Hydrogen production by steam reforming of LNG over Ni/Al2O3-ZrO2 catalysts: Effect of Al2O3-ZrO2 supports prepared by a grafting method

Article abstract of DOI:10.1016/j.molcata.2006.12.005

Al₂O₃-ZrO₂ supports with various zirconium contents were prepared by grafting a zirconium precursor onto the surface of γ-Al₂O₃. Ni(20 wt%)/Al₂O₃-ZrO₂ catalysts were t

Full text of DOI:10.1016/j.molcata.2006.12.005

Journal of Molecular Catalysis A: Chemical 268 (2007) 9–14
Hydrogen production by steam reforming of LNG over Ni/Al2O3–ZrO2
catalysts: Effect of Al2O3–ZrO2 supports prepared by a grafting method
Jeong Gil Seo, Min Hye Youn, In Kyu Song∗
School of Chemical and Biological Engineering, Research Center for Energy Conversion and Storage,
Seoul National University, Shinlim-dong, Kwanak-ku, Seoul 151-744, South Korea
Received 24 October 2006; received in revised form 2 December 2006; accepted 4 December 2006
Available online 9 December 2006
Abstract
Al2O3–ZrO2 supports with various zirconium contents were prepared by grafting a zirconium precursor onto the surface of ␥-Al2O3.
Ni(20 wt%)/Al2O3–ZrO2 catalysts were then prepared by an impregnation method, and were applied to the hydrogen production by steam reforming
of LNG. The effect of Al2O3–ZrO2 supports on the performance of the Ni(20 wt%)/Al2O3–ZrO2 catalysts was investigated. Al2O3–ZrO2 prepared
by a grafting method served as an efficient support for the nickel catalyst in the steam reforming of LNG. ZrO2 inhibited the incorporation of
nickel species into the lattice of Al2O3 and prevented the growth of metallic nickel particles during the reduction step. The crystalline structures
and catalytic activities of the Ni(20 wt%)/ZrO2–Al2O3 catalysts were strongly influenced by the amount of zirconium grafted. LNG conversion
and hydrogen yield showed volcano-shaped curves with respect to zirconium content. Among the catalysts tested, the Ni(20 wt%)/ZrO2–Al2O3
(Zr/Al = 0.17) catalyst showed the best catalytic performance in terms of both LNG conversion and hydrogen yield. The well-developed and pure
tetragonal phase of ZrO2–Al2O3 (Zr/Al = 0.17) played an important role in the adsorption of steam and the subsequent spillover of steam from the
support to the active nickel. The high reducibility of Ni(20 wt%)/ZrO2–Al2O3 (Zr/Al = 0.17) was also responsible for the enhanced performance
of the catalyst.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Alumina–zirconia support; Nickel catalyst; Grafting method; LNG; Steam reforming; Hydrogen production
1. Introduction
ported nickel catalysts generally suffer from severe catalyst
deactivation due to the sintering of nickel particles [7] and the
insufficient thermal and chemical stability of the support [8].
Several attempts have been made to overcome these problems.
These examples include the addition of second metals such as
potassium, magnesium, cerium, and molybdenum [9–12], and
the impregnation of nickel catalyst on various supports such as
ZrO2, SiO2, and mixed oxides [13–15].
It is well known that the performance of a supported nickel
catalyst depends, not only on the nature and structure of the
active nickel, but also on the chemical and textural property of
the support. The selection and modification of an appropriate
support for a nickel catalyst, therefore, can be a potential route
to improve the catalytic performance of a supported nickel cat-
alyst. It has been reported that zirconia support enhanced the
adsorption of steam onto its surface and activated the gasification
of hydrocarbons or carbon precursors adsorbed on the catalyst
surfaceinthesteamreformingreactions, resultinginanenhance-
ment in hydrogen yield and coke resistance [16]. In the steam
and CO2 reforming of methane, Ni/ZrO2 was found to show a
Catalytic reforming technology of methane has been widely
studied for use in the large-scale production of hydrogen or car-
bon monoxide [1–4]. In particular, steam reforming of methane
is generally accepted as a feasible route to produce hydrogen
for various fuel cell systems [5–7]. Liquified natural gas (LNG),
which is abundant and mainly composed of methane, can serve
as an alternate source for hydrogen production by steam reform-
ing. The extensive piping system for LNG in modern cities also
makes LNG well suited as a hydrogen source for residential
reformers.
Nickel-based catalysts have been widely used in the steam
reforming reactions. However, the nickel-based catalysts require
a high reaction temperature and excess amount of steam to
prevent the coke deposition on the catalyst surfaces [1,3]. Sup-
∗
Corresponding author. Tel.: +82 2 880 9227; fax: +82 2 889 7415.
E-mail address: inksong@snu.ac.kr (I.K. Song).
1381-1169/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2006.12.005

10
J.G. Seo et al. / Journal of Molecular Catalysis A: Chemical 268 (2007) 9–14
higher catalytic activity and long-term stability than Ni/Al2O3
catalyst [17]. It was also reported that Pt/Al2O3–ZrO2 catalyst
showed a better activity and higher stability than Pt/Al2O3 cat-
alyst in the production of synthesis gas by dry reforming of
methane [18]. Furthermore, Ni/Al2O3–ZrO2 catalyst showed a
high activity and strong resistance to coke deposition in the par-
tial oxidation and dry reforming of methane [19]. However, no
attempt has been made to utilize an Al2O3–ZrO2 support for
nickel catalysts in the hydrogen production by steam reforming
of LNG.
Al2O3–ZrO2 can be synthesized by co-precipitation [20],
sol–gel [21], and grafting methods [22–24]. Alumina, which
has many hydroxyl groups on the surface, can be modified
by grafting a zirconium precursor onto the surface of alumina.
The chemical properties of Al2O3–ZrO2 prepared by a grafting
method are different from those of Al2O3–ZrO2 prepared by a
co-precipitation method or a sol–gel method [22]. Therefore, it is
expected that Al2O3–ZrO2 prepared by a grafting method would
show interesting properties as a support for nickel catalyst.
In this work, a series of Al2O3–ZrO2 supports with various
zirconium loadings were prepared by grafting a zirconium pre-
cursor onto the surface of ␥-Al2O3. Ni/Al2O3–ZrO2 catalysts
were then prepared by an impregnation method for use in the
hydrogen production by steam reforming of LNG. The effect of
Al2O3–ZrO2 supports on the performance of Ni/Al2O3–ZrO2
catalysts in the steam reforming of LNG was investigated.
stirred at room temperature for 6 h to achieve the complete reac-
tion of the activated surface hydroxyl groups of alumina with
butoxide groups of the zirconium precursors. After removing the
unreacted zirconium precursor and butanol (by-product) by cen-
trifugation, the slurry was washed several times with anhydrous
toluene. Upon the addition of an excess amount of deionized
water to the washed slurry, a white gel was formed within a few
seconds. After maintaining the white gel in deionized water for
6 h, a solid product was obtained by filtration. The solid product
was dried overnight at 120 ◦C, and then calcined at 700 ◦C for
5 h to yield the Al2O3–ZrO2 support. The prepared Al2O3–ZrO2
support was denoted as AZ-X (X = 1–4), where X is the number
of times the whole preparation process was repeated. For exam-
ple, AZ-2 denotes an Al2O3–ZrO2 support that was prepared
by repeating the entire process two times (by adding known
amounts of Zr(OBu)4 two times through the entire preparation
process).
Ni/Al2O3–ZrO2 catalysts were prepared by impregnat-
•
ing known amounts of a nickel precursor (Ni(NO3)2 6H2O,
Aldrich) onto ␥-Al2O3 (AZ-0), AZ-1, AZ-2, AZ-3, and AZ-
4 supports. The nickel loading was fixed at 20 wt% in all cases.
The prepared catalysts were denoted as 20Ni/AZ-X (X = 0–4).
2.2. Characterization
The chemical compositions of Al2O3–ZrO2 (AZ-X) sup-
ports were determined by ICP-AES analyses (ICPS-1000IV,
Shimadzu). The crystalline phases of supports and supported
catalysts were investigated by XRD (M18XHF-SRA, MAC
2. Experimental
2.1. Preparation of Al2O3–ZrO2 (AZ-X) support and
Ni/Al2O3–ZrO2 (Ni/AZ-X) catalyst
˚
Science) measurements using Cu-K␣ radiation (λ = 1.54056 A)
operated at 50 kV and 100 mA. In order to examine the reducibil-
ity of supported catalysts, temperature-programmed reduction
(TPR) measurements were carried out in a conventional flow
system with a moisture trap connected to a thermal conductivity
detector (TCD) at temperatures ranging from room temperature
to 1000 ◦C with a ramping rate of 5 ◦C/min. For the TPR mea-
surements, a mixed stream of H2 (2 ml/min) and N2 (20 ml/min)
was used for 0.1 g of catalyst sample.
Al2O3–ZrO2 supports with various zirconium loadings were
prepared by grafting an appropriate amount of zirconium precur-
sor onto the surface of ␥-Al2O3, according to the similar method
reported in literatures [22–24]. Fig. 1 shows the schematic pro-
cedure for the preparation of Al2O3–ZrO2 support by a grafting
method. A known amount of alumina (␥-Al2O3, Degussa) was
added to 100 ml of anhydrous toluene (Aldrich) for uniform dis-
persion, and an excess amount of triethylamine (TEA, Fluka)
was then added to the alumina slurry to activate the hydroxyl
groups on the alumina surface. An appropriate amount of zir-
conium precursor (Zr(OBu)4, Aldrich) was slowly added to the
slurry with constant stirring for 1 h, and the resulting slurry was
2.3. Steam reforming of LNG
The steam reforming of LNG was carried out in a continuous
flow fixed-bed reactor at atmospheric pressure. Each calcined
catalyst (100 mg) was charged into a tubular quartz reactor, and
it was then reduced with a mixed stream of H2 (10 ml/min) and
N2 (30 ml/min) at 800 ◦C for 3 h. Water was sufficiently vapor-
ized by passing a pre-heating zone and continuously fed into the
reactor together with LNG (92.0 vol.% of CH4 and 8.0 vol.%
of C2H6) and N2 carrier (30 ml/min). The steam/carbon ratio
in the feed stream was fixed at 2.0, and the total feed rate with
respect to the catalyst was maintained at 27,000 ml h−1/g. The
catalytic reaction was carried out at 600 ◦C. The reaction prod-
ucts were periodically sampled and analyzed using an on-line
gas chromatograph (Younglin, ACME 6000) equipped with a
thermal conductivity detector. LNG conversion and hydrogen
yield were calculated according to the following equations on
Fig. 1. Schematic procedure for the preparation of Al2O3–ZrO2 support by a
grafting method.

J.G. Seo et al. / Journal of Molecular Catalysis A: Chemical 268 (2007) 9–14
11
Table 1
corresponding to the tetragonal phase of zirconia (solid lines in
Fig. 2), which were not observed in the ␥-Al2O3 (AZ-0) support.
This result strongly supports that ZrO2 was successfully grafted
onto the ␥-Al2O3. The AZ-1 and AZ-2 supports with low zir-
conium contents (Zr/Al = 0.09 and 0.17, respectively) showed
relatively broad peaks corresponding to the tetragonal phase of
zirconia. This indicates that zirconia was highly dispersed on the
alumina in the AZ-1 and AZ-2 supports. It is believed that the
co-existence of ZrO2 and Al2O3 may affect the surface struc-
ture of an Al2O3–ZrO2 support prepared by a grafting method,
as has been observed for an Al2O3–ZrO2 support prepared by a
co-precipitation method [20]. Although it has been reported that
the tetragonal phase of zirconia is unstable at room temperature
[25], the above result may be due to the fact that meta-stable
tetragonal ZrO2 is stabilized by its incorporation into Al2O3
which has a higher elastic modulus than ZrO2 [21].
On the other hand, AZ-3 and AZ-4 supports with relatively
high zirconium contents (Zr/Al = 0.31 and 0.45, respectively)
showed sharp peaks corresponding to the tetragonal phase of
zirconia, alongwithweakpeakscorrespondingtothemonoclinic
phase of zirconia. The appearance of a monoclinic phase of
zirconia at high zirconium loadings is believed to be due to
the non-homogeneous mixing of Al and Zr species during the
preparation step. It should be noted that the size of ZrO2 particles
in the AZ-3 and AZ-4 supports is larger than that in the AZ-1
and AZ-2 supports. This indicates that the grain size of ZrO2 in
the AZ-3 and AZ-4 exceeds the critical size for causing a phase
transformation from the tetragonal to the monoclinic phase [21].
The above results imply that the crystalline structures of AZ-X
supports are strongly influenced by the amount of zirconium
grafted.
Chemical compositions of Al2O3–ZrO2 (AZ-X) supports determined by ICP-
AES analyses
Support
Amount of Zr
used (wt%)
Zr loading (wt%)
Zr/Al atomic
ratio
AZ-1
AZ-2
AZ-3
AZ-4
15
30
45
60
13.4
21.9
31.8
38.4
0.09
0.17
0.31
0.45
the basis of carbon balance:
ꢀ
ꢁ
F
+ F
+ F
CH4,out
C2H6,out
C2H6,in
LNG conversion (%) = 1 −
× 100,
F
CH4,in
Fhydrogen,out
hydrogen yield (%) =
× 100
C2H6,in
2F
+ 3F
CH4,in
3. Results and discussion
3.1. Crystalline structure of AZ-X (X = 0–4)
Chemical compositions of Al2O3–ZrO2 (AZ-X) supports
determined by ICP-AES analyses are listed in Table 1. The
amount of zirconium grafted onto ␥-Al2O3 was increased with
increasing amounts of zirconium used from AZ-1 to AZ-4.
Although the amount of actual zirconium loading increased lin-
early with increasing amounts of zirconium used, the amount
of actual zirconium loading was less than that of the zirconium
used. This indicates that a considerable amount of the zirco-
nium precursors was unreacted and was washed out during the
preparation step. This is because ␥-Al2O3 and as-synthesized
Al2O3–ZrO2 supports have a limited number of hydroxyl groups
on the surface [22]. The Zr/Al ratio of the AZ-X supports
increased linearly from 0.09 to 0.45 with increasing amount of
zirconium loading from 13.4 to 38.4 wt%.
3.2. Crystalline structure of 20Ni/AZ-X (X = 0–4)
Fig. 3 shows the XRD patterns of ␥-Al2O3 (AZ-0) and
20Ni/AZ-X (X = 0–4) catalysts calcined at 700 ◦C for 5 h. The
20Ni/AZ-X (X = 0–4) catalysts showed diffraction peaks for NiO
species (JCPDS 22-1189) and nickel aluminate species. The co-
Fig. 2 shows the XRD patterns of ␥-Al2O3 (AZ-0) and AZ-
X (X = 1–4) supports calcined at 700 ◦C. The prepared AZ-X
(X = 1–4) supports showed amorphous diffraction peaks of ␥-
Al2O3. The AZ-X (X = 1–4) supports showed diffraction peaks
Fig. 2. XRD patterns of ␥-Al2O3 (AZ-0) and AZ-X (X = 1–4) supports calcined
at 700 ◦C. Solid lines represent the tetragonal phase of zirconia.
Fig. 3. XRD patterns of ␥-Al2O3 (AZ-0) and 20Ni/AZ-X (X = 0–4) catalysts
calcined at 700 ◦C for 5 h.

12
J.G. Seo et al. / Journal of Molecular Catalysis A: Chemical 268 (2007) 9–14
Fig. 5. TPR profiles of 20Ni/AZ-X (X = 0–4) catalysts.
Fig. 4. XRD patterns of 20Ni/AZ-X (X = 0–4) catalysts reduced at 800 ◦C for
3 h.
500–700 ◦C [28]. The reduction of nickel aluminate occurs at
above 800 ◦C because of the strong interaction between nickel
species and alumina [28].
existence NiO and nickel aluminate (small amount) might be
due to the low calcination temperature [26] or the high nickel
loading. It has been reported that calcined Ni/␥-Al2O3 catalysts
showed no diffraction peaks for NiO species in several cases
because of the strong interaction between the nickel species and
alumina [26–28]. It was also reported that, since the ionic radius
of nickel is larger than that of aluminum, the incorporation of
nickel into the ␥-Al2O3 increased the lattice parameter of alu-
mina, resulting in a shift of (4 4 0) diffraction peak of alumina to
a lower diffraction angle [29–32]. A shift of (4 4 0) diffraction
peak of alumina to a lower diffraction angle was also observed
for the 20Ni/AZ-0 catalyst. The shift of (4 4 0) diffraction peak
of alumina in the 20Ni/AZ-X (X = 1–4) catalysts became smaller
or negligible with increasing amounts of zirconium grafted. This
indicates that the presence of ZrO2 inhibited the incorporation of
nickel species into the lattice of ␥-Al2O3. It also implies that the
interaction between nickel species and support in the 20Ni/AZ-
X (X = 1–4) catalysts would be somewhat different from that in
the 20Ni/AZ-0 catalyst.
Fig. 4 shows the XRD patterns of 20Ni/AZ-X (X = 0–4) cata-
lysts reduced at 800 ◦C for 3 h. The reduced 20Ni/AZ-0 catalyst
showed relatively sharp XRD peaks corresponding to metallic
nickel (JCPDS 03-1051) at 2θ = 44.8◦, 52.2◦, and 76.8◦, indi-
cating the formation of large nickel particles in the 20Ni/AZ-0
catalyst. On the other hand, the 20Ni/AZ-X (X = 1–4) catalysts
showed broad and weak XRD peaks for metallic nickel with
increasing zirconium content. This indicates that the presence
of ZrO2 on the ␥-Al2O3 prevented the growth of metallic nickel
particles during the reduction process through the formation of
a new ZrO2–Al2O3 composite structure.
Fig. 5 shows the TPR profiles of 20Ni/AZ-X (X = 0–4) cat-
alysts. The 20Ni/AZ-X (X = 1–4) catalysts showed a broad
reduction peak at around 750 ◦C, while 20Ni/AZ-0 catalyst
showed two broad reduction peaks at around 450 and 750 ◦C.
The reduction peak appearing at high temperature in the
20Ni/AZ-0 catalyst can be attributed to the reduction of NiO
species that interacted strongly with ␥-Al2O3 (AZ-0) and/or to
the reduction of nickel aluminate species, while that appear-
ing at low temperature is due to the reduction of NiO that
interacted weakly with ␥-Al2O3 (AZ-0). A close examination
of the reduction profiles revealed that the reduction peaks of
20Ni/AZ-X (X = 1–4) appeared at low temperature, compared
to the reduction peak of 20Ni/AZ-0 which appeared at high
temperature. Furthermore, no reduction peak associated with
weakly interacted NiO species was observed in the 20Ni/AZ-X
(X = 1–4) catalysts due to the new interaction between Al2O3 and
ZrO2. Among the 20Ni/AZ-X (X = 1–4) catalysts, the 20Ni/AZ-
2 catalyst showed the highest reducibility (the lowest reduction
temperature).
20Ni/AZ-3 and 20Ni/AZ-4 showed another shoulder at
around 600 ◦C. This is believed to be due to the reduction of NiO
species that had interacted with ZrO2. This means that nickel
species were supported not only on the surface of Al2O3 but
also on the surface of ZrO2, when much amount of ZrO2 was
grafted on the surface of the Al2O3. Judging from the fact that the
20Ni/AZ-2 catalyst showed the highest reducibility among the
20Ni/AZ-X (X = 1–4) catalysts, it can be concluded that an opti-
mum ratio of ZrO2/Al2O3 is required for the efficient formation
of a 20Ni/Al2O3–ZrO2 catalyst.
3.3. Reducibility of 20Ni/AZ-X (X = 0–4)
3.4. Steam reforming of LNG over 20Ni/AZ-X (X = 0–4)
catalysts
TPR measurements were carried out to investigate the
reducibility of the 20Ni/AZ-X (X = 0–4) catalysts and to examine
the interaction between nickel species and AZ-X supports. It is
well known that the reduction profile of a supported nickel cat-
alyst is dependent on the interaction between the nickel species
and support. Unsupported NiO is reduced at around 400 ◦C,
while NiO species supported on ␥-Al2O3 is reduced at around
Fig. 6 shows the LNG conversion with time on stream
in the steam reforming of LNG over 20Ni/AZ-X (X = 0–4)
catalysts at 600 ◦C. The 20Ni/AZ-X (X = 0–4) catalysts showed
a stable catalytic performance during the catalytic reaction
extending over 600 min. No significant catalyst deactiva-
tion was observed in the 20Ni/AZ-X (X = 0–4) catalysts

J.G. Seo et al. / Journal of Molecular Catalysis A: Chemical 268 (2007) 9–14
13
Fig. 6. LNG conversion with time on stream in the steam reforming of LNG
over 20Ni/AZ-X (X = 0–4) catalysts at 600 ◦C: (ꢀ) 20Ni/AZ-0; (᭹) 20Ni/AZ-1;
(ꢀ) 20Ni/AZ-2; (ꢁ) 20Ni/AZ-3; (ꢂ) 20Ni/AZ-4.
Fig. 7. LNG conversion as a function of zirconium loading over 20Ni/AZ-X
(X = 0–4) catalysts in the steam reforming of LNG at 600 ◦C. The data were
obtained after a 300 min-reaction.
due to the mild reaction conditions. LNG conversion was
decreased in the order of 20Ni/AZ-2 (Zr/Al = 0.17) > 20Ni/AZ-
Figs. 7 and 8 show the LNG conversion and hydrogen
yield as a function of zirconium loading over 20Ni/AZ-X
(X = 0–4) catalysts in the steam reforming of LNG at 600 ◦C,
respectively. The data were obtained after a 300 min-reaction.
As shown in Figs. 7 and 8, LNG conversion and hydrogen
yield showed volcano-shaped curves with respect to zirconium
loading. Both LNG conversion and hydrogen yield were
decreased in the order of 20Ni/AZ-2 (Zr/Al = 0.17) > 20Ni/AZ-
3
(Zr/Al = 0.31) > 20Ni/AZ-1 (Zr/Al = 0.09) > 20Ni/AZ-4
(Zr/Al = 0.45) > 20Ni/AZ-0 (Zr/Al = 0). The catalytic per-
formance of 20Ni/AZ-X (X = 1–4) was better than that of
20Ni/AZ-0 catalyst. Among the catalysts tested, the 20Ni/AZ-2
catalyst showed the highest LNG conversion. The reasons why
the 20Ni/AZ-2 catalyst showed the best catalytic performance
in this reaction can be explained by effect of zirconia grafted
on the surface of the alumina. One possible reason is attributed
to the high reducibility of the 20Ni/AZ-2 catalyst. Although
catalyst reducibility is not the sole determining factor for
catalytic performance, the 20Ni/AZ-2 catalyst showing the
highest reducibility exhibited the best catalytic performance
(Figs. 5 and 6). It is believed that the optimized ZrO2/Al2O3
ratio of the 20Ni/AZ-2 catalyst favorably altered the interaction
between the nickel species and the support, making it more suit-
able for the steam reforming of LNG. Another possible reason
for the enhanced catalytic activity of 20Ni/AZ-2 may be due
to the presence of ZrO2 on the alumina surface. It is likely that
the presence of ZrO2 prevented the growth of metallic nickel
particles during the reduction process through the formation of
a ZrO2–Al2O3 support with a favorable structure (Fig. 4).
Hydrogen production by steam reforming of methane is
closely related to the following two adsorption mechanisms. One
is the dissociate adsorption of methane on the active nickel sur-
face, and the other is the dissociate adsorption of steam on the
active nickel surface or support [16]. The adsorption of steam
takes place competitively on the nickel and support, and zirco-
nia is known to have a high capacity for adsorbing steam. It is
believed that the zirconia in our catalyst system also played a role
in enhancing the spillover of adsorbed steam from the support to
the active nickel. The migrated steam, in turn, enhanced the gasi-
fication of surface hydrocarbons or carbon species, resulting in
an enhanced LNG conversion and hydrogen yield. It appears that
the well-developed and pure tetragonal phase of AZ-2 played
an important role in the adsorption of steam and the subsequent
spillover of steam from the support to the active nickel.
3
(Zr/Al = 0.31) > 20Ni/AZ-1 (Zr/Al = 0.09) > 20Ni/AZ-4
(Zr/Al = 0.45) > 20Ni/AZ-0 (Zr/Al = 0). Among the catalysts
tested, the 20Ni/AZ-2 catalyst showed the best catalytic
performance. These results imply that an optimum ratio of
ZrO2/Al2O3 is required for the maximum production of
hydrogen by steam reforming of LNG. It is concluded that the
Al2O3–ZrO2 (AZ-X) prepared by a grafting method served
as an efficient support for the nickel catalyst in the hydrogen
production by steam reforming of LNG, and that an optimum
ratio of ZrO2/Al2O3 was required for the maximum yield of
hydrogen over 20Ni/AZ-X catalysts.
Fig. 8. Hydrogen yield as a function of zirconium loading over 20Ni/AZ-X
(X = 0–4) catalysts in the steam reforming of LNG at 600 ◦C. The data were
obtained after a 300 min-reaction.

14
J.G. Seo et al. / Journal of Molecular Catalysis A: Chemical 268 (2007) 9–14
[3] K.J. Lee, D. Park, Korean J. Chem. Eng. 15 (1998) 658.
4. Conclusions
[4] K.D. Ko, J.K. Lee, D. Park, S.H. Shin, Korean J. Chem. Eng. 12 (1995)
478.
A series of Al2O3–ZrO2 (AZ-X) supports with various
zirconium loadings were prepared by grafting a zirconium
precursor onto the surface of ␥-Al2O3. Ni/Al2O3–ZrO2 cata-
lysts were then prepared by an impregnation method for use
in the hydrogen production by steam reforming of LNG. The
effect of Al2O3–ZrO2 supports on the performance of the
Ni/Al2O3–ZrO2 catalysts was investigated. It was found that
ZrO2 inhibited the incorporation of nickel species into the lattice
of Al2O3 and prevented the growth of metallic nickel parti-
cles during the reduction process through the formation of a
new ZrO2–Al2O3 composite structure. The crystalline structures
and catalytic activities of the 20Ni/AZ-X catalysts were strongly
influenced by the amount of zirconium grafted. In the hydrogen
production by steam reforming of LNG, LNG conversion and
hydrogen yield showed volcano-shaped curves with respect to
zirconium loading. Both LNG conversion and hydrogen yield
were decreased in the order of 20Ni/AZ-2 (Zr/Al = 0.17) >
20Ni/AZ-3 (Zr/Al = 0.31) > 20Ni/AZ-1 (Zr/Al = 0.09) > 20Ni/
AZ-4 (Zr/Al = 0.45) > 20Ni/AZ-0 (Zr/Al = 0). Among the cat-
alysts tested, the 20Ni/AZ-2 catalyst showed the best catalytic
performance. The well-developed and pure tetragonal phase of
AZ-2 played an important role in the adsorption of steam and
the subsequent spillover of steam from the support to the active
nickel. It is concluded that Al2O3–ZrO2 (AZ-X) prepared by a
grafting method served as an efficient support for the nickel cat-
alyst in the hydrogen production by steam reforming of LNG,
and that an optimum ratio of ZrO2/Al2O3 was required for the
maximum performance of 20Ni/AZ-X catalysts.
[5] Z.-W. Liu, K.-W. Jun, H.-S. Roh, S.-E. Park, J. Power Sources 111 (2002)
283.
[6] T. Takeguchi, Y. Kani, T. Yano, R. Kikuchi, K. Eguchi, K. Tsujimoto, Y.
Uchida, A. Ueno, K. Omoshiki, M. Aizawa, J. Power Sources 112 (2002)
588.
[7] Q. Ming, T. Healey, L. Allen, P. Irving, Catal. Today 77 (2002) 51.
[8] J. Sehested, J.A.P. Gelten, I.N. Remediakis, H. Bengaard, J.K. Nørskov, J.
Catal. 223 (2004) 432.
[9] T. Borowiecki, G. Wojciech, D. Andrzej, Appl. Catal. A 270 (2004) 27.
[10] T. Borowiecki, A. Goiebiowski, B. Stasinska, Appl. Catal. A 153 (1997)
141.
[11] J.S. Lisboa, D.C.R.M. Santos, F.B. Passos, F.B. Noronha, Catal. Today 101
(2005) 15.
[12] S. Natesakhawat, R.B. Watson, X. Wang, U.S. Ozkan, J. Catal. 234 (2005)
496.
[13] H.-S. Roh, K.-W. Jun, S.-E. Park, Appl. Catal. A 251 (2003) 275.
[14] R. Takahashi, S. Sato, T. Sodesawa, M. Yoshida, S. Tomiyama, Appl. Catal.
A 273 (2004) 211.
[15] M.E.S. Hegarty, A.M. O’Connor, J.R.H. Ross, Catal. Today 42 (1998)
225.
[16] Y. Matsumura, T. Nakamori, Appl. Catal. A 258 (2004) 107.
[17] Q.-H. Zhang, Y. Li, B.-Q. Xu, Catal. Today 98 (2004) 601.
[18] M. Souza, D. Aranda, M. Schmal, J. Catal. 204 (2001) 498.
[19] F. Pompeo, N.N. Nichio, O.A. Ferretti, D. Resasco, Int. J. Hydrogen Energy
30 (2005) 1399.
[20] G. Li, W. Li, M. Zhang, K. Tao, Catal. Today 93 (2004) 595.
[21] T. Klimova, M.L. Rojas, P. Castillo, R. Cuevas, Micropor. Mesopor. Mater.
20 (1998) 293.
[22] P. Iengo, M.D. Serio, V. Solinas, D. Gazzoli, G. Salvio, E. Santacesaria,
Appl. Catal. A 170 (1998) 225.
[23] P. Kim, J.B. Joo, W. Kim, H. Kim, I.K. Song, J. Yi, Carbon 43 (2005)
2409.
[24] M.V. Landau, E. Dafa, M.L. Kaliya, T. Sen, M. Herskowitz, Micropor.
Mesopor. Mater. 49 (2001) 65.
[25] T. Yamaguchi, Catal. Today 20 (1994) 199.
[26] T. Ueckert, R. Lamber, N.I. Jaeger, U. Schubert, Appl. Catal. A 155 (1997)
75.
[27] G. Li, L. Hu, J.M. Hill, Appl. Catal. A 301 (2006) 16.
[28] M.H. Youn, J.G. Seo, P. Kim, I.K. Song, J. Mol. Catal. A 261 (2007)
276.
Acknowledgements
The authors wish to acknowledge support from the Seoul
Renewable Energy Research Consortium (Seoul R & BD Pro-
gram) and RCECS (Research Center for Energy Conversion and
Storage: R11-2002-102-00000-0).
[29] P. Kim, Y. Kim, H. Kim, I.K. Song, J. Yi, Appl. Catal. A 272 (2004)
157.
[30] E.D. Dimotakis, T.J. Pinnavaia, Inorg. Chem. 29 (1990) 2393.
[31] A. Corma, V. Fornes, R.M. Aranda, F. Rey, J. Catal. 134 (1992) 58.
[32] J.A. Wang, A. Morales, X. Bokhimi, O. Novaro, Chem. Mater. 11 (1999)
308.
References
[1] J.N. Armor, Appl. Catal. A 176 (1999) 159.
[2] S.W. Nahm, S.P. Youn, H.Y. Ha, S.-A. Hong, A.P. Maganyuk, Korean J.
Chem. Eng. 17 (2000) 288.