharadwaj, D. Das, Thermochim.
Acta 2003, 399, 167–170.
[52] X. E. Verykios, Int. J. Hydrogen Energy 2003, 28, 1045–1063.
[53] A. C. Basagiannis, X. E. Verykios, Appl. Catal. A 2006, 308, 182–193.
The authors acknowledge support of this work by the Chinese
Natural Science Foundation (21263015, 21203088) and the Edu-
cation Department of Jiangxi Province (KJLD14005, GJJ14205).
Keywords: alloys · lanthanum · nickel · pyrochlores · steam
reforming
[1] L. Barreto, A. Makihira, K. Riahi, Int. J. Hydrogen Energy 2003, 28, 267–
284.
[2] S. Ayabe, H. Omoto, T. Utaka, R. Kikuchi, K. Sasaki, Y. Teraoka, K. Eguchi,
Appl. Catal. A 2003, 241, 261–269.
[3] E. Kikuchi, Y. Nemoto, M. Kajiwara, S. Uemiya, T. Kojima, Catal. Today
2000, 56, 75–81.
[4] D. L. Trimm, Catal. Today 1999, 49, 3–10.
[5] S. Freni, G. Calogero, S. Cavallaro, J. Power Sources 2000, 87, 28–38.
[6] V. V. Chesnokov, V. I. Zaikovskii, R. A. Buyanov, V. V. Molchanov, L. M.
Plyasova, Catal. Today 1995, 24, 265–267.
[7] A. T. Ashcroft, A. K. Cheetham, M. L. H. Green, P. D. F. Vernon, Nature
1991, 352, 225–226.
[8] J. R. Rostrupnielsen, J. H. B. Hansen, J. Catal. 1993, 144, 38–49.
[9] J. H. Bitter, K. Seshan, J. A. Lercher, J. Catal. 1997, 171, 279–286.
[10] K. Sutthiumporn, S. Kawi, Int. J. Hydrogen Energy 2011, 36, 14435–
14446.
[11] S. Damyanova, B. Pawelec, K. Arishtirova, J. L. G. Fierro, Int. J. Hydrogen
Energy 2012, 37, 15966–15975.
[12] J. Sehested, Catal. Today 2006, 111, 103–110.
[13] X. Du, D. Zhang, L. Shi, R. Gao, J. Zhang, J. Phys. Chem. C 2012, 116,
10009–10016.
[14] X. Du, D. Zhang, R. Gao, L. Huang, L. Shi, J. Zhang, Chem. Commun.
2013, 49, 6770–6772.
[15] X. Du, D. Zhang, L. Shi, R. Gao, J. Zhang, Nanoscale 2013, 5, 2659–2663.
[16] T. Xie, L. Shi, J. Zhang, D. Zhang, Chem. Commun. 2014, 50, 7250–7253.
[17] N. Sahli, C. Petit, A. C. Roger, A. Kiennemann, S. Libs, M. M. Bettahar,
Catal. Today 2006, 113, 187–193.
[18] S. O. Choi, S. H. Moon, Catal. Today 2009, 146, 148–153.
[19] M. Simꢂes, S. Baranton, C. Coutanceau, Appl. Catal. B 2010, 93, 354–
362.
[20] C. Petit, J. L. Rehspringer, A. Kaddouri, S. Libs, P. Poix, A. Kiennemann,
Catal. Today 1992, 13, 409–416.
[21] A. K. C. A. T. Ashcroft, M. L. H. Green, C. P. Grey, P. D. F. Vernon, J. Chem.
Soc. Chem. Commun. 1989, 1667–1669.
[22] J. M. Sohn, M. R. Kim, S. I. Woo, Catal. Today 2003, 83, 289–297.
[23] S. Gaur, D. J. Haynes, J. J. Spivey, Appl. Catal. A 2011, 403, 142–151.
[24] S. Gaur, D. Pakhare, H. Wu, D. J. Haynes, J. J. Spivey, Energy Fuels 2012,
26, 1989–1998.
Received: July 18, 2014
Revised: August 4, 2014
Published online on October 30, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2014, 6, 3366 – 3376 3376