Effect of the nature of rhodium catalyst supports on initiation of H 2 production during n-butane oxidative reforming at room temperature

Article abstract of DOI:10.1016/j.jcat.2011.12.007

Three rhodium catalysts supported on rare earth oxides with redox properties, Rh/CeO₂, Rh/Pr₆O₁₁, and Rh/Tb ₄O₇, were tested for their ability to trigger oxidative reforming of n-butane at room temper

Full text of DOI:10.1016/j.jcat.2011.12.007

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Journal of Catalysis 287 (2012) 86–92
Contents lists available at SciVerse ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Effect of the nature of rhodium catalyst supports on initiation of H2
production during n-butane oxidative reforming at room temperature
Katsutoshi Nagaoka a, , Katsutoshi Sato a,b, Yusaku Takita a
⇑
a Department of Applied Chemistry, Faculty of Engineering, Oita University, 700 Dannoharu, Oita City, Oita 870-1192, Japan
b Fuel Cell System Group, Energy Technology Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Umezono 1-1-1, Tsukuba 305-8568, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Three rhodium catalysts supported on rare earth oxides with redox properties, Rh/CeO2, Rh/Pr6O11, and
Rh/Tb4O7, were tested for their ability to trigger oxidative reforming of n-butane at room temperature.
Only Rh/CeO2 triggered the oxidative reforming, owing to the heat generated by the spontaneous oxida-
tion of the CeO2ꢀx species produced by prior reduction of the supported catalyst with H2. Although the
three rare earth oxides were reduced to CeO1.91, Pr2O3, and Tb2O3, respectively, by H2 at P873 K,
CeO1.91 was the only reduced oxide that was oxidized upon exposure to O2 at room temperature. These
results indicate that the ability of the reduced oxide support to undergo oxidation at room temperature
was crucial for triggering oxidative reforming of n-butane at room temperature.
Received 19 October 2011
Revised 26 November 2011
Accepted 2 December 2011
Available online 21 January 2012
Keywords:
Oxidative reforming
H2 production
Room temperature
Rhodium
Ó 2011 Elsevier Inc. All rights reserved.
Rare earth oxide
Redox property
OSC
1. Introduction
lysts [8]. To achieve such temperatures, researchers have explored
several strategies, including electrical heating of catalysts [18] and
Hydrogen has the potential to provide electricity much more
efficiently than combustion processes, particularly when it is used
for fuel cells, which convert chemical energy directly into electrical
energy and thus are not limited by Carnot efficiencies. Therefore,
H2 fuel cells are expected to partially solve problems arising from
the limited supply of fossil fuels [1,2]. However, innovations in
fuel-cell catalyst materials, fuel-cell engineering, and the reform-
ing of hydrocarbons to produce H2 are needed if this technology
is to compete with conventional electrical power plants [3,4]. In
this context, the oxidative reforming (OR) of liquid fuels such as
liquefied petroleum gas, gasoline, diesel fuel, and alcohol (all of
which can be easily stored and transported) to produce H2-rich
reformate gas is of great interest [5–12]. Because OR is a combina-
tion of exothermic (combustion) and endothermic (reforming)
reactions [13–16], it is expected to be more suitable than conven-
tional steam reforming for the local production of H2 at filling sta-
tions and at on-site reformers for domestic fuel cells, for which
start-up and shutdown processes are frequently repeated [17].
Efficient start-up of a reformer requires a short start-up time,
overall energy efficiency, and catalyst stability [3]. The most crucial
criterion, however, is the catalytic autoignition temperature that
normally exceeds 473 K even in the presence of noble metal cata-
homogeneous combustion [19,20]. Electrical heating is more prac-
tical because it is safer and does not require a complex start-up
procedure. However, this strategy requires external energy for
heating the catalysts from room temperature to at least 473 K, as
well as adequate time for heating; and the resulting time lapse de-
lays the onset of H2 production.
In previous work, we demonstrated a new Rh/CeO2-catalyzed
cold-start OR process triggered rapidly at room temperature with-
out the need for external energy input or a specialized start-up pro-
cedure (Fig. 1) [17]. This process can be expected to permit the
development of self-sufficient reforming systems for a new gener-
ation of fuel-cell applications. Prior to the reforming reaction, the
Rh/CeO2 catalyst is reduced in a H2 atmosphere at P873 K, and then
OR of n-butane (n-C4H10), the main component of liquid petroleum
gas, is triggered in the presence of the catalyst at room temperature
(ꢁ300 K). The heat produced by re-oxidation of CeO2ꢀx to CeO2 rap-
idly increases the catalyst temperature from room temperature to
the catalytic autoignition temperature (520 K) and drives the over-
all reaction such that 90% of the n-C4H10 is reformed. In subsequent
cycles, OR is repeatedly triggered at room temperature by in situ
reduction of CeO2 by H2 produced during OR, provided that the cat-
alyst is cooled without exposure to air.
In this new catalytic OR process, CeO2 provides oxygen storage
capacity (OSC), that is, it alternately takes up and releases oxygen.
OSC is useful in several other catalytic processes including catalytic
⇑
Corresponding author. Fax: +81 97 554 7979.
E-mail address: nagaoka@cc.oita-u.ac.jp (K. Nagaoka).
0021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.12.007

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
K. Nagaoka et al. / Journal of Catalysis 287 (2012) 86–92
87
(d = 0.5 mm) was inserted into the catalyst bed through the
a-
Catalytic cycle requiring
no external energy
Al2O3 and quartz wool packing. Cylinders of research-grade gas
were used for all experiments. The catalysts were heated in a pure
H2 atmosphere at a rate of 10 K minꢀ1 and kept at 473, 673, 873, or
1073 K for 1 h, and then the reactor was purged with pure Ar for
1 h at 873 K and subsequently cooled to room temperature
(ꢁ300 K). The furnace was switched off and opened, and the quartz
reactor was wrapped with ceramic insulation for subsequent reac-
tion under quasi-adiabatic conditions. The setup for the activity
test has been described in detail elsewhere [17]. The furnace heater
remained switched off from this point on. An n-C4H10/O2/Ar/N2
feed gas mixture (molar ratio = 1:2:7:1; space velocity = 122 L hꢀ1
gꢀ1) was passed over the catalyst at room temperature. We deter-
mined the feed gas composition by assuming the following stoichi-
ometric OR reaction:
Room temperature
O2
Rh
n-C4H10
O2
H2
CO
CeO2-x→CeO2
Reduced
catalyst
Oxidation
and
heat evolution
Rh2O3
CeO2
Rh
Rh
H2 reduction
CeO2-x
CeO2
Oxidative
reforming
H2
Rh
CeO2→CeO2-x
In situ reduction of catalyst
n-C4H10 þ 2O2 ! 4CO þ 5H2
D
H298K ¼ ꢀ316 kJ
ð1Þ
Fig. 1. Schematic of the catalytic system developed for n-C4H10 oxidative
reforming.
The composition of the exit gases was continuously monitored with
a quadrupole mass spectrometer (ANELVA, M-201QA-TDM). After
30 min, the reaction products were also analyzed by means of a
gas chromatograph equipped with a thermal conductivity detector
(TCD; Agilent, 6890 N). After 35 min, we terminated the reaction
by replacing the feed gas with Ar and cooling the catalyst to room
temperature. This feed–purge sequence was repeated five more
times.
oxidation [21–27] and the automotive three-way reaction [28–32].
Our catalytic OR process represents a new application of an oxygen
storage material. In the current study, we investigated the poten-
tial of other rare earth oxides to act as oxygen storage materials
in the OR process. The rare earth elements Ce, Pr, and Tb have tri-
valent and quadrivalent oxidation states, and oxides of these ele-
ments exist as cubic fluorite-type structures (CeO2, Pr6O11, and
Tb4O7) and as hexagonal A- or cubic bixbyite C-type rare earth ses-
quioxide structures (Ce2O3, Pr2O3, and Tb2O3) [33]. Thus, their sin-
gle-component oxides and solid solutions show OSC [24,26,31,34–
37]. We evaluated Pr6O11 and Tb4O7 as well as CeO2 as supports for
the Rh catalyst in the new catalytic OR process, and we determined
the physicochemical and redox properties that make a catalyst
support suitable for the OR process. We also discuss the role of
Rh in the OR process.
2.3. Characterization of the catalysts
The specific surface area of the catalysts after calcination was
determined by the Brunauer–Emmett–Teller method.
We measured CO chemisorption by catalyst samples using the
O2–CO2–H2–CO pulse procedure proposed by Takeguchi et al.
[38]. O2 was fed to each sample at 30 mL minꢀ1 during programed
heating to 573 K. The sample was maintained at the temperature
for 30 min, cooled to room temperature, and flushed with He for
5 min. Following the oxidation, the sample was treated with H2
at 473 K for 1 h and then cooled to 323 K. At the temperature, it
was flushed with He for 30 min and exposed to O2 for 5 min, to
CO2 for 5 min, to He for 5 min, and to H2 for 5 min and then the
sample was purged with He gas for 30 min. Finally, CO chemisorp-
tion was carried out at 323 K in a He stream (30 mL minꢀ1) by
2. Experimental
2.1. Catalyst preparation
We used three rare earth oxide–supported Rh catalysts: Rh/
CeO2, Rh/Pr6O11, and Rh/Tb4O7. Rh supported on Al2O3, which is
irreducible under the OR conditions, was used as a reference cata-
means of a pulsed-chemisorption technique with CO at 8.9
lmol
lyst. CeO2 (JRC-CEO3) and c-Al2O3 (JRC-ALO3) were supplied by the
per pulse. CO/Rh = 1 was assumed to calculate Rh dispersion.
Catalysis Society of Japan. Pr6O11 and Tb4O7 supports were precip-
itated at room temperature from suspensions prepared by adding a
solution of Pr(NO3)3 6H2O or Tb(NO3)3 6H2O to 25% NH3 solution.
The precipitates were kept in suspension by stirring at room tem-
perature overnight and were then filtered, washed with distilled
water, and dried overnight at 353 K. All the rare earth oxide sup-
ports were calcined at 1073 K and wet impregnated with an aque-
ous solution of Rh(NO3)3ꢂnH2O. The final concentration of Rh was
1 wt.%. The impregnated samples were dried at room temperature
and subsequently at 373 K overnight and then calcined at 723 K in
flowing air. The obtained catalyst powders were pressed into pel-
lets at 52 MPa for 5 min. The pellets were crushed and sieved to
Temperature-programed reduction (TPR) measurements were
taken over 200 mg of each catalyst from room temperature to
1273 K (10 K minꢀ1
)
in flowing H2/Ar gas (H2/Ar = 1/19,
30 mL minꢀ1). The H2 consumption was monitored with a TCD.
The uptake of O2 at 323 K on the reduced catalysts was mea-
sured by means of pulse injection. Each catalyst was loaded into
a tubular U-shaped quartz reactor and reduced in pure H2 at 473,
673, 873, or 1073 K, and then the reactor was flushed with Ar at
873 K. Pulses of pure O2 (95.5 lmol per pulse) were injected into
the catalyst bed at 323 K, and the O2 uptake was measured with
a TCD. The pulses were continued until the level of O2 absorption
stabilized, indicating that the maximum amount of O2 had been
absorbed.
obtain grains with diameters between 180 and 250 lm. These
resulting grains were characterized and used in the activity tests.
X-ray diffraction (XRD) analysis was performed with a Rigaku
RINT-2000 X-ray diffractometer with monochromatized Cu-K
a
2.2. Activity tests
radiation. The samples were transferred to the sample stage of
the XRD instrument after reductive treatment at the various
temperatures.
Thermogravimetric (TG) measurements were taken from room
temperature to 1073 K under flowing air with a Rigaku Thermo
plus TG8120. For the TG measurements, we used the samples that
had been used for XRD analysis after reduction at 873 K.
Catalyst (200 mg) was loaded into a tubular quartz reactor
(i.d. = 7 mm). Quartz wool was packed around the catalyst inside
the reactor, and then
the remaining space to reduce heat loss during the catalytic tests
under quasi-adiabatic conditions. K-type thermocouple
a-Al2O3 pellets (d = 1 mm) were packed into
A

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
88
K. Nagaoka et al. / Journal of Catalysis 287 (2012) 86–92
Additionally, a part of O2 must be consumed by oxidation of the Rh
and CeO2–x at that moment. This may explain higher CO concentra-
tion than CO2 concentration at 4 s.
3. Results and discussion
3.1. Activity tests for triggering OR at room temperature
The formation of CH4 at low concentrations indicates that the
methanation of CO or CO2 occurred (Eqs. (5) and (6))[39,40]:
We determined the dry gas product concentrations immedi-
ately after supplying the feed gas at room temperature to Rh/
CeO2 reduced at 873 K (Fig. 2). At the beginning of the process
(4 s), CO and CO2 were the only products, but after that time, the
H2 concentration increased drastically and eventually greatly ex-
ceeded the CO and CO2 concentrations. This H2 concentration in-
crease occurred simultaneously with the increase in the catalyst
bed temperature, indicating that first combustion (Eq. (2)) and
then reforming (Eqs. (3) and (4)) of n-C4H10 were initiated
sequentially:
CO þ 3H2 ! CH4 þ H2O
D
H298K ¼ ꢀ206 kJ
H298K ¼ ꢀ165 kJ
ð5Þ
ð6Þ
CO2 þ 4H2 ! CH4 þ 2H2O
D
Note that the occurrence of methanation is discussed in Section 3.4,
on the basis of results of activity tests. After 14 s, the H2 concentra-
tion increased to 15%, which was 79% of the steady-state concentra-
tion, and the catalyst bed temperature was 1133 K. At that time, the
heats of the exothermic and endothermic reactions, as well as the
external heat losses, were balanced. Note that atomic hydrogen
formed during reduction with H2 and remained on the catalyst,
which might have contributed to the triggering mechanism, was re-
moved by Ar purge at 873 K after H2 reduction. The results shown in
the figure indicate that OR was triggered quickly over reduced Rh/
CeO2 without any external heat input.
n-C4H10 þ 6:5O2 ! 4CO2 þ 5H2O
D
H298K ¼ ꢀ2657 kJ
ð2Þ
ð3Þ
ð4Þ
n-C4H10 þ 4CO2 ! 8CO þ 5H2
n-C4H10 þ 4H2O ! 4CO þ 9H2
D
H298K ¼ ꢀ816 kJ
D
H298K ¼ ꢀ651 kJ
OR was triggered at room temperature over Rh/CeO2 that had
been reduced at P873 K (Table 1). Even in the absence of external
heat input, complete conversion of O2 and 90% conversion of n-
C4H10 were achieved. The H2 yield and formation rate were high
(P79% and 44 L hꢀ1 gꢀ1, respectively). In contrast, OR was not trig-
gered over Rh/Pr6O11, Rh/Tb4O7, or Rh/Al2O3. The catalyst bed tem-
peratures rose only slightly upon exposure to the reaction mixture,
which indicates that only minor oxidation of the catalyst occurred.
Neither oxidation nor reforming of n-C4H10 was observed with
these catalysts. Al2O3 is irreducible under the reduction conditions
we used, and oxidation of metallic Rh alone was insufficient to heat
the catalyst to the catalytic autoignition temperature. We con-
cluded that for Rh/CeO2 reduced at 873 and 1073 K, an additional
reaction, namely oxidation of CeO2ꢀx to CeO2, provided additional
heat to bring the catalysts to the catalytic autoignition temperature
(520 K) [17]. The reason that OR was not triggered at room temper-
ature over Rh/Pr6O11 and Rh/Tb4O7 reduced with H2 was not clear.
To find an explanation, we characterized the supported Rh cata-
lysts as described in Section 3.2.
18
1473
1273
1073
873
(a)
(
(
(
(
) H2
) CO
) CO2
) CH4
16
14
12
10
8
6
673
4
473
2
0
273
0
20
40
60
Time on stream (s)
7
6
5
4
3
2
1
0
1273
1073
873
(b)
(
(
(
(
) H2
) CO
) CO2
) CH4
3.2. Characterization of the catalysts
3.2.1. Physicochemical properties of the catalysts
We determined various physicochemical properties of the sup-
ported Rh catalysts (Table 2). The specific surface areas of the three
rare earth oxide–supported catalysts were smaller than the specific
surface area of Rh/Al2O3; in particular, the specific surface areas of
Rh/Pr6O11 and Rh/Tb4O7 were lower than 10 m2 gꢀ1. The specific
surface area seemed to affect the Rh dispersion, which decreased
with decreasing specific surface area. One may suspect that differ-
ence in catalytic property is ascribed to the difference in specific
surface area of the catalysts. However, even if Rh was supported
on CeO2 of which specific surface area was 3 m2 gꢀ1 and on which
Rh dispersion was 27.0%, OR was triggered at room temperature
over the catalyst reduced at 873 K. These results indicate that cru-
673
473
273
0
2
4
6
8
10
12
Time on stream (s)
Fig. 2. Time dependence of product concentrations in dry gas and catalyst bed
temperature during OR over Rh/CeO2 reduced at 873 K: (a) 0–60 s and (b) 0–12 s.
Table 1
Catalytic activity of supported Rh catalysts reduced at different temperatures in OR, measured after 30 min.
Catalyst
Rh/CeO2
Reduction temperature (K)
Conversion (%)
Yield of products (%)
H2 formation rate (L hꢀ1 gꢀ1
)
n-C4H10
O2
H2
CO2
CO
CH4
873
1073
97
91
100
100
79
83
6
5
84
85
8
1
44
46
Neither n-C4H10 nor O2 was converted over Rh/Al2O3, Rh/Pr6O11, and Rh/Tb4O7 regardless of reduction temperature.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
K. Nagaoka et al. / Journal of Catalysis 287 (2012) 86–92
89
Table 2
800
600
400
200
0
Physicochemical properties of supported Rh catalysts.
(
(
(
(
) Rh/Al2O3
) Rh/CeO2
) Rh/Pr6O11
) Rh/Tb4O7
Catalyst
Specific
surface
areaa
CO adsorptionb
mol gꢀ1
Dispersion
of Rh (%)
(l
)
(m2 gꢀ1
)
Rh/Al2O3
Rh/CeO2
Rh/Pr6O11
Rh/Tb4O7
114
47
8
89
32
13
5
92
33
13
6
2
a
After calcination.
After reduction at 873 K.
b
cial properties of the catalysts for triggering OR are mainly depen-
dent on the kind of the rare earth elements.
3.2.2. Redox properties of the catalysts
473
673
873
1073
As already mentioned (see Fig. 1), the heat produced by oxida-
tion of the catalyst support heated the catalyst to the catalytic
autoignition temperature and initiated OR. Therefore, reduction
and subsequent oxidation of the catalyst supports were key steps
for triggering OR at room temperature. We evaluated the reduction
of the supported catalysts by TPR (Fig. 3). In the TPR profile of Rh/
Al2O3, a tiny broadened peak for H2 uptake due to the reduction of
Rh2O3 was visible at around 463 K. In contrast, larger H2 uptake
peaks compared to the reduction of Rh2O3 were observed for the
other catalysts, which indicate the reduction of the rare earth oxi-
des. In the case of Rh/CeO2, H2 uptake due to reduction of CeO2
started above 773 K. For Rh/Pr6O11 and Rh/Tb4O7, rapid H2 uptake
due to reduction of Rh2O3 and Pr6O11 and Tb4O7 started at temper-
atures below 473 K. These results indicate that Rh/Pr6O11 and Rh/
Tb4O7 were reduced at lower temperatures than Rh/CeO2, and thus
the inability of Rh/Pr6O11 and Rh/Tb4O7 to trigger OR could not be
ascribed to insufficient reduction of the catalytic supports.
Next, we evaluated O2 absorption by the reduced catalysts by
pulsing O2 into the catalyst bed at 323 K (Fig. 4). Because Al2O3 is
irreducible under the reduction conditions employed, the O2
absorption was nearly equal to the theoretical value for the oxida-
Reduction temperature (K)
Fig. 4. O2 absorption capacity at 323 K for supported Rh catalysts reduced at
different temperatures.
CeO2 was reduced to CeO2ꢀx at 873 and 1073 K, in agreement with
the TPR results, and that the oxidation of CeO2ꢀx increased the up-
take of O2 at 323 K.
Although TPR showed that Rh/Pr6O11 and Rh/Tb4O7 were easily
reduced, the uptake of O2 by these catalysts did not exceed
87 l
mol gꢀ1 even after reduction at 873 K. These results indicate
that Pr2O3 and Tb2O3 were not oxidized at 323 K.
3.2.3. Catalyst structures before and after H2 reduction
To elucidate changes in the catalyst structure during H2 reduc-
tion, we measured the XRD patterns of the supported Rh catalysts
before and after reduction (shown for Rh/CeO2 in Fig. 5). All the
samples, fresh and reduced at 473 or 873 K, showed the diffraction
pattern for the fluorite structure (a = b = c = 0.541 nm). Peak shift
due to the change of lattice parameter was not observed. However,
the peak intensity was higher at the higher reduction temperature,
owing to sintering of the material.
tion of Rh to Rh2O3 (73 l
mol gꢀ1). The uptake of O2 on Rh/CeO2 re-
duced at 673 K was similar to that observed for Rh/Al2O3, whereas
the uptake of O2 increased drastically for the catalysts reduced at
the higher reduction temperatures. These results indicate that
(d)
(c)
(
) CeO2
(c)
(b)
(a)
(b)
(a)
273
473
673
873
1073
1273
10
20
30
40
50
60
70
80
Temperature (K)
2θ (degrees)
Fig. 3. Temperature-programed reduction profiles for (a) Rh/Al2O3, (b) Rh/CeO2, (c)
Fig. 5. XRD patterns of Rh/CeO2 catalysts reduced at different temperatures: (a)
Rh/Pr6O11, and (d) Rh/Tb4O7.
fresh, (b) 473 K, and (c) 873 K.