Direct decomposition of NO into N2 and O2 on SrFe0.7Mg0.3O3 perovskite oxide

Article abstract of DOI:10.1246/bcsj.80.2039

NO direct decomposition on doped SrFeO₃ perovskite oxide was investigated. The ability of SrFeO₃ for direct decomposition of NO is strongly affected by the dopant in Fe sites. Among the examined dopants and compositions, the highest yield of N₂ was achieved on SrFe _0.7Mg_0.3O₃. When SrFe_0.7Mg _0.3O₃ was loaded with Pt, the N₂ yield further improved, and the light-off temperature fell by 100K. On this catalyst, the yields of N₂ and O₂ were 56 and 35%, respectively, at 1123 K. On the Pt-loaded SrFe_0.7Mg_0.3O₃ catalyst, the NO decomposition rate increased with increase in the NO partial pressure with P_NO^1.31. The presence of oxygen slightly decreased the N₂ yield with P_oo₂^-0.12 Therefore, the effect of oxygen poisoning on NO decomposition upon Pt-loaded SrFe_0.7Mg_0.3O₃ is small. From the result of O_2-TPD, Pt loading possibly weakens the adsorption strength of surface oxygen and enhances NO adsorption. In summary, this study shows that the substitution of Fe with lower valence cation in SrFeO₃ and also loading a small amount of Pt are highly effective for increasing the NO decomposition activity.

Full text of DOI:10.1246/bcsj.80.2039

Ó 2007 The Chemical Society of Japan
Bull. Chem. Soc. Jpn. Vol. 80, No. 10, 2039–2046 (2007) 2039
Direct Decomposition of NO into N2 and O2
on SrFe0:7Mg0:3O3 Perovskite Oxide
ꢀ
Hideharu Iwakuni, Yusuke Shinmyou, Hiroshige Matsumoto, and Tatsumi Ishihara
Department of Applied Chemistry, Faculty of Engineering, Kyushu University,
744 Motooka, Nishi-ku, Fukuoka 819-0395
Received April 27, 2007; E-mail: ishihara@cstf.kyushu-u.ac.jp
NO direct decomposition on doped SrFeO3 perovskite oxide was investigated. The ability of SrFeO3 for direct
decomposition of NO is strongly affected by the dopant in Fe sites. Among the examined dopants and compositions,
the highest yield of N2 was achieved on SrFe0:7Mg0:3O3. When SrFe0:7Mg0:3O3 was loaded with Pt, the N2 yield further
improved, and the light-off temperature fell by 100 K. On this catalyst, the yields of N2 and O2 were 56 and 35%,
respectively, at 1123 K. On the Pt-loaded SrFe0:7Mg0:3O3 catalyst, the NO decomposition rate increased with increase
in the NO partial pressure with PNO1:31. The presence of oxygen slightly decreased the N2 yield with PO ꢁ0:12. Therefore,
2
the effect of oxygen poisoning on NO decomposition upon Pt-loaded SrFe0:7Mg0:3O3 is small. From the result of O2-
TPD, Pt loading possibly weakens the adsorption strength of surface oxygen and enhances NO adsorption. In summary,
this study shows that the substitution of Fe with lower valence cation in SrFeO3 and also loading a small amount of Pt
are highly effective for increasing the NO decomposition activity.
Nitrogen oxides (NOx), which mainly form in lean combus-
tion engines, such as diesel engines, are extremely toxic to the
human body and also harmful to environment, because they
are one of the main sources of both acid rain and photochem-
ical smog. At present, because of the increase in the number of
the diesel engine cars, the amount of NOx emission in urban
area has increased significantly. Several methods have been
proposed for NOx removal.1–3 Among them, the selective re-
duction of NOx by hydrocarbons has been studied extensively,
and various catalysts,4–14 in particular, Cu-ZSM-5, have been
proposed as an active catalyst for this reaction.15–19 However,
due to the low selectivity of the reductant, this method is not
practical. Instead of hydrocarbon reduction, selective reduction
of NOx with urea is now considered the most promising meth-
od for the removal of NOx under oxygen containing atmo-
sphere. However, for this method, a slight excess of urea to
NOx is required and so, environmental problems from urea
or ammonia formed are also possible. On the other hand, direct
decomposition of NO into N2 and O2 (2NO = N2 + O2) is the
most ideal reaction for NOx removal, because the process is
quite simple.20,21 However, it is well known that the formed
oxygen adsorbs strongly on the catalyst, resulting in the deac-
tivation of the catalyst. Some catalysts, such as Cu-ZSM-5,22
Co-ZSM-5 (which contains Co in the framework23), La2O3,24
In our previous study, NO direct decomposition over
LaMnO3 perovskite oxide doped with Ba for the La site and
In for the Mn site has been investigated, and it has been found
that La0:7Ba0:3Mn0:8In0:2O3 perovskite oxide exhibits high ac-
tivity toward NO direct decomposition over 1073 K.28 On the
other hand, NO direct decomposition with Fe4þ, which is an
anomalous valence number of Fe at high temperature, has
hardly been studied. Shin et al. have reported that brownmil-
lerite-like compound of Sr2Fe2O5 exhibits rather high NO de-
composition activity in the range of 973–1173 K.29–31 In addi-
tion, they have also shown that an oxygen-deficient perovskite
of composition SrFeO3ꢁx selectively absorbs NO gas above
approximately 373 K, and analysis of the infrared spectrum
of the system 14NO–15NO–SrFeO3ꢁx has shown that the ab-
sorbed species in SrFeO3ꢁx is possibly the nitrosyl ion (NOꢁ)
bound to the iron ion accommodate in some oxygen vacan-
cies.32 However, except for their IR study on NO adsorption
species on SrFeO3, the number of studies on the NO decompo-
sition activity of oxides consisting of FeIV is limited. Perov-
skite oxide, which contains Fe4þ, is expected to be a new NO
direct decomposition catalyst, since reduction of Fe4þ to Fe3þ
proceeds easily and the removal of the surface oxygen can be
easily achieved at reasonably low temperature. In this study,
effects of various dopants at the Fe site in SrFeO3 on NO de-
composition activity were investigated. NO decomposition
activity was greatly improved by a small amount of dopant
in case of LaMnO3 or BaMnO3 catalyst.
26
Ba/MgO,25 and LaCoO3 based perovskite oxides, are active
toward the direct decomposition of NO. In particular, Teraoka
et al. have reported that La0:8Sr0:2CoO3 is highly active toward
NO decomposition and that the N2 yield is 40% at 1073 K.27
Although the reaction temperature is higher, high NO decom-
position activity is expected on perovskite oxides. High reac-
tion temperatures are sometimes better from the viewpoint of
practical applications, because the negative effects of oxygen,
water, and sulfur compounds seem to decrease with an in-
crease in the temperature.
Experimental
Doped SrFeO3 were prepared by a conventional solid-state re-
action method. The precursor of SrFeO3 was obtained by evapo-
rating an aqueous solution of a calculated amount of Sr(NO3)2,
Fe(NO3)2, and metal nitrates. The mixtures obtained were pre-
calcined in air at 673 K for 2 h to decompose the metal nitrates,
Published on the web October 12, 2007; doi:10.1246/bcsj.80.2039

2040 Bull. Chem. Soc. Jpn. Vol. 80, No. 10 (2007)
NO Direct Decomposition on SrFeO3-Based Oxide
Table 1. NO Direct Decomposition into N2 and O2 over SrFe0:7M0:3O3
Conversion/%
NO
Yield/%
Specific surface
area/m2 gꢁ1
Catalyst
a)
N2
O2
N2O
NO2
SrFe0:7M0:3O3
M = Mg
Sn
Ni
1.7
0.8
1.2
1.3
1.6
0.7
0.9
0.6
1.4
2.0
0.6
0.9
0.8
0.6
86.4
79.1
76.1
77.7
78.1
62.4
60.1
48.2
29.6
26.1
11.1
6.8
47.3
42.7
41.6
41.3
37.5
27.0
25.5
17.3
13.2
11.1
3.8
23.4
19.0
17.2
14.2
11.8
5.2
4.3
2.6
1.7
1.6
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
31.5
30.1
29.5
31.8
33.2
28.6
27.9
22.8
14.0
12.3
5.3
Ce
Zr
Mn
Ga
Co
Ru
Ir
Si
0.5
0.4
0.9
0.4
Ge
Cu
Ti
3.9
3.3
0.8
3.2
8.8
1.8
18.6
3.9
a) Estimated based on material balance of nitrogen. NO = 1%, He balance, W=F ¼
3:0 g s cmꢁ3, temperature: 1123 K.
and then calcined in air at 1473 K for 6 h. X-ray diffraction was
ment pan and evacuated at 773 K for 3 h to remove the adsorbed
water. After evacuation, a background spectrum was acquired at
773 K, and gaseous NO (ca. 10 kPa) was introduced into the
measurement cell at 773 K. NO adsorption was performed at
773 K for 30 min, and the catalyst was cooled to room tempera-
ture. IR spectra of adsorbed NO were measured after evacuation
of the catalyst at the described temperature for 30 min.
preformed on this sample using a Cu Kꢀ line, and the formation
of perovskite phase was confirmed. This catalyst powder was press-
ed into disks, crushed and sieved into 16 to 32 meshes. Pt (1 wt %)
was loaded on SrFe0:7Mg0:3O3 by conventional impregnation with
tetraammineplatinum(II) chloride, 1-hydrate (Pt(NH3)4Cl2 H2O).
ꢂ
Direct decomposition of NO was performed with a convention-
al fixed-bed gas-flow reactor with a quartz glass tube (12-mm di-
ameter). Gaseous mixture of 1% NO diluted with He was fed to
the catalyst bed for NO direct decomposition reactions. The cata-
lyst (1 g) was always set in the reactor by using quartz wool (ca.
0.5 cm in catalyst height), and the feed rate of the reactant was
fixed at W=F ¼ 3:0 g s cmꢁ3, where W and F are the catalyst
weight and the gas flow rate, respectively. Reaction was per-
formed at 773–1123 K, and N2, O2, and fed NO were analyzed
with online gas chromatography with a 5A molecular sieve col-
umn. N2O was analyzed using a Porapak-Q column with a thermal
conductivity detector (TCD). Because NO2 cannot be analyzed
with gas chromatography, NO2 formation rate was estimated by
the amount of nitrogen and oxygen molar amounts before and
after the catalyst bed, namely average of (NO conversion–N2
yield) and (N2 yield–O2 yield) values. In this study, the activity
of the catalyst toward NO decomposition is mainly discussed
based on the N2 yield. It is also noted that the formation of N2O
was not observed in this study. Effects of the coexistence of O2
were measured by mixing 1–5% O2 diluted with He with the
reactant gas. In order to keep the total flow rate of 20 mL minꢁ1
constant, the feed rate of He was varied as a balance gas.
Temperature-programmed desorption (TPD) of NO and O2 was
performed to analyze the adsorption state of O2 and NO. After
evacuation of the catalyst at 773 K for 1 h, the catalyst was ex-
posed to 101 kPa NO or O2 for 1 h, followed by cooling to room
temperature. After evacuation at room temperature for 30 min, the
catalyst was heated at 10 K minꢁ1, and the desorbed gas was
monitored by using a quadrupole mass spectrometer (ANELVA,
AQR-100R).
Results and Discussion
Effects of Dopant in SrFeO3 on NO Direct Decomposi-
tion Activity. Table 1 shows the NO decomposition activities
of SrFeO3 perovskite oxide doped with various cations (Mg,
Sn, Ni, Ce, Zr, Mn, Ga, Co, Ru, Ir, Si, Ge, Cu, and Ti). The
NO direct decomposition activity of SrFeO3 without dopant
was not measured in this study, because it was not stable and
its single phase cannot be obtained. Similar to LaMnO3, the
NO decomposition activity of the SrFeO3 catalyst was strongly
affected by additives. Among the catalysts examined, Mg-
doped SrFeO3 catalyst afforded the highest N2 yield. Other
catalysts, such as Sn, Ni, Ce, and Zr-doped catalysts, also ex-
hibited fairly high N2 formation rate. On the contrary, N2 for-
mation rates on Si, Ge, Cu, and Ti-doped catalysts were small-
er than 4%. Therefore, doping with these cations, which are
mainly tetravalent cations, does not enhance the NO decom-
position activity. BET surface area of the doped SrFeO3 is also
shown in Table 1. Since BET surface area did not significantly
change with the dopant, change in NO decomposition activity
was assigned to the dopant effects. Since it was found that the
Mg-dopant was the most effective for the direct decomposition
of NO, Mg-doped SrFeO3 was further studied in the following
part.
Figure 1 shows the XRD patterns of SrFe1ꢁxMgxO3 cata-
lysts. The X-ray diffraction peaks shifted to a lower angle with
an increase in the amount of Mg, suggesting an increase in the
lattice constant. Considering the effective ionic radius of 6
coordination of Mg2þ is 86.0 pm, which is larger than that of
6-coordinate Fe4þ (72.5 pm) and that of 12-coordinate Sr2þ
Adsorption state of NO was measured with FT-IR (JASCO type
610) with a diffuse reflection unit equipped with a heating unit and
KBr window. The catalyst (ca. 0.15 g) was set into the measure-

H. Iwakuni et al.
Bull. Chem. Soc. Jpn. Vol. 80, No. 10 (2007) 2041
100
80
60
40
20
0
NO conversion
SrFeO3-δ
Sr3Fe2O7
N2 yield
x=40 mol%
x=30 mol%
x=20 mol%
x=10 mol%
O2 yield
10
20
30
40
50
60
70
80
773
873
973
1073
1173
2θ
/degree
Temperature / K
Fig. 1. XRD patterns of SrFe1ꢁxMgxO3 prepared.
Fig. 3. Temperature dependence of NO decomposition ac-
tivity upon SrFe0:7Mg0:3O3 in the range of 773–1123 K
(PNO ¼ 1%, W=F ¼ 3:0 g s cmꢁ3).
100
NO conversion
80
amount of Mg into Fe site is at x ¼ 0:3. As a result, the
optimum amount for Mg doping is 30 mol % of Fe sites in
SrFeO3.
60
Temperature dependence of NO decomposition activity
upon SrFe0:7Mg0:3O3 is shown in Fig. 3. It was observed that
NO decomposition started around 873 K, at which N2 and O2
start to form. N2 and O2 yield monotonically increased with
the reaction temperature. N2 yield reached a value of 47% at
1123 K. Although the contact time is much longer and the NO
concentration is much higher than that of the actual exhaust
gas from engines, it was found that fairly high NO conversion
could be achieved on SrFe0:7Mg0:3O3 catalyst. Therefore,
perovskite oxide of SrFe0:7Mg0:3O3 is an interesting NO de-
composition catalyst, although a high reaction temperature is
required.
N2 yield
40
O2 yield
20
0
0
0.1
0.2
0.3
0.4
x in SrFe1-xMgxO3
Fig. 2. N2 and O2 yield on NO direct decomposition over
SrFe1ꢁxMgxO3 at 1123 K as a function of x value
(PNO ¼ 1%, W=F ¼ 3:0 g s cmꢁ3).
Effects of Pt Loading on SrFe0:7O3 on NO Direct Decom-
position Activity. In order to improve the NO decomposition
activity, loading a small amount of a precious metal is some-
times effective. For example, it has been reported that loading
Pt on TiO2, SiO2, and Al2O3 is effective for increasing NO de-
composition activity.34 It has also been reported that Pt loading
is effective for increasing the NO decomposition activity on a
BaO/Al2O3 catalyst.35 In this study, effects of 1 wt % Pt load-
ing for SrFe0:7Mg0:3O3 on the NO decomposition activity were
also studied. Temperature dependence of NO decomposition
activity upon Pt-loaded SrFe0:7Mg0:3O3 is shown in Fig. 4.
The NO decomposition activity improved greatly due to Pt-
loading. In addition, the N2 yield improved from 47 to 56%,
and the O2 yield increased from 23 to 36% at 1123 K by load-
ing Pt. Therefore, it can be said that the loading Pt is quite
effective for improving the NO decomposition activity of
SrFe0:7Mg0:3O3. At present, a detailed mechanism for the im-
proved NO decomposition activity by Pt loading is not clear;
however, loading Pt did not always improve the NO decom-
position activity. In fact, loading Pt into BaMnO3 simply de-
creased the NO decomposition activity. Therefore, improve-
ment in NO decomposition activity is not a simple summation
of the activity of Pt and SrFeO3 catalyst. It is considered that
there are synergy effects between Pt and SrFe0:7Mg0:3O3. The
light-off temperature of NO decomposition activity also shift-
ed by as much as 100 K to a lower temperature. Effects of
(158 pm),33 it is thought that Mg replaces Fe, but not Sr, in
SrFeO3. Although the formation of a small amount of Sr3-
Fe2O7 phase was observed, the main diffraction peaks were as-
signed to those of SrFeO3. An impurity phase appeared when
the x value was higher than 0.3 and became significant with an
increase in the amount of Mg. Therefore, it seems likely that
the solubility limit of Mg into Fe site of SrFeO3 exists around
x ¼ 0:3 and Mg replaces Fe in SrFeO3 lattice. Valence number
of Fe in SrFeO3 is +4 from the molecular formula, which is an
unusual valence number, and by doping lower valence cations
of Mg, the number of oxygen vacancies is increased for charge
compensation. The Fe4þ ion is not stable at high temperatures,
and so, the solid solubility of Mg into Fe site is not large due to
the low lattice stability. In any case, at x values less than 0.3,
the main phase of Mg-doped sample is a perovskite structure
of SrFeO3.
Figure 2 shows the N2 and O2 yield on NO direct decompo-
sition over SrFe1ꢁxMgxO3 catalyst at 1123 K as a function
of x value. It was observed that the N2 yield as well as O2 yield
increased with an increase in the amount of Mg and the yields
reached a maximum at x ¼ 0:3, which is close to the solid
solubility limit of Mg. After the N2 and O2 yield became satu-
rated at ca. 47 and 23%, respectively, a further increase in
the amount of Mg decreased the N2 and O2 yield. Considering
the solubility limit of Mg, it is seen that the optimum

2042 Bull. Chem. Soc. Jpn. Vol. 80, No. 10 (2007)
NO Direct Decomposition on SrFeO3-Based Oxide
60
10
N2 yield of Pt/SrFe(Mg)O3
N2 yield of SrFe(Mg)O3
50
O2 yield of Pt/SrFe(Mg)O3
µ
40
30
20
10
0
O2 yield of SrFe(Mg)O3
1
∆
=1.31
0.1
773
873
973
1073
1173
10-3
10-2
Temperature / K
log(PNO / atm)
Fig. 4. Temperature dependence of NO decomposition
activity upon 1 wt % Pt/SrFe0:7Mg0:3O3 in the range of
773–1123 K (PNO ¼ 1%, W=F ¼ 3:0 g s cmꢁ3).
Fig. 6. N2 formation rate from NO direct decomposition
over 1 wt % Pt/SrFe0:7Mg0:3O3 as a function of NO partial
pressure at 1123 K (W=F ¼ 3:0 g s cmꢁ3).
100
100
80
60
40
20
0
80
60
40
20
0
NO conversion
N2 yield
0
0.2
0.4
0.6
0.8
1.0
0
1
2
3
4
5
PNO / 10-2 atm
PO2 (%)
Fig. 7. N2 yield from NO direct decomposition over 1 wt %
Pt/SrFe0:7Mg0:3O3 as a function of O2 partial pressure at
1123 K (PNO ¼ 0:5%, W=F ¼ 3:0 g s cmꢁ3).
Fig. 5. NO conversion and N2 yield from NO direct decom-
position over 1 wt % Pt/SrFe0:7Mg0:3O3 as a function of
NO partial pressure at 1123 K (W=F ¼ 3:0 g s cmꢁ3).
reaction conditions on NO decomposition activity are also
reported.
is essentially required in some step for the formation of N2
under NO decomposition, and for this, a high concentration
of surface NO is needed.
Figure 5 shows the N2 formation rate as a function of NO
partial pressure. From our previous study, the N2 formation
rate increases with an increase in the NO partial pressure,
and the rate-determining step might be adsorption or activation
of NO.28 Similar to a Mn-based perovskite oxide, the NO con-
version increases with an increase into the NO partial pressure.
High NO conversion at low NO partial pressure is an important
for a NO decomposition catalyst. However, at 1000 ppm NO,
conversion of NO was still reasonably high (ca. 45%). There-
fore, this SrFeO3-based catalyst also exhibited high NO de-
composition activity in a low PNO range. N2 formation rate
is plotted against NO partial pressure in Fig. 6. The N2 forma-
tion rate monotonically increased with an increase in the log-
arithm of NO partial pressure, and the PNO dependence was
PNO1:31, which is almost the same value with that of La(Ba)-
Mn(In)O3 in our previous study. This suggests that the adsorp-
tion and activation of NO are also important steps on Pt/
SrFe0:7Mg0:3O3 catalyst. Coupling of NO adsorption species
Figure 7 shows the N2 yield as a function of the O2 partial
pressure. In the conventional study, the presence of O2 signifi-
cantly lowered the catalyst activity for NO decomposition, and
the number of the catalyst exhibiting a high NO conversion in
the presence of O2 was quite small. As shown in Fig. 7, the N2
yield decreased from 56 to 42% in the presence of 1% O2;
however, the decrease in the N2 yield was not significant in
the PO range from 1 to 5%. As a result, even under a PO of
2
2
5%, a N2 yield of 35% was obtained. Therefore, the negative
effects of O2 are not serious, and Pt-loaded SrFe0:7Mg0:3O3
is highly tolerant against O2. Figure 8 shows the N2 formation
rate as a function of oxygen partial pressure. The N2 formation
rate slightly decreased with an increase in the PO , and the PO
2
2
dependency was as small as PO ꢁ0:12, which is much smaller
2
than that of the similar perovskite oxides reported.28 Therefore,
the negative effects of O2 are much smaller than that
for other perovskite catalysts, and it is expected that a high

H. Iwakuni et al.
Bull. Chem. Soc. Jpn. Vol. 80, No. 10 (2007) 2043
10
12
8
µ
X=0.4
X=0.3
X=0.2
X=0.1
∆
= −0.12
1
4
0
273 373 473 573 673 773 873 973 1073
Temperature /K
0.1
0.01
0.1
log(PO2 / atm)
Fig. 10. O2 desorption profiles from SrFe1ꢁxMgxO3 (O2
adsorption: 773 K for 1 h).
Fig. 8. N2 formation rate from NO direct decomposition
over 1 wt % Pt/SrFe0:7Mg0:3O3 as a function of O2 partial
pressure at 1123 K (PNO ¼ 0:5%, W=F ¼ 3:0 g s cmꢁ3).
the conventional NO decomposition catalyst. Considering the
BET surface area of the present catalyst, NO decomposition
reaction at a large space velocity might be diffusion limited,
and so, by increasing the surface area, it is thought that the
NO decomposition activity should improve. Long-term stabili-
100
80
60
40
20
0
ty of the NO decomposition activity over Pt-loaded SrFe0:7
-
Mg0:3O3 catalyst was also examined, and it was confirmed that
the NO decomposition activity decreased negligibly up to 90 h
after the reaction started. These results indicate that Pt-loaded
SrFe0:7Mg0:3O3 catalyst is active and good NO decomposition.
Effects of Mg Dopant and Pt Loading to O2 and NO Ad-
sorption State. Since the NO decomposition activity was
greatly improved by doping Mg for Fe site and Pt loading,
effects of the dopant and metal addition on NO and O2 ad-
sorptions were investigated by O2 and NO-TPD. Normally, a
higher NO decomposition activity tends to be achieved on a
catalyst that weakly absorbs oxygen. In this study, change in
oxygen adsorption state was investigated by O2-TPD.
Figure 10 shows the O2-TPD curves on SrFe1ꢁxMgxO3.
Oxygen desorption was observed at temperature higher than
523 K and desorption peaks at 573, 773, and 973 K, were ob-
served on all catalysts. Comparing the desorption temperature
of oxygen from the Mn-based perovskite oxide,28 O2 desorp-
tion temperature from SrFeO3 is much lower, and the amount
is also larger. This suggests that the strength of surface oxygen
on SrFeO3 is much weaker than that on LaMnO3. The desorp-
tion temperature of oxygen around 573 K did not change with
an increase in the Mg; however, the amount of total oxygen
desorption increased with an increase in the Mg amount and
reached a maximum at x ¼ 0:3, at which the highest NO de-
composition activity is achieved. Although quantitative analy-
sis by mass spectrometry is not precise, it is noted that the
peak area of TPD peaks from 323 to 1023 K was ca. 4% larger
by addition of 30% Mg to Fe site of SrFeO3. Therefore, the
high NO decomposition activity on SrFe0:7Mg0:3O3 can be
explained by how easily the surface O2 desorbs. The small
0
10000
20000
30000
40000
SV / h-1
Fig. 9. N2 yield from NO direct decomposition over 1 wt %
Pt/SrFe0:7Mg0:3O3 as a function of space velocity at 1123
K (PNO ¼ 1:0%).
NO decomposition activity will be maintained for a longer pe-
riod on this Pt-loaded SrFe0:7Mg0:3O3 catalyst in the presence
of O2.
In order to use a catalyst in an actual NO removal processes
from automobile exhaust gas, a high NO removal rate should
be maintained at a high space velocity. Since the NO decom-
position reaction is considered a rather slow reaction, com-
pared to the selective reduction of NO, the NO decomposition
activity is generally lower at an elevated space velocity. In ad-
dition, BET surface area of perovskite oxide is small, that is,
a few m2 gꢁ1, because high-temperature calcination is needed
to obtain a single phase. In fact, the BET surface area of
SrFeO3-based oxide is less than 4 m2 gꢁ1 as shown in Table 1.
Figure 9 shows the N2 yield as a function of the space velocity
of NO. N2 formation rate decreased with an increase in the
space velocity; however, a reasonably high N2 formation rate
was sustained at a high space velocity of NO, considering the
BET surface area was about 1.7 m2 gꢁ1. In addition, a N2 yield
of 20% was still obtained at SV = 40000 hꢁ1. Although the
high activity is needed even at a few 100000 hꢁ1 for a com-
mercial de-NOx catalyst, the NO decomposition activity of
Pt-loaded SrFe0:7Mg0:3O3 catalyst is reasonably high among
PO dependence of NO decomposition activity is also in good
agreement with these results.
2
In contrast to oxygen desorption, NO desorption was not
significantly affected by doping Mg for Fe site in SrFeO3.
Figure 11 shows the NO desorption curves from SrFe0:7
Mg0:3O3. As shown in Fig. 11, O2 and N2 formation was also
-

2044 Bull. Chem. Soc. Jpn. Vol. 80, No. 10 (2007)
NO Direct Decomposition on SrFeO3-Based Oxide
10
4
8
NO
Pt / SrFe0.7Mg0.3O3
3
2
6
4
O2
N2
2
1
0
SrFe0.7Mg0.3O3
0
273 373 473 573 673 773 873 973 1073
Temperature / K
273 373 473 573 673 773 873 973 1073
Temperature /K
Fig. 11. NO desorption profiles from SrFe0:7Mg0:3O3 (NO
adsorption: 773 K for 1 h).
Fig. 12. O2 desorption profiles from 1 wt % Pt/SrFe0:7
Mg0:3O3 and SrFe0:7Mg0:3O3 (O2 adsorption: 773 K for
1 h).
-
observed in NO-TPD. It was also observed that the molecular
NO desorption occurred around 573 K, at which a fairly large
amount of N2 desorption was also observed. However, no oxy-
gen desorbed at this temperature. This suggests that the formed
oxygen was trapped on the surface of SrFeO3-based oxide
around 573 K. This trapped oxygen started to desorb above
773 K, followed by N2 desorption. Therefore, it is thought that
some amount of surface NO decomposed as the temperature
increased and adsorption of NO is dissociative. Considering
desorption of molecular NO and N2 but not O2, NO dispropor-
tionation reaction (for example; 3NO = N2 + NO3ad) occurs
around 573 K, and the nitrogen oxide surface species may de-
compose at temperatures higher than 873 K. Therefore, desorp-
tion of O2 and N2 is observed at temperatures higher than
873 K. Formation of NO3 surface species has also been ob-
served on LaMnO3 oxide.28 This means that the surface of
SrFe0:7Mg0:3O3 is covered with oxygen formed by the decom-
position of NO around 573 K. However, once the surface oxy-
gen is removed, i.e., by decomposition of NO3, NO direct de-
composition seems to proceed, because N2 desorption was ob-
served after O2 desorption was observed around 973 K. There-
fore, SrFe0:7Mg0:3O3 exhibits high activity toward NO direct
decomposition at temperatures higher than 973 K due to de-
sorption oxygen. If the adsorbed oxygen can be removed, then
it can be expected that direct decomposition of NO will pro-
ceed at a temperature around 773 K. In any case, doping with
Mg at the Fe sites of SrFeO3 makes oxygen desorption easier.
Pt loading was also studied by O2 and NO-TPD. Figure 12
shows the comparison of oxygen desorption profiles on
SrFe0:7Mg0:3O3 with and without Pt loading. Obviously, the
desorption temperature of oxygen shifted to a lower tempera-
ture, and oxygen desorption was observed at temperatures as
low as 473 K. In addition, large oxygen desorption peaks were
also observed above 873 K, which were not observed on the
catalyst without Pt. The total area of O2-TPD peak on Pt
loading one was 1.13 times larger than that of non Pt-loading
catalyst. Therefore, it is clear that Pt loading is effective for
decreasing oxygen desorption temperature and also enlarging
the desorption amount. Since Pt is well-known reduction cata-
lyst, small amount of surface Pt may work as the reduction
10
8
6
NO
4
N2
O2
2
0
273 373 473 573 673 773 873 973 1073
Temperature / K
Fig. 13. NO desorption profiles from 1 wt % Pt/SrFe0:7
Mg0:3O3 (NO adsorption: 773 K for 1 h).
-
catalyst and is effective for removing the surface oxygen. This
means that the surface Pt might be active to break the metal–
oxygen bonds on the surface of SrFeO3. Figure 13 shows the
NO desorption profiles from Pt-loaded SrFe0:7Mg0:3O3 cata-
lyst. Similar to the profile for Mg doping, there were no signifi-
cant changes in NO desorption profiles of Pt-loaded samples,
except for a decrease in NO desorption amount and slight
increase in oxygen desorption amount. Since the amount of
desorbed N2 and O2 is slightly improved by loading Pt, the de-
crease in the amount of NO desorption (the peak area is almost
half of non Pt-loaded one) suggests that Pt causes the NO dis-
sociation on the surface and the amount of molecular NO ad-
sorption decreases. However, amount of surface activated NO
seems to increase with Pt loading. Therefore, it seems that the
loaded Pt is effective for improving the surface active NO con-
centration. As discussed earlier, NO adsorption and activation
seems to be important factors for NO decomposition activity
on this catalyst, and so, Pt has a positive effect on the NO de-
composition activity, because oxygen desorbs more readily
and thus active sites are recovered.
NO adsorption state on Pt-loaded SrFe0:7Mg0:3O3 catalyst

H. Iwakuni et al.
Bull. Chem. Soc. Jpn. Vol. 80, No. 10 (2007) 2045
a small amount of Pt. The N2 yield increased from 47 to 56%
and the light-off temperature of NO decomposition activity
shifted by as much as 100 K to a lower temperature.
0.4
(f)
0.2
(e)
On Pt-loaded SrFe0:7Mg0:3O3 catalyst, the NO decomposi-
tion rate increased with an increase in the NO partial pressure
with PNO1:31. Therefore, the key step for NO decomposition on
Pt-loaded SrFe0:7Mg0:3O3 seems to be related with the adsorp-
tion or activation of NO. The presence of oxygen decreased
(d)
(c)
0
-0.2
-0.4
-0.6
the N2 yield with PO ꢁ0:12; however, a N2 yield of 35% was
2
(b)
(a)
afforded even in the presence of 5% O2 at 1123 K. Therefore,
oxygen scarcely affects the decomposition reaction, and the
effect of oxygen poisoning on NO decomposition upon Pt-
loaded SrFe0:7Mg0:3O3 is small. From the O2-TPD results, Pt
loading is thought to weaken the adsorption strength of the sur-
face oxygen and to increase the decomposition of nitrate sur-
face species. In summary, this study shows that substituting
some of the Fe with a cation in a lower valence state in SrFeO3
and also loading a small amount of Pt are highly effective for
increasing the NO decomposition activity.
1450
1225
1050
850
1550
1780
3000
2500
2000
1500
1000
Wavenumber / cm-1
Fig. 14. IR spectra of adsorbed NO on 1 wt % Pt/SrFe0:7
-
Mg0:3O3: (a) NO adsorption at 773 K, (b) evacuation at
room temperature, (c) evacuation at 473 K, (d) evacuation
at 573 K, (e) evacuation at 673 K, (f) evacuation at 773 K.
was studied. Figure 14 shows the IR spectra of NO after ad-
sorption at 773 K. After adsorption of NO at 773 K, absorption
peaks at 1780, 1550, 1450, 1225, 1050, and 850 cmꢁ1 were
observed. In comparison to the reported values,28 the peak at
1780 and those around 1500 cmꢁ1 were assigned to bent-type
NO (NOþ) and nitrate oxides, such as NO3 species. Therefore,
adsorption of NO seems to be electron donative, i.e., the cata-
lyst is reduced by adsorption of NO, and this is reasonable con-
sidering that Fe4þ is easily reduced to Fe3þ, which is a more
stable oxidation state. Since the absorption peak at 1780
cmꢁ1 disappeared after the sample was placed under a vacuum
at 473 K for 30 min, the desorption peak of NO in NO-TPD
around 623 K was assigned to that of bent-type NO adsorption
species. On the other hand, absorption peaks around 1500
and 1000 cmꢁ1 remained after evacuation at 673 K, but dis-
appeared after evacuation at 773 K evacuation, for 30 min.
Therefore, O2 and N2 desorption peaks observed in NO-TPD
at temperatures higher than 773 K were assigned to the decom-
position products of nitrate spices. Since nitrate species strong-
ly cover the active sites for NO decomposition, removal of
nitrate species from the surface is highly important to achieve
high NO decomposition activity. Pt-loaded SrFe0:7Mg0:3O3
catalyst needs a relatively low temperature for desorption of
nitrate-like surface species resulting in the high NO decom-
position activity in an intermediate temperature range. In any
case, the decomposition of nitrate species, which forms by dis-
proportionation of NO, is important step for NO decomposi-
tion reaction on SrFeO3-based catalyst, and this seems to be
accelerated by addition of small amount of Pt.
References
1
2
3
4
M. Iwamoto, Shokubai 1995, 37, 614.
S. Kagawa, Y. Teraoka, Hyomen 1993, 31, 913.
H. Hamada, Shokubai 1991, 33, 320.
H. Hamada, Y. Kintaichi, M. Sasaki, T. Ito, M. Tabata,
Appl. Catal. 1990, 64, L1.
5 S. Sato, Y. Yu-u, H. Yahiro, N. Mizuno, M. Iwamoto,
Appl. Catal. 1991, 70, L1.
6 K. Yogo, M. Umeno, H. Watanabe, E. Kikuchi, Chem.
Lett. 1993, 131.
7 H. Hamada, Y. Kintaichi, M. Sasaki, T. Ito, T. Yoshinari,
Appl. Catal., A 1992, 88, L1.
8 M. Iwamoto, H. Yahiro, H. K. Shin, M. Watanabe, J. Guo,
M. Konno, T. Chikahisa, T. Murayama, Appl. Catal., B 1994, 5,
L1.
9 K. Yogo, M. Ihara, I. Terasaki, E. Kikuchi, Chem. Lett.
1993, 229.
10 Y. Li, J. N. Armor, Appl. Catal., B 1992, 1, L31.
11 Y. Nishizaka, M. Misono, Chem. Lett. 1993, 1295.
12 A. Obuchi, A. Ohi, M. Nakamura, A. Ogata, K. Mizuno,
H. Ohuchi, Appl. Catal., B 1993, 2, 71.
13 M. Inaba, Y. Kintaichi, H. Hamada, Catal. Lett. 1996, 36,
223.
14 H. Iwakuni, A. Takami, K. Komatsu, Science and
Technology in Catalysis 1998, Proceedings of the Third Tokyo
Conference on Advanced Catalytic Science and Technology,
1999, p. 251.
15 M. Iwamoto, H. Yahiro, S. Shundo, Y. Yu-u, N. Mizuno,
Appl. Catal. 1991, 69, L15.
16 W. Held, A. Koenig, T. Richter, L. Puppe, SAE Tech. Pap.
Ser. 900496.
Conclusion
17 H. Hamada, N. Matsubayashi, H. Shimada, Y. Kintaichi,
T. Ito, A. Nishijima, Catal. Lett. 1990, 5, 189.
18 C. J. Bennett, P. S. Bennett, S. E. Golunski, J. W. Hayes,
A. P. Walker, Appl. Catal., A 1992, 86, L1.
19 K. Kharas, Appl. Catal., B 1993, 2, 207.
20 H. Hamada, Y. Kintaichi, M. Sasaki, T. Ito, Chem. Lett.
1990, 1069.
The NO decomposition activity increased in the order of
Mg > Sn > Ni > Ce > Zr after doping the Fe site in SrFeO3
perovskite oxide. Among the investigated dopants and com-
positions, the highest N2 yield was achieved with SrFe0:7
-
Mg0:3O3. On this catalyst, NO conversion increased with an
increase in the reaction temperature, and conversion of NO
into N2 and O2 at 1123 K was 47 and 23%, respectively.
The NO decomposition activity improved greatly by loading
21 H. Yasuda, N. Mizuno, M. Misono, J. Chem. Soc., Chem.
Commun. 1990, 1094.

2046 Bull. Chem. Soc. Jpn. Vol. 80, No. 10 (2007)
NO Direct Decomposition on SrFeO3-Based Oxide
Mater. Res. Bull. 1979, 14, 133.
30 S. Shin, M. Yonemura, H. Ikawa, Bull. Chem. Soc. Jpn.
1979, 52, 947.
22 M. Iwamoto, H. Yahiro, K. Tanda, N. Mizuno, Y. Mine,
S. Kagawa, J. Phys. Chem. 1991, 95, 3727.
23 Y. F. Chang, J. G. McCarty, J. Catal. 1998, 178, 408.
24 M. A. Vannice, A. B. Walters, X. J. Zhang, J. Catal. 1996,
159, 119.
31 S. Shin, M. Yonemura, H. Ikawa, Mater. Res. Bull. 1978,
13, 1017.
25 S. Xie, M. P. Rosynek, J. H. Lunsford, J. Catal. 1999,
188, 24.
26 Y. Teraoka, H. Fukuda, S. Kagawa, Chem. Lett. 1990, 1.
27 Y. Teraoka, T. Harada, S. Kagawa, J. Chem. Soc., Faraday
Trans. 1998, 94, 1887.
32 S. Shin, H. Arakawa, Y. Hatakeyama, K. Ogawa, K.
Shimomura, Mater. Res. Bull. 1979, 14, 633.
33 R. D. Shannon, C. T. Prewitt, Acta Crystallogr., Sect. B
1969, 25, 925; R. D. Shannon, Acta Crystallogr., Sect. A 1976,
32, 751.
28 T. Ishihara, A. Ando, K. Takiishi, K. Yamada, H.
Nishiguchi, Y. Takita, J. Catal. 2003, 220, 104.
29 S. Shin, Y. Hatakeyama, K. Ogawa, K. Shimomura,
34 X. Wang, S. M. Sigmon, J. J. Spivey, H. H. Lamb, Catal.
Today 2004, 96, 11.
35 Z. Liu, J. A. Anderson, J. Catal. 2004, 224, 18.