Effective hydrogen production from propane steam reforming using M/NiO/YSZ catalysts (M = Ru, Rh, Pd, and Ag)

Article abstract of DOI:10.1016/j.cattod.2017.08.056

This study has investigated the propane steam reforming (PSR) performance of M/Ni/YSZ catalysts (M = Ru, Rh, Pd, and Ag) for effective hydrogen production. To improve the catalytic performance, particularly with respect to the duration of the catalyst, YSZ (yttria-stabilized zirconia) is used as a support and noble metals such as Ru, Rh, Pd, and Ag are introduced as promoters. YSZ is used to suppress carbon deposition as it provides lattice oxygen to carbon coke on the catalyst. Noble metals provide synergy, assisting Ni metal in improving dehydrogenation. Preferentially, the PSR performance of 1.0 wt.% M/Ni/YSZ catalysts are measured to find which noble metal is suitable as a promoter. The result confirms that the Rh component is the best promoter for M/Ni/YSZ. This study also demonstrates the appropriate amount of Rh to use, based on the PSR performance over Ni/YSZ and 0.1, 0.5, and 1.0 wt.% Rh/Ni/YSZ for 100 h. The results confirm that Ni/YSZ becomes deactivated at 76 h, but the catalysts with the Rh promoter have much higher durability. Thus, this study concludes that 0.5 wt.% Rh/Ni/YSZ is the best catalyst because it does not have excessive Rh content and also does not show any discernable decrease in the PSR performance. Moreover, the Rh/Ni/YSZ catalysts had a higher resistance to carbon deposition because of the presence of the Rh promoter, which could improve lattice oxygen mobility of the Rh/Ni/YSZ catalyst.

Full text of DOI:10.1016/j.cattod.2017.08.056

CatalysisTodayxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Effective hydrogen production from propane steam reforming using M/NiO/
YSZ catalysts (M = Ru, Rh, Pd, and Ag)
Younghwan Im, Jae Hyung Lee, Byeong Sub Kwak, Jeong Yeon Do, Misook Kang⁎
Department of Chemistry, College of Natural Sciences, Yeungnam University, Gyeongsan, Gyeongbuk 38541, Republic of Korea
A R T I C L E I N F O
A B S T R A C T
Keywords:
Propane steam reforming
Noble metals
Yttria-stabilized zirconia
Rh/Ni/YSZ
This study has investigated the propane steam reforming (PSR) performance of M/Ni/YSZ catalysts (M = Ru,
Rh, Pd, and Ag) for effective hydrogen production. To improve the catalytic performance, particularly with
respect to the duration of the catalyst, YSZ (yttria-stabilized zirconia) is used as a support and noble metals such
as Ru, Rh, Pd, and Ag are introduced as promoters. YSZ is used to suppress carbon deposition as it provides
lattice oxygen to carbon coke on the catalyst. Noble metals provide synergy, assisting Ni metal in improving
dehydrogenation. Preferentially, the PSR performance of 1.0 wt.% M/Ni/YSZ catalysts are measured to find
which noble metal is suitable as a promoter. The result confirms that the Rh component is the best promoter for
M/Ni/YSZ. This study also demonstrates the appropriate amount of Rh to use, based on the PSR performance
over Ni/YSZ and 0.1, 0.5, and 1.0 wt.% Rh/Ni/YSZ for 100 h. The results confirm that Ni/YSZ becomes deac-
tivated at 76 h, but the catalysts with the Rh promoter have much higher durability. Thus, this study concludes
that 0.5 wt.% Rh/Ni/YSZ is the best catalyst because it does not have excessive Rh content and also does not
show any discernable decrease in the PSR performance. Moreover, the Rh/Ni/YSZ catalysts had a higher re-
sistance to carbon deposition because of the presence of the Rh promoter, which could improve lattice oxygen
mobility of the Rh/Ni/YSZ catalyst.
1. Introduction
generalized for use in households. However, the supply of feed gas can
be limited in places that do not have well-developed infrastructures to
Hydrogen is an important energy carrier because it is eco-friendly
and has a high energy density. Moreover, as the world will soon con-
front a reality in which hydrogen plays a much larger role in society,
supported by the widespread commercialization of hydrogen energy
conversion and storage, the methods of hydrogen production will be-
come more important. Currently, water electrolysis, gasification of coal
and biomass, partial oxidation of hydrocarbons, and steam reforming of
hydrocarbons are some of the commercially methods to obtain hy-
drogen; steam reforming is well known as a method that affords the
most economical way to produce hydrogen [1,2].
transfer or supply feed gas to a fuel cell, for example, rural, wild, and
camping areas. Therefore, propane is one of the most promising can-
didates that can be accepted in these situations because it can be
commonly stored as liquefied petroleum gas (LPG). Hence, studies on
hydrogen production through PSR need to be conducted. Theoretically,
the overall reaction for PSR can be written as follows [10,11]:
C3H8 + 6H2O → 3CO + 10H2.
ΔH = 499 kJ mol−1
The reaction can be typically distinguished as follows:
Steam reforming:
Feed gases for steam reforming are usually light hydrocarbons such
as methane [3], ethane [4], propane [5], butane [6], and oxidized
compounds such as methanol [7], ethanol [8], and dimethyl ether [9].
Particularly, propane has many advantages; for example, it can be used
worldwide as fuel; it has a well-developed infrastructure and relatively
high energy density; and though it is a gas, it is compressible to a
transportable liquid at normal temperature. In addition, compactible
steam reformers for providing hydrogen to fuel cells in households,
rather than industries, should also be developed, as fuel cells can be
CmHn + mH2O → mCO + (m + 1/2n)H2
+ 7H2)
(C3H8 + 3H2O → 3CO
Water gas shift:
CO + H2O → CO2 + H2.
ΔH = −41.1 kJ mol−1
In practice, steam reforming is usually performed at high tempera-
tures, and common catalysts based on Ni/Al2O3 are widely used for
steam reforming. However, these catalysts, such as NiAl2O4, can easily
⁎
Corresponding author.
E-mail address: mskang@ynu.ac.kr (M. Kang).
http://dx.doi.org/10.1016/j.cattod.2017.08.056
Received 5 June 2017; Received in revised form 19 August 2017; Accepted 25 August 2017
0920-5861/©2017ElsevierB.V.Allrightsreserved.
Pleasecitethisarticleas:Im,Y.,CatalysisToday(2017),http://dx.doi.org/10.1016/j.cattod.2017.08.056

Y. Im et al.
CatalysisTodayxxx(xxxx)xxx–xxx
aggregate and deactivate due to carbon deposition and phase transi-
tions. Moreover, carbon deposition tends to be more severe if the feed
gas is heavier. Hence, in industrial operation, excess steam is used to
minimize carbon deposition; however, it inevitably shortens the cata-
lytic life.
treated under air condition at 1000 °C for 6 h. Finally, the catalysts
prepared in this work were 30.0 wt.% NiO/YSZ, 1.0 wt.% Ru/30.0 wt.
% NiO/YSZ, 1.0 wt.% Rh/30.0 wt.% NiO/YSZ, 1.0 wt.% Pd/30.0 wt.%
NiO/YSZ, 1.0 wt.% Ag/30.0 wt.% NiO/YSZ, 0.5 wt.% Rh/30.0 wt.%
NiO/YSZ, and 0.1 wt.% Rh/30.0 wt.% NiO/YSZ, and all the catalysts
were reduced before PSR performance tests and characterizations. To
avoid confusion, these are simplified to Ni/YSZ, 1%Ru/Ni/YSZ, 1%Rh/
Ni/YSZ, 1%Pd/Ni/YSZ, 1%Ag/Ni/YSZ, 0.5%Rh/Ni/YSZ, and 0.1%Rh/
Ni/YSZ, respectively.
To improve the catalytic lifetime, many studies have attempted
experiments such as novel preparation or promotion techniques that
resist catalytic deactivation due to carbon deposition and phase tran-
sition. Studies focusing on adding promoters, such as alkali metals like
K2O [12] or MgO [13], have shown increased carbon coke resistance by
enhancing carbon gasification, but with decreased catalytic perfor-
mance. Other promoters like precious metals such as Ru, Rh, or Pt have
also been studied [14–17], and the use of these precious metals in-
creased hydrogen selectivity but decreased carbon deposition. In ad-
dition to the introduction of promoters, bi- or tri-metallic catalysts have
been researched [18–21]. Natesakhawat et al. [22] reported that the
effect of the lanthanide promoters (La, Ce, and Yb) significantly en-
hanced catalytic activity and stability because the lanthanide helped to
inhibit both the growth of nickel crystallization and the re-oxidation of
metallic nickel sites. Hardiman et al. [23] also found that Co-Ni/γ-
Al2O3 catalysts could perform well at low carbon/steam ratios. Malai-
bari et al. [24] also researched the effects of the element Mo as a
promoter in the Ni/Al2O3 catalyst, studied at a very low temperature of
450 °C. The promoter Mo showed both the effects of reducing the
carbon deposition rate and increasing the catalytic activity.
2.2. Characterizations of Ni/YSZ and M/Ni/YSZ catalysts
Powder X-ray diffraction (XRD) patterns of the catalysts were
measured by an XRD instrument (model from PANalytical, The
Netherlands) using nickel filtered-CuKα radiation (λ = 1.54 Å, 40 kV,
30 mA) in the range from 20° to 80°. The morphologies of catalysts
before and after the 100-h reaction were obtained by transmission
electron microscope (TEM, H-7600, Hitachi, Japan) with accelerated
voltage 120 kV. H2-temperature programmed reduction (H2-TPR) ex-
periments were conducted using a gas adsorption analyzer, BELCAT-M
(BEL Japan Inc., Japan). Here, 0.1 g of catalyst was pre-treated under
50 mL min−1 Ar flow at 300 °C for 1 h and then cooled down to 50 °C.
Finally, the analysis was conducted by increasing the sample tem-
perature from 50 °C to 800 °C, at
a
rate of 10 °C min−1 under
50 mL min−1 of 5.0 vol.% H2/Ar flow. In addition, the C3H8 adsorption
abilities of the catalysts were determined by C3H8-temperature pro-
grammed desorption (C3H8-TPD) using the same instrument, BELCAT-
M. In this experiment, 0.2 g of the catalyst was pre-treated under
50 mL min−1 of 5.0 vol.% H2/Ar flow at 800 °C for 30 min and then
naturally cooled down to 50 °C. Afterwards, 50 mL min−1 of 25.0 vol.%
C3H8/Ar flow was injected for C3H8 adsorption, and the pretreated
catalyst was purged under 50 mL min−1 of He flow for 20 min. Finally,
an analysis was conducted by increasing the sample temperature from
50 °C to 800 °C under 50 mL min−1 of He flow. To determine how much
carbon coke was deposited on the surface of the catalyst after the 100-h
PSR reaction, thermogravimetric analysis (TGA) was carried out using a
thermogravimetric analyzer, TGA N-3000 (Scinco M & T Inc., Korea).
Approximately 0.02 g of the used catalyst for 100 h was loaded onto a
ceramic pan, and the measurement temperature was increased from
40 °C to 800 °C at a heating rate of 10 °C min−1 under 50 mL min−1 of
air flow. X-ray photoelectron spectroscopy (XPS) measurements were
carried out using K-Alpha™ (Thermo Scientific, UK) with a mono-
chromated AlKα X-ray source (1486.6 eV). The powders were pelletized
at 1.2 × 104 kPa for 1 min and the 1.0-mm pellets were stored over-
night at 1.0 × 10−7 Pa to remove the extra gas molecules adsorbed on
the surface of the pellets before XPS measurement. The base pressure in
the system was less than 1 × 10−9 Pa. The experiments were per-
formed using a 200 W source power and an angular acceptance of 5°,
and the analyzer axis was formed at a 90° angle with the specimen
surface. The Shirley function was used to subtract the background for
XPS data analysis.
To achieve a better catalyst performance and a higher hydrogen
yield on propane steam reforming, this study used yttrium-stabilized
zirconia (YSZ) as a catalytic support because YSZ has more lattice
oxygen mobility than Al2O3, and it can suppress severe carbon de-
position by offering lattice oxygen to carbon coke [25,26]. In addition,
YSZ has higher thermal stability, and therefore, it cannot be trans-
formed to another phase during a reforming reaction as opposed to
NiAl2O4, which forms in Ni-based alumina catalysts. For this reason,
much higher catalytic durability is expected with the use of YSZ than
for catalysts using Al2O3 as a support. Moreover, this study also tested
precious metals—Ru, Rh, Pt, and Ag—to find which metal is more de-
sirable as a promoter for the Ni/YSZ catalyst and to determine how
much precious metal should be added to the Ni/YSZ catalyst. The
prepared catalysts are characterized by XRD, TEM, C3H8-TPD, H2-TPR,
TGA, and XPS.
2. Experimental
2.1. Preparations of Ni/YSZ and M/Ni/YSZ catalysts (M = Ru, Rh, Pd,
and Ag)
The Ni/YSZ catalyst was synthesized by the following method. First,
47.734 g of NiCl2·6H2O (as 15 g of NiO, Junsei Co., Japan) was dis-
solved into 450-mL distilled water, and then 35 g of YSZ powder (8%
Yttrium, Sigma Aldrich) was dispersed into the Ni aqueous solution,
while NH4OH was added to adjust the pH to 9, being stirred for 90 min.
This colloidal solution was transferred to 1.0 L of titanium liner and a
thermal treatment was performed at 180 °C for 12 h, with stirring oc-
curring at 200 RPM. Afterwards, this colloidal solution was evaporated
at 80 °C and then the obtained powder was thermally treated under air
condition at 1000 °C for 6 h.
The synthesis of M/Ni/YSZ catalysts was the same as above before
hydrothermal treatment. After the colloidal solutions were evaporated
at 80 °C, the obtained powders (Ni/YSZ) were thermal-treated under air
condition at 500 °C for 4 h. In the next step, 10 g of the dried powders
(Ni/YSZ) were dispersed in the solvent, mixed with 45-mL distilled
water and 5.0-mL NH4OH. Each noble metal salt, RhCl3·xH2O (Daejung
Chemical, Korea), RhCl3·xH2O (Alfa Aesar, USA), PdCl2 (Inuisho
Precious Metals. Co., Japan), or AgNO3 (Junsei Co., Japan) were added
into the colloidal solutions, including Ni/YSZ, with stirring taking place
for 12 h. The solutions were evaporated at 80 °C and then thermally
2.3. Propane steam reforming (PSR) reaction
PSR was carried out in a fixed bed reactor type which designed in
our previous study [27]. 1.5 g was pelleted at 50 MPa for 30 min, and
then, the pelleted catalyst was sieved to the selected size of the catalyst
granules, from 1 mm to 2 mm. Here, 1.0 g of catalyst granules was
packed with 0.25-g ceramic wool to prevent the catalyst from moving in
the fixed bed quartz reactor, which was arranged vertically inside an
electric furnace. All of the catalysts were reduced under 50 mL of 5.0
vol.% H2/Ar flow at 800 °C for 2 h before PSR reaction, and then cooled
down to ambient temperature. The PSR performance was tested at a
steam to carbon ratio of 3.0 with a gas hourly space velocity (GHSV) of
12,000 h−1. The amount of steam was adjusted by controlling the vapor
temperature according to the partial pressure law. The flow rate of the
2

Y. Im et al.
CatalysisTodayxxx(xxxx)xxx–xxx
feed gas was 40 mL min−1 of 25.0 vol.% C3H8/Ar. Ar was used as a
diluent gas with a steam carrier at a flow rate of 70 mL min−1. The
product gases during the PSR reaction were detected using a gas
chromatograph (7200 series, DS Science, Korea), equipped with a
thermal conductivity detector (TCD) and flame ionization detector
(FID). C3H8 conversion and H2 selectivity can be calculated as follows:
at high temperatures did not seem to be preferred compared with Ni/
YSZ at over 700 °C, based on the CO/CO2 ratio. Therefore, it was judged
that the addition of Ru promoter onto Ni/YSZ did not provide any
benefit because there was no notable catalyst improvement [29].
Meanwhile, catalytic improvement seemed to be evident for 1% Rh/Ni/
YSZ and 1% Pd/Ni/YSZ. These catalysts exhibited greater than 70%
C3H8 conversion for the entire temperature range. The absolute abun-
dances of H2, CH4, CO, CO2, and C3H8 detected in outlet gases when the
1% Rh/Ni/YSZ catalyst was used at 650 °C were 10.0, 0.13, 1.60, 1.70,
and 0.0011 μmol, respectively. For comparison, on 1% Pd/Ni/YSZ, the
detected amounts of H2, CH4, CO, CO2, and C3H8 gases were 9.20, 0.11,
1.80, 1.30, and 0.0011 μmol, respectively. Intriguingly, more CH4
outlet gases were produced below 650 °C for both catalysts, compared
to Ni/YSZ. However, upon heating, the quantities of CH4 were pro-
gressively reduced, and the CH4 gases almost completely underwent
reaction at about 700 °C, as predicted by thermodynamics [30,31].
Meanwhile, the C3H8 conversion for 1% Ag/Ni/YSZ did not reach 90%,
even though the reaction temperature was increased to 700 °C, and the
C2H6 gas appeared as an intermediate at 700 °C. This is probably caused
by the fact that the Ag component inhibits CeC bond cracking and
dehydrogenation during the reaction. Thus, it was concluded that the
PSR performance of 1% Rh/Ni/YSZ was better than the other catalysts.
[C3H8]in − [C3H8]out
C3H8 conversion (%)=
× 100%
[C3H8]in
[H2]out
H2 selectivity (%)
× 100%
[allgases]out
∑
3. Results and discussion
3.1. PSR performances of the Ni/YSZ and 1.0 wt.% M/Ni/YSZ catalysts
(M = Ru, Rh, Pd, and Ag)
First, to determine which noble metal promoter was most suitable
for Ni/YSZ in PSR reaction, the PSR performance of the M/Ni/YSZ
catalysts (M = Ru, Rh, Pd, and Ag) was analyzed as a function of the
reaction temperature. The PSR performance for each catalyst is shown
in Fig. 1. The product gases, detected by GC during the PSR reaction,
were H2, CH4, C2H6, CO, and CO2. In Ni/YSZ, C3H8 conversion in-
creased as the reaction temperature increased; however, it did not reach
90% until the reaction temperature reached 650 °C, and the selectivity
to H2 also increased with the same tendency. At 650 °C, the abundances
of outlet gases were as follows: H2 = 9.30 μmol, CH4 = 0.24 μmol,
CO = 1.30 μmol, CO2 = 1.70 μmol, and C3H8 = 0.0083 μmol. The CO-
gases were detected in greater amounts than expected because the WGS
(water gas shift) reaction prefers temperatures lower than 400 °C [28].
The absolute quantities of H2, CH4, CO, CO2, and C3H8 detected in
outlet gases when the 1% Ru/Ni/YSZ catalyst was used at 650 °C were
9.60, 0.20, 1.30, 1.90, and 0.055 μmol, respectively. The catalytic
performance of 1% Ru/Ni/YSZ was nearly similar to Ni/YSZ for the
entire temperature range, even though its C3H8 conversion and H2 se-
lectivity were slightly better than Ni/YSZ. Moreover, its WGS reaction
3.2. Characteristics of the Ni/YSZ and 1.0 wt.% M/Ni/YSZ (M = Ru, Rh,
Pd, and Ag)
Steam reforming reaction basically uses metallic catalysts to dehy-
drogenate hydrocarbons. It is therefore important to know the catalytic
reduction temperature, because the catalytic activity depends on the
reduction temperature. Therefore, the H2-TPR experiments were con-
ducted, and the results are presented in Fig. 2. Ni/YSZ exhibited a ty-
pical reduction shape of NiO, and the curve was relatively weak and
broad from 320 to 600 °C, compared to the other catalysts with noble
metals. Namely, this indicates that several chemical environments of
NiO exist in Ni/YSZ [32]. However, the catalyst that has noble metals as
a promoter exhibited more peaks, which can be distinguished from each
Fig. 1. Outlet gas distribution of PSR performance for Ni/YSZ and 1.0% M/Ni/YSZ catalysts (M = Ru, Rh, Pd, and Ag), and shows the outlet gas selectivity at 650 °C.
3

Y. Im et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 2. H2-TPR profiles of Ni/YSZ and 1.0% M/Ni/YSZ catalysts.
other. The first peak appears below 200 °C, the second peak appears at
around 400 °C, and the last peak appears above 500 °C. In particular,
the second peaks in the M/Ni/YSZs were changed to be higher and
sharper than those for Ni/YSZ, because the change of the second peak
was associated with the reduction in the interaction of NiO with the
noble metal chemical species. Namely, the trace amount of noble metal
addition could affect the reduction properties of NiO by which the
noble metals rearranged the NiO chemical environment [33]. The last
peaks were considered as NiO residue that could not interact with the
noble metals but were strongly influenced by YSZ supporter, and the
peaks appeared at the highest temperature because YSZ gives lattice
oxygen to the NiO species. Finally, the first peaks below 200 °C in 1%
Rh/Ni/YSZ and 1% Pd/Ni/YSZ are related to the reduction in isolated
noble metal oxide [34]. The first peaks of 1% Rh/Ni/YSZ and 1% Pd/
Ni/YSZ were exhibited at 132 and 117 °C, respectively, but the peaks
were not apparent for 1% Ru/Ni/YSZ and 1% Ag/Ni/NiO. This means
that it was possible that the amount of Rh and Pd components could be
reduced more to optimize the catalyst cost.
The C3H8-TPD catalyst profiles are arranged in Fig. 3. In general, the
chemical adsorption and desorption properties of the catalyst are very
important because the catalytic performance depends on reactant ad-
sorption and product desorption. If the strength is too strong or weak,
the catalytic performance would be poor. Therefore, the adsorption and
desorption strength should be moderate, depending on the catalytic
species and reaction conditions. However, the catalytic performance
can be improved if the amount of adsorption sites with moderate
strength generally increases. The energy of chemical adsorption is as
great as that of physical adsorption, and the energy of physical ad-
sorption is similar to the energy of condensation. Hence, C3H8 deso-
rption over ambient temperature can be deemed as chemical adsorp-
tion, shown in Fig. 3, because the C3H8 boiling point is −42 °C [35]. In
all of the catalysts, the adjacent desorption peaks were fitted by
Gaussian. The 1% Rh/Ni/YSZ catalyst showed the biggest desorption of
C3H8, especially for the fourth peak. However, compared to 1% Rh/Ni/
YSZ, it can be expected that the strength for C3H8 desorption was not
too strong for all the catalysts through the entire measured temperature
range in C3H8-TPD. Generally, the stronger the C3H8 adsorption site is,
the better the PSR performance is. As a result, the distribution of C3H8
desorption over 300 °C was unrivaled in 1% Rh/Ni/YSZ compared to
the other catalyst, and the 1% Pd/Ni/YSZ barely caught up on the 1%
Rh/Ni/YSZ. Conclusively, 1% Rh/Ni/YSZ can be considered as the best
performed catalyst among the catalysts tested, so more experiments
were conducted to study the quantitative effects of Rh promoter by
synthesizing 0.1% and 0.5% Rh/Ni/YSZ.
3.3. PSR performance of the Ni/YSZ and x wt.% Rh/Ni/YSZ catalysts
(x = 0.1, 0.5, and 1.0)
Before checking the lifetime of the catalyst, the catalytic PSR per-
formance based on reaction temperature was tested. As a result, the
performance of the PSR reaction over x wt.% Rh/Ni/YSZ catalysts, as a
function of reaction temperature, was similar. However, as shown in
Fig. 4, the PSR performance at 650 °C for 100 h as a function of the Rh
amount was different. Only product gases H2, CH4, CO, and CO2 were
detected, and other gases did not appear during the PSR reaction, using
the Ni/YSZ and x wt.% Rh/Ni/YSZ. In addition, a significant fluctuation
in the product gases in Ni/YSZ was observed compared to the x wt.%
Rh/Ni/YSZ. It was expected that there were several Ni states in Ni/YSZ;
evidently, its active sites have more states than the x wt.% Rh/Ni/YSZ,
despite the reduction treatment. Therefore, the relatively unstable
surface of the Ni/YSZ caused greater fluctuations in the amount of
emitted H2 [36]. Notably, the Ni/YSZ was completely deactivated at
76 h due to severe carbon deposition along with the catalyst layer, but
the amount of selectivity to H2 and other gases did not noticeably
change (the average value of H2 = 9.5 μmol, δ = 0.60) until the reac-
tion was stopped. The catalytic duration can possibly be lengthened if
severe carbon deposition can be suppressed. The catalysts using Rh
promoter did not show catalyst deactivation during the 100-h reaction,
and the H2 amount increased
a bit (the average value of
H2 = 10.3 μmol and δ = 0.39 in 0.1% Rh/Ni/YSZ; the average value of
H2 = 10.3 μmol and δ = 0.34 in 0.5% Rh/Ni/YSZ; the average value of
H2 = 10.4 μmol and δ = 0.40 in 1.0% Rh/Ni/YSZ), because WGS
4

Y. Im et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 3. C3H8-TPD profiles of Ni/YSZ and 1.0% M/Ni/YSZ catalysts.
equilibrium was improved due to the addition of Rh promoter. This can
be confirmed by the CO/CO2 ratio. With respect to the CH4 abundance,
Rh/Ni/YSZ had a smaller amount of CH4 compared to Ni/YSZ, because
the added Rh promoter, which is well-known as a good catalyst, as-
sisted in CH4 steam reforming.
the 100-h reaction seemed to be MWCNT and an aggregated carbon.
The MWCNT over Ni/YSZ was much thicker than the one over 0.5%
Rh/Ni/YSZ, and the more aggregated carbon deposition also appeared
on Ni/YSZ. From this result, it was confirmed that the catalyst with Rh
promoter generated less carbon deposition during PSR reaction.
Fig. 7 presents the TGA results to quantitatively examine how the
carbon species were deposited on the catalysts during the 100-h reac-
tion. Ni/YSZ showed a 28.7% weight loss, which is similar to the CNT
decomposition at 550 °C without any weight loss before 500 °C [39]. In
addition, the other Rh/Ni/YSZ catalysts also experienced weight loss at
the same temperature of 550 °C. However, there was an unusual 2%
weight gain in the catalysts with Rh promoter at 550 °C, due to the
oxidation of the catalysts. It is believed that there was no weight in-
creases in Ni/YSZ because Ni/YSZ was slowly oxidized during PSR re-
action. Moreover, the extent of the weight loss was different based on
the amount of Rh promoter present. The weight loss regarding CNT
decomposition was 19.1% in 0.1% Rh/Ni/YSZ, and the weight losses of
0.5% Rh/Ni/YSZ and 1.0% Rh/Ni/YSZ were 14.5% and 13.2%, re-
spectively. As expected, the growth of carbon deposition easily de-
creased with increase in the Rh content. However, the suppression of
carbon deposition was not directly proportional to Rh content, as the
weight loss gap was just 1.3% between 0.5% Rh/Ni/YSZ and 1.0% Rh/
Ni/YSZ. Nevertheless, the conclusion can be drawn that 0.5% Rh pro-
moter is sufficient to suppress carbon deposition.
To determine the temperatures at which catalyst reduction and
C3H8 gas adsorption can occur, based on the amount of Rh promoter,
the H2-TPR and C3H8-TPD experiments were conducted and the results
are presented in Fig. 5. In H2-TPR, there was no significant difference
with respect to Rh amount, but the first peak indicated that the isolated
rhodium oxide chemical species disappeared as the amount of Rh ad-
dition was decreased. Namely, the addition of less than 0.5 wt.% Rh
was the most efficient amount, without creating an excess amount of
Rh. Additionally, in C3H8-TPD, the distribution of C3H8 desorption of
0.5% Rh/Ni/YSZ at 300 °C was more distinguishable than 1.0% Rh/Ni/
YSZ. The positions of the fourth peaks in 0.1% and 0.5% Rh/Ni/YSZ
were shifted towards higher temperatures, and the amounts of C3H8
desorption in the fourth peak were decreased as the amount of Rh
promoter was decreased. Consequently, the strongest C3H8 adsorption
site corresponded with Rh promoter. In other words, the number of
relatively strong C3H8 adsorption sites was in direct proportion to the
abundance of Rh promoter, but the strength of C3H8 adsorption sites
was associated with the appropriate Rh addition ratio [37].
Fig. 6a shows the XRD patterns of Ni/YSZ, 0.1%, 0.5%, and 1.0%
Rh/Ni/YSZ after the 100-h reaction, and Fig. 6b displays the TEM
images of Ni/YSZ and 0.5% Rh/Ni/YSZ before and after the 100-h re-
action. Unlike the results in which only rhombohedral NiO (JCPDS 44-
1159) and tetrahedral YSZ (JCPDS 48-0224) were observed in all the
catalysts before the reaction, the reduced metal species were together
observed after the reaction. In Fig. 6a, the Rh/Ni/YSZ catalysts, having
Rh promoter, exhibited two structural peaks which were assigned to
YSZ and metallic Ni (JCPDS 04-0850) located at 44.52° (111) and
52.87° (200). However, in Ni/YSZ, there were other peaks assigned to
NiO and graphite (JCPDS 00-041-1487) at 26.43° (002). Here, the
graphite peak is associated with the layer distance of multilayer carbon
nanotube (MWCNT) [38], indicating that Rh promoter could prevent
the oxidization of catalytic species during reaction, and carbon de-
position appeared as a MWCNT but was suppressed due to Rh promoter.
The TEM images of Ni/YSZ and 0.5% Rh/Ni/YSZ before and after the
100-h reaction were seen in Fig. 6b. Ni/YSZ and 0.5% Rh/Ni/YSZ had
similar shapes and dimensions without notable difference. However,
the growth and shape of carbon species were quite different for both
catalysts after the 100-h reaction. The generated carbon species during
To consider how the oxidation state of the catalytic species changed
before and after the 100-h PSR reaction, the element peaks for Zr3d, O1s
,
Ni2p, C1s, and Rh3p in Ni/YSZ and 0.5% Rh/Ni/YSZ were surveyed by
XPS, as shown in Fig. 8. The reduced metallic peaks for both catalysts
were observed because of the pre-reduction treatment. However, the
divergence between the Ni/YSZ and 0.5% Rh/Ni/YSZ were apparent
after the 100-h PSR reaction. Both of the catalysts seemed to be oxi-
dized after the 100-h PSR reaction, but the extent to which this oc-
curred varied. First, the C1s peaks in Ni/YSZ and 0.5% Rh/Ni/YSZ
correspond to a graphitic compound; thus, carbon deposition mainly
consisted of MWCNT. In addition, the intensity of C1s showed the
amount of carbon deposition. The O1s peaks in both catalysts were al-
most the same, exhibiting a typical O1s peak in metal oxide before re-
action. However, the peaks were changed to the O1s locations in hy-
droxide by being oxidized because H2O was supplied onto YSZ or Ni
metal. Finally, a hydroxyl group remained on the catalyst surface.
Likewise, the Zr3d peak was assigned to the reduced ZrO2 at 182.0 eV
for both catalysts of Ni/YSZ and 0.5% Rh/Ni/YSZ, and the peak posi-
tions at 183.3 and 183.0 eV of the small amount of oxidation were seen,
5

Y. Im et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 4. Outlet gas distribution of PSR performance over Ni/YSZ and 0.1%, 0.5%, and 1.0% Rh/Ni/YSZ catalysts for 100 h.
respectively. This means that ZrO2 could not be influenced by Rh
content primarily but supplied lattice oxygen to an outer layer element
like Ni with similar oxygen mobility, as the binding energy of Zr3d
would be distinguished. The Ni2p peaks changed depending on the Rh
content. The position of the Ni2p peak before reaction was located at
852.3 eV, but the positions of Ni2p peaks after the 100-h reaction were
seen at 854.4 and 853.2 eV in Ni/YSZ and 0.5% Rh/Ni/YSZ, respec-
tively. This indicates that the ratio of NiO/Ni in 0.5% Rh/Ni/YSZ was
less than that of Ni/YSZ [40]. Conclusively, as mentioned in Fig. 2, the
NiO in the Ni/YSZ was more difficult to be reduced and not strongly
influenced by YSZ supporter, compared to 0.5% Rh/Ni/YSZ, but Fig. 8
showed that some of the NiO content in Ni/YSZ had more strongly
bound lattice oxygen atoms than 0.5% Rh/Ni/YSZ. In other words, the
Rh/Ni/YSZ catalyst with the Rh promoter could possibly have fast
oxygen mobility, and it eventually could cause greater CeC cracking
ability and less carbon deposition [41]. Consequently, the PSR perfor-
mance on the Rh/Ni/YSZ catalyst improved.
6

Y. Im et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 5. H2-TPR and C3H8-TPD profiles of Ni/YSZ and 0.1%, 0.5%, and 1.0% Rh/Ni/YSZ catalysts.
4. Conclusions
the growth of carbon deposition by supplying lattice oxygen to carbon
coke on the surface and enhanced CeC cracking during PSR. In addi-
tion, the use of noble metals created a synergy effect by helping Ni
metal to improve dehydrogenation, achieving much longer catalytic
This study has focused on the PSR performance of M/Ni/YSZ
(M = Ru, Rh, Pd, and Ag) catalysts. Using YSZ as a support suppressed
Fig. 6. (a) XRD patterns of Ni/YSZ and 0.1%, 0.5%, and 1.0% Rh/Ni/YSZ catalysts after the 100-h PSR reaction and (b) TEM images of Ni/YSZ and 0.5% Rh/Ni/YSZ before and after the
100-h PSR reaction.
7

Y. Im et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 7. TGA profiles of Ni/YSZ and 0.1%, 0.5%, and 1.0% Rh/Ni/YSZ catalysts after the 100-h PSR reaction.
durability. The noble metals of Ru, Rh, Pd, and Ag were proposed, and
then the Rh promoter was selected as a prospective candidate, de-
pending on the PSR performance. As a result, 1.0% Rh/Ni/YSZ ex-
hibited 74.4% of H2 selectivity and 99.9% of C3H8 conversion at 650 °C.
The 1.0% Rh/Ni/YSZ catalyst had the largest C3H8 desorption area and
distribution over 300 °C, and the Rh promoter was believed to most
strongly adsorb C3H8, depending on the PSR performance. To minimize
the amount of Rh promoter, 0.1% and 0.5% Rh/Ni/YSZ were
Fig. 8. XPS of Ni/YSZ and 0.5% Rh/Ni/YSZ catalysts before and after the 100-h PSR reaction.
8

Y. Im et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Scheme 1. Expected mechanism for suppressing
carbon deposition on the Rh/Ni/YSZ catalyst.
[11] M.A. Rakib, J.R. Grace, C.J. Lim, S.S.E.H. Elnashaie, B. Ghiasi, Int. J. Hydrogen
Energy 35 (2010) 6276–6290.
[12] A. Carrero, J.A. Calles, A.J. Vizcaíno, Chem. Eng. J. 163 (2010) 395–402.
[13] A.J. Vizcaino, A. Carrero, J.A. Calles, Catal. Today 146 (2009) 63–70.
[14] A. Kaftan, M. Kusche, M. Laurin, P. Wasserscheid, J. Libuda, Appl. Catal. A: Gen.
201 (2017) 169–181.
[15] Y. Li, X. Wang, C. Song, Catal. Today 263 (2016) 22–34.
[16] I.N. Buffoni, M.N. Gatti, G.F. Santori, F. Pompeo, N.N. Nichio, Int. J. Hydrogen
Energy 42 (2017) 12967–12977.
[17] H. Iida, N. Onuki, T. Numa, A. Igarashi, Fuel Process. Technol. 142 (2016) 397–402.
[18] A.J. Vizcaíno, A. Carrero, J.A. Calles, Fuel Process. Technol. 146 (2016) 99–109.
[19] H. Wu, G. Pantaleo, V.L. Parola, A.M. Venezia, X. Collard, C. Aprile, L.F. Liotta,
Appl. Catal. B: Environ. 156 (2014) 350–361.
introduced. The 0.5 wt.% Rh was observed to be the optimal amount of
Rh addition to suppress the carbon deposition growth, based on the PSR
performance. A mechanism was expected based on the activity and
physical properties of the catalyst, as shown in Scheme 1. The Rh
promoter strongly affected the chemical environment of NiO, rearran-
ging it to increase reducibility, meaning that oxygen atoms can be ea-
sily provided. Even though this rearranged NiO was reduced after the
pre-treatment, Ni could somewhat sustain this chemical environment
derived from the rearranged NiO properties. Therefore, this Ni could
influence the lattice oxygen mobility when lattice oxygen moved along
the Ni metal surface. Hence, Ni/YSZ, in which the less reducible Ni
shows more distribution than 0.5% Rh/Ni/YSZ, could lead to resistance
against carbon deposition.
[20] K.M. Kim, B.S. Kwak, Y. Im, N.-K. Park, T.J. Lee, S.T. Lee, M. Kang, J. Ind. Eng.
Chem. 51 (2017) 140–152.
[21] A. Carrero, J.A. Calles, L.G. Moreno, A.J. Vizcaíno, Catalysts 7 (2017) 55–71.
[22] S. Natesakhawat, R.B. Watson, X. Wang, U.S. Ozkan, J. Catal. 234 (2005) 496–508.
[23] K.M. Hardiman, T.T. Ying, A.A. Adesina, E.M. Kennedy, B.Z. Dlugogorski, Chem.
Eng. J. 102 (2004) 119–130.
[24] Z.O. Malaibari, E. Croiset, A. Amin, W. Epling, Appl. Catal. A: Gen. 490 (2015)
80–92.
Acknowledgement
[25] 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) 558–595.
[26] N. Laosiripojana, S. Assabumrungrat, J. Power Sources 163 (2007) 943–951.
[27] K.M. Kim, B.S. Kwak, N.-K. Park, T.J. Lee, S.T. Lee, M. Kang, J. Ind. Eng. Chem. 46
(2017) 324–336.
[28] D. Devaiah, P.G. Smirniotis, Ind. Eng. Chem. Res. 56 (2017) 1772–1781.
[29] D. Li, K. Nishida, Y. Zhan, T. Shishido, Y. Oumi, T. Sano, K. Takehira, Appl. Clay Sci.
43 (2009) 49–56.
This study was supported by the Human Resource Training Program
for Regional Innovation and Creativity through the Ministry of
Education and National Research Foundation of Korea (NRF-
2015H1C1A1035639), for which the authors are very grateful.
References
[30] B.T. Schädel, M. Duisberg, O. Deutschmann, Catal. Today 142 (2009) 42–51.
[31] C. Resini, M.C.H. Delgado, L. Arrighi, L.J. Alemany, R. Marazza, G. Busca, Catal.
Commun. 6 (2005) 441–445.
[32] H. Roh, K. Jun, S. Baek, S. Park, Catal. Lett. 81 (2002) 147–151.
[33] Z. Wei, J. Sun, Y. Li, A.K. Datye, Y. Wang, Chem. Soc. Rev. 41 (2012) 7994–8008.
[34] M. Jabłońska, Catal. Commun. 70 (2015) 66–71.
[35] S.S. Kum, B.Y. Jo, S.H. Moon, Appl. Catal. A: Gen. 365 (2009) 79–87.
[36] R. Prins, Chem. Rev. 112 (2012) 2714–2738.
[1] J.D. Holladay, J. Hu, D.L. King, Y. Wang, Catal. Today 139 (2009) 244–260.
[2] P. Nikolaidis, A. Poullikkas, Renew. Sustainable Energy Rev. 67 (2017) 597–611.
[3] B.S. Kwak, K.M. Kim, S.W. Jo, J.Y. Do, S. Kang, M. Park, N. Park, T.J. Lee, S.T. Lee,
M. Kang, J. Ind. Eng. Chem. 37 (2016) 57–66.
[4] M. Matsuka, K. Shigedomi, T. Ishihara, Int. J. Hydrogen Energy 39 (2014)
14792–14799.
[5] S.W. Jo, Y. Im, J.Y. Do, N.-K. Park, T.J. Lee, S.T. Lee, M.S. Cha, M.-K. Jeon, M. Kang,
Renew. Energy 113 (2017) 248–256.
[6] H. Jeong, M. Kang, Appl. Catal. B: Environ. 95 (2010) 446–455.
[7] S. Kim, M. Kang, J. Ind. Eng. Chem. 18 (2012) 969–978.
[8] B.S. Kwak, J.S. Lee, J.S. Lee, B.-H. Choi, M.J. Ji, M. Kang, Appl. Energy 88 (2011)
4366–4375.
[37] P.O. Graf, B.L. Mojet, J.G.V. Ommen, L. Lefferts, Appl. Catal. A: Gen. 332 (2007)
310–317.
[38] D.K. Singh, P.K. Iyer, P.K. Giri, Diam. Relat. Mater. 19 (2010) 1281–1288.
[39] N.M. Mubarak, J.N. Sahu, E.C. Abdullah, N.S. Jayakumar, P. Ganesan, Diam. Relat.
Mater. 48 (2014) 52–59.
[40] C.D. Wagner, W.M. Riggs, L.E. Davis, J.F. Moulder, G.E. Muilenberg, Handbook of
X-ray Photoelectron Spectroscopy, Perkin-Elmer co., 1979.
[9] S. Goicoechea, E. Kraleva, S. Sokolov, M. Schneider, M. Pohl, N. Kockmann,
H. Ehrich, Appl. Catal. A: Gen. 514 (2016) 182–191.
[10] G. Kolb, R. Zapf, V. Hessel, H. Löwe, Appl. Catal. A: Gen. 277 (2004) 155–166.
[41] C. Descorme, D. Duprez, Appl. Catal. A: Gen. 202 (2000) 231–241.
9