In situ infrared spectroscopic studies on the mechanism of the selective catalytic reduction of NO by C3H8 over Ga 2O3/Al2O3: High efficiency of the reducing agent

Article abstract of DOI:10.1021/jp051214b

The partial oxidation of propane and the mechanism of selective catalytic reduction (SCR) of NO by propane over Ga₂O₃/Al ₂O₃ in excess of O₂ have been investigated using in situ Fourier transform infrared spectroscopy. An optimized Ga ₂O₃/Al₂O₃ catalyst shows high activity and efficiency of the reducing agent propane (100% conversion of NO at 623 K, GHSV: 10 000 h^-1). One molecule of propane converts more than 4 NO molecules to N₂. The reaction starts with the partial oxidation of C₃H₈ by O₂ and carboxylates (acetate, formate) are formed on the catalyst surface above 573 K. This oxidation represents the rate-determining step of the SCR reaction. These surface carboxylates represent a dominating intermediate and (easily) react with (adsorbed) NO forming nitrogen-containing organic species. The latter are proposed to react with NO to form N₂. Total oxidation of propane was enhanced at temperatures above 773 K leading to decreased reductant efficiency. Surface nitrite and nitrate species can also be observed, but they were found to be spectators only. This could be concluded from the electron balance (conversion of propane relative to NO) and from the relative rates of the single reaction steps. On the basis of these investigations and stoichiometric calculations, a conclusive reaction mechanism is proposed. ? 2005 American Chemical Society.

Full text of DOI:10.1021/jp051214b

15906
J. Phys. Chem. B 2005, 109, 15906-15914
In situ Infrared Spectroscopic Studies on the Mechanism of the Selective Catalytic
Reduction of NO by C3H8 over Ga2O3/Al2O3: High Efficiency of the Reducing Agent
†
†
‡
‡
§
Chuanhua He, Martin Paulus, Josef Find, Julius A. Nickl, Hans-J u1 rgen Eberle,
§
|
,†
J o1 rg Spengler, Wei Chu, and Klaus K o1 hler*
Department of Chemistry, Technische UniVersit a¨ t M u¨ nchen, Lichtenbergstrasse 4, D-85747 Garching,
Germany, Gesellschaft f u¨ r Werkstoffpr u¨ fung mbH, Georg-Wimmer-Ring 25, D-85604 Zorneding, Germany,
Consortium f u¨ r Elektrochemische Industrie GmbH, Zielstattstrasse 20, D-81379 M u¨ nchen, Germany, and
Chemical Engineering School of Sichuan UniVersity, 610065 Chengdu, People’s Republic of China
ReceiVed: March 9, 2005; In Final Form: June 23, 2005
The partial oxidation of propane and the mechanism of selective catalytic reduction (SCR) of NO by propane
over Ga
An optimized Ga
conversion of NO at 623 K, GHSV: 10 000 h ). One molecule of propane converts more than 4 NO molecules
to N . The reaction starts with the partial oxidation of C by O and carboxylates (acetate, formate) are
O /Al O
2 3 2 3
2
in excess of O have been investigated using in situ Fourier transform infrared spectroscopy.
2
O
3
/Al catalyst shows high activity and efficiency of the reducing agent propane (100%
2 3
O
-1
2
3
H
8
2
formed on the catalyst surface above 573 K. This oxidation represents the rate-determining step of the SCR
reaction. These surface carboxylates represent a dominating intermediate and (easily) react with (adsorbed)
NO forming nitrogen-containing organic species. The latter are proposed to react with NO to form N . Total
2
oxidation of propane was enhanced at temperatures above 773 K leading to decreased reductant efficiency.
Surface nitrite and nitrate species can also be observed, but they were found to be spectators only. This could
be concluded from the electron balance (conversion of propane relative to NO) and from the relative rates of
the single reaction steps. On the basis of these investigations and stoichiometric calculations, a conclusive
reaction mechanism is proposed.
Introduction
mechanism since this information may lead us to the design of
more active catalyst systems.
2
6-28
It is widely accepted that selective catalytic reduction (SCR)
of nitrogen oxides using hydrocarbons in an oxidizing atmo-
sphere is a promising method to remove NOx from lean-burn
and diesel exhausts and could be an interesting cheaper
alternative to ammonia SCR, e.g., for power stations.1 Although
the first catalysts reported and many studies focused on SCR
of NO by hydrocarbons are related to zeolite or metal ion
exchanged zeolite catalysts, metal oxide or supported metal
oxide catalysts are probably the best candidates for this reaction
due to their high activity and stability.3 Among them, Ga2O3/
Al2O3 is a promising candidate for practical applications because
of its high stability and activity even in the presence of water
and SO2.2
Depending on the catalyst, the reducing agent, and reaction
conditions, various reaction mechanisms for the SCR of NO
by hydrocarbons have been proposed in the literature:2
6-37
On
the basis of the knowledge of the mechanism of decomposition
of NO over supported noble metal catalysts, a mechanism that
involves dissociation of NO to N(ads) and O(ads) with subsequent
desorption of N2 or N2O and removal of O(ads) by the reducing
,2
29,30
31
agent has been proposed by Burch et al.,
Acke et al.,
32
33
Wogerbaur et al., and Cho et al. independently. Engler et
al. studied the mechanism of the SCR of NO by hydrocarbons
-21
34
on alumina-supported platinum catalysts and proposed that both
NO and hydrocarbons are first activated by O2 forming adsorbed
surface species (oxygenates, nitrates). Then, these adsorbed
species react with each other to give CO2, N2, and H2O. Haneda
0,21
The most important limitation of Ga2O3/Al2O3 for practical
application is that the active temperature is too high and the
activity-temperature relationship shows a volcano shape which
27
28
et al. and Shimizu et al. investigated the mechanism of SCR
of NO using C3H6 over Ga2O3/Al2O3 and Al2O3 independently.
20,21
results in a small operation temperature range.
Recently,
The authors suggested that NO was first oxidized to NO and
2
-
-
-
several research groups have studied modifications of these
catalysts by changing the catalyst preparation methods, adding
a second or third metal oxide, or mixing different catalysts
2 3
adsorbed as NOx (NO , NO ) species; subsequently, the
adsorbed NOx species were reduced to N by C H or by
-
2
3
6
oxygenates generated from C H . On the other hand, Obuchi et
3
6
3
5
physically in order to increase the activity at lower temperature
al. proposed that carbonaceous or carbon deposits generated
from partial oxidation of the reducing agents play a crucial role
as intermediates in the SCR reaction in their studies on the
reduction of NO using MTBE over γ-alumina. The last
and broaden the temperature range.2
0-25
Alternatively, research-
ers have focused on structure-activity relationships and reaction
36
mechanism is further supported by Shibata et al. and Meunier
*
To whom correspondence should be addressed. Fax: + 49 89 289
3
7
1
3183. E-mail: klaus.koehler@ch.tum.de.
et al. by studies of the SCR of NO using C3H8 and C3H6 over
Ag/γ-Al2O3 and by Radtke et al. by investigations on the SCR
of NO using ethene or propylene over CuOx/γ-Al2O3.38 To
reveal the reaction mechanism of the hydrocarbon SCR reaction,
†
Technische Universit a¨ t M u¨ nchen.
Gesellschaft f u¨ r Werkstoffpr u¨ fung.
Consortium f u¨ r Elektrochemische Industrie.
Chemical Engineering School of Sichuan University.
‡
§
|
1
0.1021/jp051214b CCC: $30.25 © 2005 American Chemical Society
Published on Web 07/28/2005

Selective Catalytic Reduction of NO by C3H8
J. Phys. Chem. B, Vol. 109, No. 33, 2005 15907
it can be extremely important to investigate the formation and
consumption of the possible intermediates and to calculate the
electron balance for each reaction step. However, little detailed
work focused on the stoichiometry of the reaction.
TABLE 1: Composition of Different Gas Mixtures Used in
This Paper
no.
gas mixture
concentration
1
2
3
4
5
6
7
He
O2
NO
pure helium (99.999%)
In the present work, we report a first systematic in situ Fourier
transform infrared (FTIR) study including time-dependent
experiments on the mechanism of the selective catalytic
reduction of NO by C3H8 over Ga2O3/Al2O3. The catalyst has
been optimized, and its catalytic properties have been investi-
gated extensively (conversions as function of various temper-
atures, detailed selectivity based on complete nitrogen and
carbon balance). It is this catalyst Ga2O3/Al2O3 that is particu-
larly interesting and suitable for mechanistic investigations for
several reasons: (i) In comparison to all other systems reported,
it shows the highest efficiency of the reducing agent: one
propane molecule converts (more than) 4 NO molecules
exclusively to N2. This makes the system clearer, and several
reaction mechanisms proposed in the literature can surely be
excluded for Ga2O3/Al2O3 due to the overall electron balance.
10% O /He
2
1000 ppm NO/He
3000 ppm C H /He
3 8
1000 ppm NO + 10% O
C
3
H
8
NO + O
2
2
/He
C
C
3
H
H
8
+ O
+ O
2
3000 ppm C
3000 ppm C
3
H
H
8
+ 10% O
2
/He
3
8
2
+ NO
3
8
+ 10% O
2
+
1
000 ppm NO/He
TABLE 2: Catalytic Performance of Al
2 3 2 3 2 3
O and Ga O /Al O
in SCR of NO by C
3
H
8
a
b
NOCFc
(%)
temperature
(K)
X
NO
X
C3H8
(%)
catalyst
(%)
X
NO:XC3H8
Al
2
O
3
623
673
10.1
14.7
28.4
86.5
93.4
83.5
14.2
28.6
46.5
1.7
3.7
10.1
52.9
95.3
99.3
3.1
5.9
4.0
2.8
1.6
1.0
0.8
4.6
3.4
4.3
4.2
1.0
1.0
0.9
59.4
39.7
28.1
16.4
9.6
7
7
8
8
23
73
23
73
(
ii) It is generally accepted that such oxide systems neither
8.4
26
directly decompose NO nor oxidize NO to NO2. (iii) Neither
Ga2O3 nor Al2O3 are known to be redox-active materials, i.e.,
partially reduced catalyst intermediates can also be excluded.
On the basis of detailed in situ FTIR experiments on the
oxidation of C3H8 in different gas atmospheres and on the
consumption of surface carboxylates and nitrates and by careful
correlation of the results with catalytic properties, a conclusive
reaction mechanism will be presented.
Ga O /Al O
3
573
623
45.8
33.7
43.0
41.5
10.4
9.6
2
3
2
8.5
6
7
7
8
8
73
23
73
23
73
10.8
24.1
91.8
98.5
99.2
100
95.3
95.0
92.1
9.3
a
b
NO conversion: ([NO]inlet - [NO]outlet)/[NO]inlet × 100%.
3 8
C H
c
conversion: ([C
3 8
H ]
inlet - [C
3 8
H
]outlet)/[C
H
3 8
]
inlet × 100%. NO com-
petitiveness factor: ([NO]inlet - [NO]outlet)/{10 × ([C
3
H
8
]
3 8
inlet - [C H ]outlet)}
39
Experimental Section
× 100%.
Catalyst Preparation. The catalysts applied in this work (10
mol % Ga2O3/Al2O3) were prepared by the impregnation
method. A certain amount of γ-Al2O3 (specific surface area,
min) were recorded as a function of time. The composition of
different gas mixtures are listed in Table 1. The short description
in column two ‘‘gas mixture” is used in this paper.
In addition to the steady-state reaction, the time-dependent
experiments were employed to clarify the role and reactivity of
particular surface species. The catalyst was first exposed to either
C3H8 + O2 or NO + O2 at 623 K for 120 min to accumulate
the surface species. After purging in flowing He for 30 min,
the feed gas was switched to various gas mixtures, and the
changes in the intensity of IR bands were measured with time
on the stream.
2
Brunauer-Emmett-Teller: 198 m /g) support was soaked into
a solution of gallium nitrate (Merck, >99%) in a predetermined
amount of H2O and was allowed to stand at room temperature
for 2 h with frequent stirring. Then the water was removed by
heating under reduced pressure. The catalyst precursor thus
obtained was dried at 393 K overnight, followed by calcinations
at 723 K in air for 4 h.
Catalytic Activity Measurements. The catalytic activity was
measured by using a fixed bed-flow reactor. The reaction
mixture consisted of 1000 ppm NO, 1000 ppm C3H8, and 10%
O2 (Helium as balance gas). The gas flow rate GHSV is 10 000
Results
-1
h . The analysis of feed gas and products was carried out with
a mass spectrometer, a CLD NOx analyzer, a NDIR CO
analyzer, and a gas micro chromatograph with a thermal
conductivity detector and Porapack Q and molecular sieve A
as separation columns. The catalytic activity was evaluated in
terms of NO conversion to N2, C3H8 conversion to COx (CO +
CO2), and the NO competitiveness factor.39 The formation of
N2O was found negligible in the present work.
1. Catalytic Performance of SCR of NO by C3H8 over
Al2O3 and Ga2O3/Al2O3. Table 2 shows the NO conversion
for Al2O3 and Ga2O3/Al2O3 catalysts in the SCR of NO by C3H8
as a function of temperature in a fixed bed continuous flow
reactor. Ga2O3/Al2O3 appears to be more active than Al2O3 over
the whole temperature range. Both catalysts show a maximum
in NO conversion (slight volcano shape), which is shifted to
lower temperature for Ga2O3/Al2O3. The conversion of NO
decreased more quickly for Al2O3 for temperatures higher than
the optimum value. The impregnation of Ga2O3 onto Al2O3
reduced the temperature of maximum NO conversion and
broadened the temperature-activity window for the reduction
of NO by C3H8.
In situ FTIR Experiments. Catalyst powder (5 mg) was
pressed into self-supporting wafers with a diameter of 8 mm
2
(
10 mg/cm ) and placed in a high-temperature flow cell with
two pieces of CaF2 windows. IR spectra were recorded on an
FTIR spectrometer (FTS-575C, BIO-RAD) with 16 scans at a
-1
resolution of 2 cm . Prior to each experiment, the catalyst wafer
was first activated in situ by heating in a flow of O2 (total
flow: 50 mL/min) at 673 K for 2 h, followed by cooling to the
desired temperature and purging in He for 30 min. Then, the
spectrum of clean surface was recorded which was used as the
background of the in situ experiment. The IR spectra of surface
species in a flow of various gas mixtures (total flow: 50 mL/
Particular emphasis has been placed on the careful quantitative
determination of all (by)products of the SCR reaction (complete
carbon and nitrogen balance by a complex gas analysis system).
In contrast to NO conversion, the conversion of C3H8 continu-
ously increases for temperatures higher than an optimum value
leading to a reduction of efficiency of the reducing agent (so-
3
9
called NO competitiveness factor, NOCF). At temperatures

15908 J. Phys. Chem. B, Vol. 109, No. 33, 2005
He et al.
TABLE 3: Wavenumber and Assignment of Main
Absorption Bands in the FTIR Spectra
wavenumber/
-
1
cm
surface species
Al3+-OH
vibration
ref
3
3
3
3
768
755
686
608
ν(OsH)
ν(OsH)
ν(OsH)
ν(OsH)
28, 47
28, 47
28, 47
47
Ga-OH-Al
hydrogen-bonded
hydroxyl group
2230
2180
1716
1590
1580
1564
1552
1456
1392
1376
1306
1294
1240
out of phase (Al-NCO) ν(NdCdO) 42, 43
cyanide (-CN)
ν(CtN)
27, 37, 42
44
carbonyl
formate
bidentate nitrate (NO
acetate
unidentate nitrate
acetate
formate
formate
nitrate
ν
as(COO)
-) ν(NdO)
as(COO)
ν(NdO)
(COO)
27
27,36
27
3
ν
ν
s
27, 36
27, 36
27, 36
37
δ(CH)
ν
s
(COO)
unidentate nitrate
bidentate nitrate (NO
ν
ν
as(ONO)
as(ONO)
36
36
-)
3
Figure 1. IR spectra of adsorbed species under different gas
atmospheres over Ga /Al at 623 K after 200 min in (a) C
, (b) C + NO, (c) C + O + NO, and (d) NO + O
in a flow of NO +O2 (Figure 1, spectrum d) has been assigned
2
O
3
2
O
3
3 8
H +
4
5
by Chi et al. to NO2 stretching in bidendate nitrate.
O
2
H
3 8
3
H
8
2
2
.
In flowing C3H8 + O2 + NO, four strong bands at 1580,
-
1
1
552, 1294, and 1240 cm have been observed in addition to
lower than 673 K, the NOCF is ca. 40%, while at temperatures
higher than 673 K, the NOCF decreased. This suggests the
occurrence of side reactions at T ) 673 K and higher (propane
combustion). Thereby, propane is not completely oxidized, i.e.,
on the average 45% CO2 and 55% CO are detected as oxidation
products. On the other hand, NO is reduced nearly completely
to N2 (only 1% N2O were detected).
the bands observed in flowing C3H8 + NO and C3H8 + O2.
Couples of bands in the regions of 1650-1500 and 1170-1300
-
1
cm are assigned to the ν3 split stretching vibration of nitrates
2
8,46
on metal oxides in the literature.
Thus, the bands at 1552
-
1
-1
and 1294 cm and 1580 and 1240 cm have been assigned
to ν(NdO) and νas(ONO) of unidentate and bidentate nitrates,
respectively.
Two negative bands at 3768 and 3686 cm- (O-H stretching
region of the spectra) are observed in flowing C3H8 + O2 and
C3H8 + NO as well as C3H8 + O2 + NO. In flowing NO +
O2, these negative bands have been found to be slightly shifted
2
. Nature of Adsorbed Species Formed in Different Gas
1
Atmospheres. The IR spectra of the adsorbed species on Ga2O3/
Al2O3 in dependence on the composition of the gas atmosphere
at 623 K are shown in Figure 1. In flowing C3H8 + O2 (spectrum
a), the following bands have been observed: two strong bands
-1
to 3755 and 3686 cm . Signals of OH groups of the oxide
-
1
centered at 1564 and 1456 cm , two weak bands at 1392 and
surface should not be observed in the IR spectra because of
background subtraction. These OH groups are consumed by
interaction with acetate and/or nitrates or by formation of acetate
and/or nitrates. Thus they appear as negative bands in the
-
1
-1
1
376 cm , and a broad band at 3550 cm . In the literature,
-
1
the bands in the region between 3800 and 3200 cm are
assigned to the stretching vibration of OH groups. We assign
the broad band at 3550 cm to bridging hydroxyl groups
perturbed by interaction with C3H8 or with other organic species
generated from C3H8. The bands at 1564 and 1456 cm are in
good agreement with those of acetic acid adsorbed on Ga2O3/
Al2O3 as reported by Haneda et al.27 and by blank adsorption
tests performed by us for all relevant species discussed here
-
1
-1
spectra. The OH band around 3780-3760 cm has been
identified as being coordinated to a terminal Al-OH group,
-
1
46,47
which is generally ascribed to a basic hydroxyl group.
46
-1
Pushkar et al. assigned the band between 3688 and 3680 cm
to a Ga-OH-Al bridging structure on Ga2O3/Al2O3 catalysts.
20
Shimizu et al. suggested that Ga-O-Al bonds observed for
(not shown). These two bands are assigned to vibrations
low gallium loadings on Ga2O3/Al2O3 catalysts exhibit high
activity for SCR of NO by C3H6. According to this, it can be
expected that the OH groups of terminal basic Al-OH and of
bridged Ga-OH-Al are reactive and removed or exchanged
upon the formation of acetate and nitrates as has been observed
in this work. The main observed IR bands of the adsorbed
species on Ga O /Al O in various gas atmospheres and their
νas(COO) and νs(COO) of acetate formed by the reaction
between C3H8 and O2. The frequencies of the bands at 1392
-1
and 1376 cm are in good agreement with those of formic acid
adsorbed on the same catalysts (not shown) and those found in
previous studies with formate adsorbed on Al2O3 as reported
2
8
40
by Shimizu et al. and Amenomiya et al. Thus, the two bands
2
3
2
3
-
1
at 1392 and 1376 cm have been assigned to νs(COO) and
δ(CH) of adsorbed formate. They are also observed in flowing
C3H8 + NO and C3H8 + NO + O2. In comparison to the
C3H8 + O2 flow, four additional bands have been observed for
flowing C3H8 + NO: 2230, 2180, 1716, and 1306 cm . The
bands at 2230 and 2180 cm are assigned to adsorbed
isocyanate and cyanide, respectively.
at 1716 and 1306 cm are suggested to be due to vibrations of
adsorbed carbonyl and the asymmetric vibration of nitrate
according to recent literature by Meunier et al. and Shimizu
et al. An additional band in the region of 1400-1360 cm
assignments are listed in Table 3.
. Dynamic Changes of Surface Species under Reaction
Conditions (C H + NO + O ). The results of catalytic activity
3
3
8
2
measurements suggested that NO reduction to N over Ga O /
2
2
3
-
1
2 3
Al O was effectively catalyzed at temperatures above 623 K
-
1
(Table 2). To clarify the behavior of surface species formed
during the SCR reaction, the time-dependent quantitative
changes of surface species formed in flowing NO + C H +
4
1-43
The bands centered
-
1
3
8
O were measured at 623 K. As shown in Figure 2, the formation
2
37
-1
-1
of formate (2908, 1392, and 1378 cm ), acetate (1465 cm ),
and nitrite or nitrate (1580, 1552, and 1238 cm ) species is
4
4
-1
-1

Selective Catalytic Reduction of NO by C3H8
J. Phys. Chem. B, Vol. 109, No. 33, 2005 15909
Figure 2. FTIR spectra of adsorbed species in flowing NO + C
3
H
8
2 2 3 2 3
+ O on the Ga O /Al O catalyst surface at 623 K as a function of time: (a)
0
.5, (b) 2.5, (c) 5, (d) 15, (e) 30, (f) 60, (g) 90, (h) 120, and (i) 180 min.
in a flow of C3H8 + O2, C3H8 + NO, and C3H8 + NO + O2,
respectively, at 623 K. This figure illustrates that the corre-
sponding reaction rates and the concentration of adsorbed
surface acetate follows the sequence of C3H8 + O2 > C3H8 +
NO > C3H8 + NO + O2. C3H8 itself can hardly be adsorbed
4
8
on the catalyst surface at 623 K. Propane should be activated
by adsorbed O2 or adsorbed species generated from NO (e.g.,
-
48
NOx ). According to proposals from the literature, the decrease
in the concentration of adsorbed acetate in an order of C3H8 +
O2 > C3H8 + NO > C3H8 + NO + O2 is explained by the
difference in the oxidizing potential of the oxidizing agents and
4
9,50
the inhibition of propane oxidation by NO.
The difference
in reactivity of surface acetate to NO or O2 also explains the
relations of the surface concentrations. The initial rate of acetate
formation in flowing C3H8 + NO + O2 is higher than in flowing
C3H8 + NO (Figure 3). After 40 min, the acetate concentration
in flowing C3H8 + NO + O2 becomes lower than the one in a
flow of C3H8 + NO. In a flow of C3H8 + NO + O2, propane
is mainly oxidized by O2, and the acetate formed is further
oxidized by NO (or other nitrites/nitrates generated from
NO + O2). The intensity of the IR bands characteristic of surface
acetate increased very quickly at the beginning, and an equi-
librium between formation and consumption of surface acetate
is achieved after ca. 80 min.
Figure 3. Time dependence of integrated area of the band characteristic
-1
of acetate (1480-1420 cm ) under different gas atmospheres on Ga
Al at 623 K: (9) C + O , (b) C + NO, and (2) C
+ NO.
2 3
O /
O
2 3
3
H
8
2
H
3 8
3 8
H +
O
2
4. Studies on Single Steps of the Reaction Cycle: Forma-
tion and Consumption of Surface Acetate. The IR spectra of
adsorbed carbonaceous species over Ga2O3/Al2O3 in flowing
C3H8 + O2 at different temperatures are shown in Figure 4.
There are only quantitative differences in the IR spectra of
adsorbed species on Ga2O3/Al2O3 and Al2O3 (not shown). The
main adsorbed species in flowing C3H8 + O2 over both catalysts
observed immediately after the gas mixture introduction, while
weak bands due to nitrogen-containing organic species (-CN
and -NCO) are only observed after an induction period of ca.
5
min. Obviously they are formed by the reaction between
adsorbed carboxylates and NO or by species generated from
NO. The IR intensity of the formate bands increases at the
beginning and became constant after ca. 5 min; however, the
IR intensity of acetate increases upon time on stream until 80
min. This could result from the differences in the reactivity
between acetate and formate toward the other gases O2 or NO.
Acetate is also formed in C3H8 + O2 and C3H8 + NO as
shown in Figure 1. Figure 3 includes the change in intensity of
-
1
-1
are acetate (1558, 1468 cm ) and formate (1392, 1376 cm ).
The bands due to acetate and formate are observed over Ga2O3/
Al2O3 starting at 523 K. Over Al2O3, these bands can be
observed only at 573 K and higher temperatures. Furthermore,
the rate of formation of surface acetate over Ga O /Al O is
2
3
2
3
higher than the one over Al2O3 catalysts at the same temperature
(Figure 5). These results indicate that the impregnation of Ga2O3
onto Al2O3 reduces the reaction temperature and improves the
-
1
the IR band assigned to adsorbed acetate (1480-1420 cm )

15910 J. Phys. Chem. B, Vol. 109, No. 33, 2005
He et al.
Figure 4. IR spectra of adsorbed species on Ga
+ O at (a) 523 K, (b) 573 K, (c) 623 K, and (d) 648 K for
20 min.
2
O
3
/Al
2
O
3
in flowing
Figure 6. IR spectra of adsorbed species on Ga O /Al O at 623 K
2
3
2
3
C
1
H
3 8
2
taken (a) after soaking into a flow of C H + O for 120 min, followed
3
8
2
by He purge for 30 min and then changing the gas to (b) He for
8
9
0 min, (c) O
0 min, respectively.
2
2
for 90 min, (d) NO for 90 min, and (e) NO + O for
observed in flowing C3H8 + O2 (spectrum a). After purging in
flowing He for (80 min, spectrum b) and then in O2 for 90 min
(spectrum c), no significant changes have been observed in the
IR spectra. However, the intensity of the IR bands characteristic
of adsorbed acetate and formate decreased significantly when
the feed gas was switched to NO or NO + O2 during 90 min.
-
1
Between 2000 and 1000 cm , the bands due to acetate (1564,
-
1
1
458 cm ) disappeared completely, and bands at 1552, 1294
-1
-1
cm and 1580, 1240 cm assigned to ν(NdO) and νas (ONO)
of unidentate and bidentate nitrates, respectively, have been
observed. In the OH vibration region, negative bands were
shifted from 3768 and 3686 cm to 3755 and 3686 cm . The
surface acetate was consumed by NO or NO + O2.
The time dependence of the intensity of the acetate band
-
1
-1
-1
(
1480-1420 cm ; normalized by initial intensity) in the various
gas flows at different temperatures is shown in Figure 7. In a
flow of He as well as O2, a small decrease in the intensity of
the acetate band has been observed. This small decrease can be
interpreted by desorption of the adsorbed species (in both gases)
and by slow oxidation of the adsorbed species by oxygen (in
O2). However, the intensity of IR bands due to acetate
significantly decreased when the feed gas was switched to NO
or NO + O2. This can be due to (i) the reaction of acetate with
NO and NO/O2 or (ii) site competition with other species (e.g.,
Figure 5. Time dependence of the integrated area of the band due to
-
1
acetate (1480-1420 cm ) in flowing C
3
H
8
+ O
(blank) at different temperatures: (9, 0) 523 K,
b, O) 573 K, (2, 4) 623 K, and (1, 3) 648 K;
2 2 3 2 3
on Ga O /Al O
(
(
2 3
solid) and Al O
activity for the partial oxidation of propane. This correlates well
with the increased catalytic activity in SCR of the Ga2O3/Al2O3
system at lower temperatures (Table 2).
-
NOx ). All experimental observations of this paper and results
28,44
from the literature (e.g., of Shimizu et al.
) support that the
chemical reaction of surface oxygenates with NO (point i) is
the reason for this decrease. The temperatures at which the
reaction (acetate consumption) takes place correlates exactly
with the catalysis results and the formation of N2. The reaction
rates as a function of temperature reflect the expectations for
the reaction and not for adsorption-desorption equilibria. The
small differences in the reaction rate of acetate consumption in
NO and NO + O2, respectively, suggest that O2 does not
significantly promote the reaction between adsorbed acetate and
NO. The consumption of the surface oxygenates in O2 increased
significantly with increasing temperature. At lower temperatures,
The reactivity of adsorbed acetate over Ga2O3/Al2O3 toward
different oxidizing agents was evaluated by time-dependent
changes of the IR spectra at different temperatures. In these
experiments, the catalyst was first exposed to a flow of
C3H8 + O2 for 120 min (Figure 6, spectrum a), followed by
purging in flowing He for 30 min. Then the gas flow was
switched to He, O2, NO, and NO + O2, respectively. Figure 6
also shows the IR spectra recorded at 623 K for 90 min after
change of the gas composition. Adsorbed acetate (1564, 1456
-1
-1
cm ) and formate (1392, 1376 cm ) are the dominating species

Selective Catalytic Reduction of NO by C3H8
J. Phys. Chem. B, Vol. 109, No. 33, 2005 15911
-
1
Figure 7. Time dependence of the relative intensities of the acetate band in the range 1420-1480 cm normalized by initial intensity in (9) He,
b) O , (2) NO, and (1) NO + O . Before the measurements, the catalyst was pretreated in C + O /He for 120 min and purged by He at
23 K for 30 min: (A) 573 K; (B) 623 K; (C) 648 K.
(
2
2
H
3 8
2
6
Figure 9. Time dependence of the nitrate concentration (left) and
Figure 8. IR spectra of adsorbed species on Ga
73 K taken (a) after soaking into a flow of NO +O
followed by He purge for 30 min and then change the gas to (b) He
2
O
3
/Al
2
O
3
at T )
-1
relative intensities of nitrate bands in the region 1330-1180 cm
normalized by the intensity of 150 min in different gas flows: (0),
NO + O in helium; (b) He; (2) C /He; (9) C + O /He.
Temperature ) 573 K.
5
2
for 120 min
2
H
3 8
H
3 8
2
for 80 min, (c) C
respectively.
H
3 8
for 90 min, and (d) C
3
H
8
2
+ O for 90 min,
NO is the predominant reaction partner of surface acetate (even
in extreme excess of oxygen). For higher temperatures, the
reaction with oxygen is enhanced. This leads to a decrease in
efficiency of the reducing agent propane (competitiveness factor)
and is reflected by a (slight) volcano shape (shown in Table 2).
It should be mentioned that the temperatures consciously chosen
for the IR experiments correspond well to the temperatures
where Ga2O3/Al2O3 is active and shows measurable high NO
conversions to N2. They are below the optimum temperature
with highest activity (total conversion, T ) 723 K).
to C3H8 and C3H8 + O2, respectively, for 90 min. Figure 8 also
shows the IR spectra recorded at 573 K after changing the feed
-
1
gas for 90 min. Four bands centered on 1578 and 1240 cm
-
1
as well as 1554 and 1294 cm assigned to surface bidentate
and monodentate nitrate species were observed in flowing
NO + O2. After purging in flowing He for 80 min (spectrum
b) and/or in flowing C3H8 for 90 min, only small quantitative
changes of the IR bands are observed. That means adsorbed
nitrates do not react with C3H8 over Ga2O3/Al2O3 at 573 K.
However, after purging in flowing He and then in flowing
5
. Studies on Single Steps of the Reaction Cycle: Reactiv-
C3H8 + O2 for 90 min, a new band characteristic of acetate
-
(1458 cm-1) is observed. Simultaneously, the intensity of the
ity of Surface Nitrates. The reactivity of adsorbed NOx species
toward propane was evaluated by transient response of the IR
spectra at 573 K. First, the catalyst was exposed to a flow of
NO + O2 for 120 min (Figure 8, spectrum a) and then purged
in flowing He for 30 min. Finally, the gas flow was switched
IR bands characteristic of surface nitrates decreased.
Figure 9 shows the changes in the integrated intensity of the
-
1
nitrite/nitrate bands in the range of 1330-1180 cm (normal-
ized by initial intensity) as a function of time after the feed gas

15912 J. Phys. Chem. B, Vol. 109, No. 33, 2005
He et al.
was switched from NO + O2 to He, C3H8, or C3H8 + O2 at
73 K. The intensity of the IR bands characteristic of sur-
C H + xNO + (10 - x)/2O f 3CO + x/2N + 4H O
3
8
2
2
2
2
5
(1)
face nitrates decreased only slightly in flowing He, C3H8, or
C3H8 + O2. This indicates that nitrates are stable in C3H8 and
C3H8 + O2. In flowing He, the small decrease in the concentra-
tion of nitrates can be interpreted by desorption of adsorbed
species. In flowing C3H8, there is no significant difference in
the consumption rate of surface nitrate compared to flowing
C H + 10NO f 5N + 3CO + 4H O
(2)
(3)
3
8
2
2
2
C H + 5O f 3CO + 4H O
3
8
2
2
2
For the reduction of one molecule of NO, two electrons must
be delivered by propane (formal reduction from +2 for N in
NO to 0 in N2). On the other hand, one molecule C3H8 provides
0 electrons when it is oxidized to CO2 and H2O. One molecule
of propane can thus formally reduce 10 NO molecules to N2
(eq 2).
-
He. This illustrates that the adsorbed NOx species are not
reactive to C3H8 at 573 K. Interestingly, an increase of the
2
-
1
intensity of the IR bands due to nitrates (1330-1180 cm ) at
the beginning is observed when the feed gas was switched to
C3H8 + O2. This might be due to the formation of acetate species
by C3H8-O2 reaction, because weak IR bands due to acetate
As experimentally observed (Table 2), one molecule of
propane converts four NO exclusively to N , i.e., the overall
2
-
1
species appear in the regions of 1200-1235 cm and 1290-
reaction at temperatures of T e 673 K can be formally described
-
1 27,28
1
340 cm .
After 10 min, the intensity of the IR bands due
by eq 4
-
-1
to adsorbed NOx species (1330-1180 cm ) decreased with
time. Although the consumption rate of surface nitrates is also
slow, it is faster than in flowing He and in flowing C3H8. This
indicates that the surface acetate formed under these conditions
C H + 4NO + 3O f 3CO + 2N + 4H O
(4)
3
8
2
2
2
2
That means 12 electrons must be accepted by O2 and that O2
has been involved in the hydrocarbon SCR reaction. (More
exactly, the incomplete conversion of C3H8 to carbon dioxide
must be taken into account: 45% CO2, 55% CO. In concern of
the carbon balance, approximately 50% of the carbon com-
pounds are not completely oxidized, i.e., in fact only 11 electrons
are accepted by oxygen, or 4 NO molecules are reduced by
even less than 1 C3H8.)
(
C3H8 + O2) acts as the active intermediate in the consumption
-
of adsorbed NOx species.
Discussion
1. Particular Catalytic Properties of the Optimized System
Ga2O3/Al2O3. The optimized catalyst Ga2O3/Al2O3 investigated
in the present work shows high activity and selectivity to N2
for the selective catalytic reduction of nitric oxide by propane.
It converts NO completely at temperatures of 723K and above
This is still in agreement with reaction mechanisms proposed
in the literature for different catalyst/reducing agent systems,
according to which (i) NO initially activated (oxidized) by O2
2
7,28
reacts with the hydrocarbon,
(ii) the hydrocarbon is first
-
1
(
GHSV ) 10 000 h , Table 2) selectively to nitrogen (99%
3
5-38
partially oxidized and then reacted with NO,
or (iii) both
N2, 1% N2O).
NO and hydrocarbon are preoxidized by O2 and then the
The particular performance of this catalyst is, however, the
high and reproducible efficiency of the reducing agent (propane),
which is in general not regarded and achieved for other oxidic
34
generated species react to each other. If we assume however
that propane is preoxidized to surface acetate (what is generally
35-38
accepted by the majority of literature reports
and also
39
catalyst systems before. Yan et al. described this efficiency
by a so-called “NO competitiveness factor” (Table 2, superscript
c), which is near 40% (43.0% for 673 K and 41.5% for 723 K)
in the present work.
demonstrated in this work) then the reaction partner cannot be
-
an adsorbed nitrate (*-NO3 ). This reaction path (a variation
of iii) can be excluded by the electron balance considerations
for Ga O /Al O . In this case, only two NO molcules could be
2
3
2
3
As shown in Table 2, one molecule of propane converts rather
exactly four NO molecules exclusively to N2 (feed gas composi-
tion: NO:C3H8 ) 1000:1000 ppm). The same relation and
selectivity are observed for lower temperatures and lower
conversions (T ) 573, 623, and 673 K), i.e., at the temperatures,
at which the in situ IR studies were performed (Table 2). This
relation is clearly more efficient than for most other SCR
catalysts reported in the literature like silver-alumina or cobalt-
reduced by one molecule of propane in contrast to the
experimental finding of NO:C3H8 conversion ) 4:1.
Even the preoxidation of NO to NO2 becomes less probable
due to the following calculations: By assumption that propane
is oxidized first to surface acetate, the following reaction fits
the mechanism best
“
*-CH COOH” + 4NO f 2N + 2CO + 2H O (5)
3 2 2 2
1
1,39
alumina systems.
It is this that makes the catalyst Ga2O3/
This equation reflects the highest possible efficiency for NO
conversion by surface acetate and fits very well the proposed
reaction mechanism presented below for the Ga2O3/Al2O3
system, i.e., NO oxidation is neither expected on Ga2O3/Al2O3
or on Al2O3 nor it is in agreement with the electron balance
and reaction mechanism. Although these considerations do not
give any information on the detailed reaction mechanism as,
e.g., the role of nitrogen containing organic species or possible
NO decomposition, it allows valuable conclusions to exclude
some reaction pathways (the pathway iii where both reactants
are first oxidized) for the Ga2O3/Al2O3 system.
3. Formation and Consumption of Surface Species. Ad-
sorbed nitrates, oxygenates (partially oxidized hydrocarbons,
mainly acetate and formate), and nitrogen-containing organic
species (-CN and -NCO) have been observed on the surface
of Ga2O3/Al2O3 exposed to SCR atmosphere at reaction tem-
Al2O3 particularly interesting for mechanistic investigations also
because of several additional reasons: (i) several reaction
mechanisms proposed in the literature can be surely excluded
for Ga2O3/Al2O3 due to the overall electron balance (see below);
(
ii) it is generally assumed that such oxide systems neither
26
directly decompose NO nor oxidize NO to NO2; (iii) further-
more, neither Ga2O3 nor Al2O3 are known to be redox active
materials, i.e., partially reduced catalyst intermediates can also
be excluded. In the following, the electron balance of the
reaction system will be regarded in detail.
2
. Electron Balance and Stoichiometric Calculations for
Possible Reaction Pathways. There are two oxidizing agents
(NO and O2) and one reducing agent (C3H8) in the SCR feed
(
3
reaction eq 1), and two main competing reactions (eqs 2 and
) must be taken into account for the overall SCR process.
5
1

Selective Catalytic Reduction of NO by C3H8
J. Phys. Chem. B, Vol. 109, No. 33, 2005 15913
peratures (Figures 1 and 2). The same surface species have also
been reported by Sibata et al.36 during SCR of NO by propane
over Ag/Al2O3 catalysts and by Haneda et al.2 during SCR of
NO by propene over Ga2O3/Al2O3. In the present work, the
formation and consumption of these adsorbed species on Ga2O3/
Al2O3 in different gas atmospheres were investigated in order
to elucidate their role in the SCR reaction. As shown in Figures
SCHEME 1: Proposed Reaction Pathways of NO
Reduction by Propane over Ga2O3/Al2O3
a
7
1
and 3, hydrocarbon-derived surface oxygenates (acetate,
formate) can be formed in flowing C3H8 + O2, C3H8 + NO, as
well as C3H8 + O2 + NO at 623 K. The concentration of
adsorbed surface acetate decreases in the order C3H8 + O2 >
C3H8 + NO > C3H8 + NO + O2, suggesting that only O2 but
not NO is a useful oxidizing agent for the partial oxidation of
propane over Ga2O3/Al2O3. The partial oxidation of propane is
a
Pathways 1 and 4 are the main reaction pathways.
can react with hydrocarbon derivates (acetate, formate) generated
from C3H8 + O2 only. However, according to the discussion
above the surface nitrite/nitrate species cannot be involved in
the intrinsic SCR mechanism and we attribute them to spectators
for the catalytic system under study. As discussed above, the
formation of surface nitrites/nitrates cannot be the rate-determin-
ing step of the SCR reaction over Ga2O3/Al2O3. Even more,
the role of oxidized NOx surface species in the reaction
mechanism is called into question at all by experimental results
and electron balance considerations presented in this work.
Nitrogen-containing organic species (-CN and -NCO) are
only detected in a flow of C3H8 + NO + O2 after an induction
period of about 5 min (Figure 2). Probably they are formed by
the reaction between adsorbed carboxylates and NO or NO
derived species only.
even inhibited by NO. These results are in good logical
agreement with De Soete et al.49 by studies on combustion of
coal char in gas atmospheres with and/or without NO and with
Chambers et al.,52 who reported that the introduction of NO
into a flow of C3H8 + O2 caused an inhibition of propane
oxidation over Pt/SiO2 catalysts. This is explained by NO2
formation and decomposition cycles on the catalyst surface.
In flowing C3H8 + O2, the formation of surface acetate starts
from 523 K and its concentration increases with increasing
temperature until 673 K (Figure 3). The impregnation of Ga2O3
onto Al2O3 leads to an enhancement of the partial oxidation of
27
propane. Explanations are given by Haneda et al. according
to their catalysis testing and catalyst characterization results. It
is this partial oxidation of propane to acetate (formate) that
represents the rate-determining step of the SCR reaction. There
is a perfect correlation between this partial oxidation reaction
and the activity of the Ga2O3/Al2O3 catalyst in the SCR reaction.
The formation of surface nitrite/nitrate species cannot be the
rate-determining step, because the rate of nitrite/nitrate formation
and its surface concentration decreases with increasing temper-
ature, whereas the NO conversion increases with increasing
temperature when the reaction temperature is lower than
4
. Reaction Mechanism of SCR of NO by C3H8 over
Ga2O3/Al2O3. As discussed above and illustrated in Scheme 1,
the C3H8-SCR of NO over Ga2O3/Al2O3 starts with the partial
oxidation of C3H8 by O2 to surface carboxylates such as acetate
and formate (step 1). This represents the rate-determining step
of the whole reaction cycle. The surface oxygenates (mainly
acetate, formate) are highly reactive to (adsorbed) NO (or NO-
derived species) forming nitrogen-containing organic surface
species (isocyanides, cyanides, step 4). The coupling of nitrogen
atoms to form N2 most probably occurs via the comproportion-
ation reaction between NO (gas or adsorbed) and reduced
nitrogen containing species (e.g., -CN, -NCO, step 5). This
mechanism is in very good agreement with the catalytic results
7
23 K.
Differences in reactivity of surface acetate with O2, NO, or
NO + O2 are also reflected by the time dependence of the
intensity of acetate IR bands in corresponding atmospheres. As
shown in Figures 6 and 7, surface acetate species are stable in
flowing O2 but reactive in flowing NO or NO + O2. The
presence of NO (and/or N-derived species) decreases the surface
acetate concentration. The small differences in the rate of
consumption of surface acetate in flowing NO and NO + O2
(
N2 selectivity, C3H8 conversion/competitiveness factor) as a
function of temperature and the presence, concentration, and
rate of formation and consumption of all relevant surface species
observed by IR spectroscopy.
The possible oxidation of propane by NO or NOx (oxidized
NOx species, step 2) is much slower than the main reaction (step
in Scheme 1; Figure 3). Nitrite and/or nitrate species, which
are observed on the Ga2O3/Al2O3 surface, are found to be
spectators only. Their participation in the main reaction cycle
(Figure 7) indicate that O2 does not enhance the reaction between
1
acetate and NO. This is an additional indication that the
participation of nitrite/nitrate species in the SCR reaction
mechanism over Ga2O3/Al2O3 can be neglected. The rate of
acetate consumption depends on the reaction temperature. The
increase in reaction rate with temperature and the absolute
temperature values correlate very well with the catalytic results
(
steps 1, 4, and 5) can be excluded from considerations of the
electron balance and the reaction rates. At higher temperatures
surface carboxylates can be oxidized totally by oxygen (step 3
in Scheme 1; Figure 7). This side reaction is the reason for the
decrease of the NO competiveness factor in the reaction at
temperatures above 773 K and probably for the clearly less
efficient NO reduction over other HC-SCR catalysts such as
(NO conversion, Table 2).
Surface nitrite and/or nitrate species are observed in flowing
NO, NO + O2, as well as NO + O2 + C3H8 (Figures 2 and 8).
The intensity of the IR bands due to surface nitrites/nitrates
decreases only slightly in flowing He and C3H8 at 573 K.
Haneda et al.27 and Shimizu et al. suggested for C3H6 that
this alkene can be oxidized by O2 as well as surface nitrates.
C3H8, however, does not react with surface nitrates on Ga2O3/
Al2O3 as found in this study (as shown in Figure 8). This can
be explained by the difference in the reactivity of C3H8 and
C3H6. The decrease of intensity of nitrite/nitrate bands in flowing
C3H8 + O2 is faster than in flowing C3H8 suggesting that nitrates
11,39
Ag/Al2O3 and Co/Al2O3 in comparison to Ga2O3/Al2O3.
28
Conclusions
The formation and consumption of surface species and the
role of the adsorbed oxygenates and nitrates during SCR of NO
by propane on Ga2O3/Al2O3 were investigated by in situ IR
spectroscopy. The reaction starts with the formation of surface
carboxylates (acetate, formate) by the reaction of C3H8 + O2.

15914 J. Phys. Chem. B, Vol. 109, No. 33, 2005
He et al.
Carboxylates become the predominant surface species in the
steady state and act as an important intermediate during the NO
reduction. This propane oxidation is the rate-determining step
of the whole SCR reaction. The surface carboxylates react easily
with (adsorbed) NO. Nitrogen-containing organic species are
believed to be intermediates in the final formation of N2. The
stoichiometry of the SCR reaction, in particular the comparison
of NO and hydrocarbon conversion, allow interesting conclu-
sions concerning the reaction mechanism. Surface nitrites/
nitrates are found to be only spectators in the system under study
(19) Bennici, S.; Gervasini, A.; Ravasio, N.; Zaccheria, F. J. Phys. Chem.
B 2003, 107, 5168.
(
20) Shimizu, K.; Satsuma, A.; Hattori, T. Appl. Catal., B 1998, 16,
19.
(21) Haneda, M.; Kintaichi, Y.; Hamada, H. Appl. Catal., B 2001, 31,
51.
3
2
1
(22) Haneda, M.; Kintaichi, Y.; Shimada, H.; Hamada, H. J. Catal. 2000,
92, 137.
(23) Haneda, M.; Kintaichi, Y.; Hamada, H. Catal. Today 1998, 54,
391.
(24) Haneda, M.; Kintaichi, Y.; Hamada, H. Appl. Catal., B 1999, 20,
2
89.
(
25) Chen, L.; Horiuchi, T.; Mori, T. React. Kinet. Catal. Lett. 2000,
69, 265.
(propane, Ga2O3/Al2O3).
(26) Burch, R.; Breen, J. P.; Meunier, F. C. Appl. Catal., B 2002, 39,
2
Acknowledgment. The authors thank the Bayerische For-
schungsstiftung for financial support; C.H.H. acknowledges the
Bayerische Forschungsstiftung for a grant. Prof. Janos Mink
Hungarian Academy of Sciences, Budapest, Hungary) is
acknowledged for his useful discussion about the assignment
of the IR bands.
D.; Hamada, H. J. Catal. 2002, 206, 114.
(28) Shimizu, K.; Kawabata, H.; Satsuma, A.; Hattori, T. J. Phys. Chem.
B 1999, 103, 5240.
(
65.
(
32) Wogerbauer, C.; Maciejewski, M.; Baiker, A. Appl. Catal., B 2001,
References and Notes
3
4, 11.
(33) Cho, B. K. J. Catal. 1998, 178, 395.
(34) Engler, B. H.; Leyrer, J.; Lox, E. S.; Ostgathe, K. Stud. Surf. Sci.
(
1) Amiridis, M. D.; Zhang, T.; Farranto, R. J. Appl. Catal., B 1996,
1
0, 203.
(
(
(
2) Fokema, M. D.; Ying, J. Y. Catal. ReV. 2001, 43, 1.
Catal. 1995, 96, 529.
3) Armor, J. N. Appl. Catal., B 1992, 1, 221.
(35) Obuchi, A.; Ogata, A.; Mizuno, K.; Ohi, A.; Nakamura, M.; Ohuchi,
H. J. Chem. Soc., Chem. Commun. 1992, 247.
(36) Shibata, J.; Shimizu, K.; Satokawa, S.; Satsuma, A.; Hattori, T.
Phys. Chem. Chem. Phys. 2003, 5, 2154.
(37) Meunier, F. C.; Breen, J. P.; Zuzaniuk, V.; Olsson, M.; Ross, J. R.
H. J. Catal. 1999, 187, 493.
(38) Radtke, F.; Koeppel, R. A.; Minard, E. G.; Baiker, A. J. Catal.
1997, 167, 127.
(39) Yan, J. Y.; Lei, G. D.; Sachtler, W. M. H.; Kung, H. H. J. Catal.
1996, 161, 43.
(40) Amenomiya, Y. J. Catal. 1979, 57, 64.
(41) Kameoka, S.; Ukisu, Y.; Miyadera, T. Phys. Chem. Chem. Phys.
2000, 2, 367.
(42) G u¨ nzler, H.; Gremlich, H. IR Spectrascopy; WILEY-VCH: 1998;
p 243.
4) Iwamoto, M.; Yahiro, H.; Shundo, S.; Yoshihiro, Y.; Mizuno, N.
Appl. Catal. 1991, 69, L15.
(
5) Walker, A. P. Catal. Today 1995, 26, 107.
(6) Park, Y. K.; Lee, J. W.; Lee, C. W.; Park S. E. J. Mol. Catal., A
2
000, 158, 173.
7) Sadovskaya, E. M.; Suknev, A. P.; Pinaeva, L. G.; Goncharov, V.
B.; Balzhinimaev, B.; Chupin, C.; Mirodatos, C. J. Catal. 2001, 201, 159.
(
(
81.
8) Wang, X.; Chen H.-Y.; Sachlter, W. M. H. J. Catal. 2001, 197,
2
(
9) Wogerbauer, C.; Maciejewski, M.; Baiker, A.; Gobel, U. Top. Catal.
2
001, 16, 181.
(
10) Shen, S.-C.; and Kawi, S. J. Catal. 2003, 213, 241.
(11) Iliopoulou, E. F.; Evdou, A. P.; Lemonidou, A. A.; Vasalos, I. A.
Appl. Catal., A 2004, 274, 179.
12) Kung, M. C.; Kung, H. H. Top. Catal. 2000, 10, 21.
(43) Bion, N.; Saussey, J.; Haneda, M.; Daturi, M. J. Catal. 2003, 217,
47.
(44) Shimizu, K.; Kawabata, H.; Maeshima, H.; Satsuma, A.; Hattori,
T. J. Phys. Chem. B 2000, 104, 2885.
1
81, 1.
14) Konin, G. A.; Il’ichev, A. N.; Matyshak, V. A.; Khomenko, T. I.;
(
Korchak, V. N.; Sadykov, A.; Doronin, V. P.; Bunina, R. V.; Alikina, G.
M.; Kuznetsova, T. G.; Paukshtis, E. A.; Fenelonov, V. B.; Zaikovski, V.
I.; Ivanova, A. S.; Beloshapkin, S. A.; Rozovskii, A. Y.; Tretyakov, V. F.;
Ross, J. R. H.; Breen, J. P. Top. Catal. 2001, 16/17, 193.
(45) Chi, Y.; Chuang, S. S. C. J. Catal. 2000, 190, 75.
(46) Pushkar, Yu. N.; Sinitsky, A.; Parenago, O. O.; Kharlanov, A. N.;
Lunina, E. V. Appl. Surf. Sci. 2000, 167, 69.
(47) Morterra, C.; Magnacca, G. Catal. Today 1996, 27, 497.
(48) Burch, R.; Halpin, E.; Sullivan, J. A. Appl. Catal. B 1998, 17, 115.
(49) De Soete, G. G. Proc. 23rd Int. Symp. Combust. New Orleans,
LA, 1990, 1257.
(50) Hucknall, D. J.; Chemistry of Hydrocarbon Combustion; Chapman
and Hall: London, 1985.
(51) Chajar, Z.; Primet, M.; Praliaud, H. J. Catal. 1998, 180, 279.
(52) Chamber, D. C.; Cant, N. W. Appl. Catal. B 2003, 41, 61.
(
15) Amalric-Popescu, D.; Bozon-Verduraz, F. Catal. Lett. 2000, 64,
25.
16) Pasel, J.; Speer, V.; Albrecht, C.; Richter, F.; Papp, H. Appl. Catal.,
B 2000, 25, 105.
17) Pietrogiacomi, D.; Sannino, D.; Tuti, S.; Ciambelli, P.; Indovina,
V.; Occhiuzzi, M.; Pepe, F. Appl. Catal., B 1999, 21, 141.
18) Fokema, M. D.; Ying, J. Y.Appl. Catal., B 1998, 18, 71.
1
(
(
(