Mechanism of O2 desorption during N2O decomposition on an oxidized Rh/USY catalyst

Article abstract of DOI:10.1006/jcat.2001.3197

Pulsed N₂O decomposition on a reduced or oxidized Rh/USY catalyst has been carried out to study the dependence of oxygen coverage (θ_o) on activity and to study the mechanism of O₂ desorption using an isotopic tracer technique. Decomposition activity decreased to the minimum (formation of only N₂) at θ_o = 0.5, but increased with increasing coverage (θ_o > 0.5), and finally high activity with steady-state O₂ production was attained at high θ_o (> 1.2). In the isotope study, N₂ ¹⁶O was pulsed onto ¹⁸O/oxidized Rh catalyst at low temperatures (220-240°C), and desorbed O₂ molecules were monitored by means of mass spectrometry. The ¹⁸O fraction in the desorbed oxygen had almost the same value as that on the surface oxygen. The result shows that O₂ desorption does not proceed via the Eley-Rideal mechanism, but via the Langmuir-Hinshelwood mechanism, i.e., desorption of dioxygen through the recombination of adsorbed oxygen. On the other hand, O₂-TPD measurements in He showed that desorption of oxygen from the Rh catalyst occurred at much higher temperatures. Therefore, it was proposed that reaction-assisted desorption of O₂ occurs during N₂O decomposition at low temperatures.

Full text of DOI:10.1006/jcat.2001.3197

Journal of Catalysis 200, 203–208 (2001)
doi:10.1006/jcat.2001.3197, available online at http://www.idealibrary.com on
Mechanism of O2 Desorption during N2O Decomposition
on an Oxidized Rh/USY Catalyst
Shin-ichi Tanaka, Koichi Yuzaki, Shin-ichi Ito, Satoshi Kameoka, and Kimio Kunimori1
Institute of Materials Science, University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan
Received July 10, 2000; revised February 6, 2001; accepted February 6, 2001; published online April 24, 2001
Rh surface (9). Furthermore, an O2-TPD measurement (in
He) showed that O2 was not desorbed up to 600 C (9).
In practice, TPD experiments (in vacuo) show that oxygen
atomson a Rh(111) surface desorb to form O2 at 800 C (11).
Therefore, why O2 is desorbed at such low temperatures
during catalytic N2O decomposition over Rh/USY catalyst
remains an open question.
Pulsed N2O decomposition on a reduced or oxidized Rh/USY
catalyst has been carried out to study the dependence of oxygen
coverage ( o) on activity and to study the mechanism of O2 des-
orption using an isotopic tracer technique. Decomposition activity
decreased to the minimum (formation of only N2) at o = 0.5, but in-
creased with increasing coverage ( o > 0.5), and finally high activity
with steady-state O2 production was attained at high o (>1.2). In
the isotope study, N2 16O was pulsed onto 18O/oxidized Rh catalyst
at low temperatures (220–240 C), and desorbed O2 molecules were
monitored by means of mass spectrometry. The 18O fraction in the
desorbed oxygen had almost the same value as that on the surface
oxygen. The result shows that O2 desorption does not proceed via
the Eley–Rideal mechanism, but via the Langmuir–Hinshelwood
mechanism, i.e., desorption of dioxygen through the recombination
of adsorbed oxygen. On the other hand, O2-TPD measurements
in He showed that desorption of oxygen from the Rh catalyst oc-
curred at much higher temperatures. Therefore, it was proposed
that reaction-assisted desorption of O2 occurs during N2O decom-
c
The mechanisms of N2O decomposition have been given
as follows (1, 10):
N2O → N2 + O(a),
2O(a) → O2,
N2O + O(a) → N2 + O2.
[1]
[2]
[3]
Step [1] is dissociative N2O adsorption followed by the
production of N2 and adsorbed oxygen on the catalyst sur-
face. Step [2] is O2 desorption by the recombinative des-
orption of adsorbed oxygen, and the so-called Langmuir–
Hinshelwood (LH) mechanism is described by steps [1] and
[2]. As stated above, however, it may be difficult to expect
that step [2] prevails at low temperatures. Step [3] is oxygen
removal via the Eley–Rideal (ER) mechanism (4), which
might be possible at relatively low reaction temperatures
(4, 12). In the case of the Rh/USY catalyst described above
(9), it appeared that O2 is freed on collision of N2O with
the adsorbed oxygen (step [3], ER mechanism). Dandl and
Emig (12) proposed a mechanistic model from the kinetics
simulation, where the ER mechanism prevails at lower tem-
peratures and the LH mechanism prevails at higher temper-
atures. On the other hand, the hot-atom (HA) mechanism
(13) may also be considered, where only hot (nascent) O(a)
atoms produced by step [1] are desorbed via step [2].
We have already studied the mechanism of N2O decom-
position on an oxidized Rh black catalyst (without the sup-
port) using an 18O tracer technique (i.e., using N2 16O pulse
onto 18O(a)-covered Rh surface) (14). Surprisingly, the
O2 desorption was found to proceed via step [2] (i.e., LH
mechanism) at 220 C, and O2-TPD (in He) showed that de-
sorption of oxygen occurred at higher temperatures (14).
Because the recombinative desorption of oxygen did not
occur in He at 220 C, we have proposed reaction-assisted
position at low temperatures.
2001 Academic Press
Key Words: N2O decomposition; 18O isotope; Rh/USY cata-
lyst; mechanism of O2 desorption; oxygen coverage; temperature-
programmed desorption.
INTRODUCTION
Nitrous oxide (N2O), which has been considered a strong
greenhouse gas, also contributes to catalytic stratospheric
ozone destruction (1). Therefore, the catalytic decomposi-
1
tion of N2O (N2O → N2 + 2 O2) has been attracting much
attention (1–10). Various research groups have studied
supported noble or transition metal catalysts (2–5) and
oxides (6, 7) for N2O decomposition at different tempera-
tures (250–500 C). We have reported that a Rh/USY cata-
lyst was very active for the catalytic decomposition of N2O
1
even at low temperatures around 250 C (SV = 60,000 h
)
(8). A pulsed N2O experiment over the Rh/USY catalyst (at
260 C) showed that only N2 was produced on a reduced Rh
surface, but O2 production started on an oxygen-covered
1 To whom correspondence should be addressed. E-mail: kunimori@
ims.tsukuba.ac.jp. Fax: +81-298-55-7440.
203
0021-9517/01 $35.00
c
Copyright
2001 by Academic Press
All rights of reproduction in any form reserved.

204
TANAKA ET AL.
desorption of O2 during N2O decomposition at low tem- for the isotopic tracer study. The temperature was increased
perature (14).
from room temperature to 800 C at a constant heating rate
In this work, the 18O tracer technique was applied to the of 10 C/min and was kept at 800 C for 20 min.
Rh/USY catalyst to elucidate the mechanism of O2 desorp-
tion. It is of interest to make clear whether LH-type recom-
binative desorption of oxygen (i.e., step [2]) also prevails
RESULTS AND DISCUSSION
or not on the supported Rh catalyst. Before the isotopic
Dependence of Oxygen Coverage on N2O Decomposition
study, pulsed N2O experiments were carried out to study
N2O pulses were injected onto the Rh/USY catalyst after
the oxygen coverage dependence on activity (N2 and O2
H2 reduction at 500 C. Figure 1 shows amounts of products
formation), and O2-TPD studies were also performed to
(N2 and O2) from N2O decomposition at 260 C to each
elucidate the nature of the adsorbed oxygen.
pulse (0.517% N2O in He). The pulse interval was 3 min.
Until the 19th pulse, the product was only N2 . The amount
EXPERIMENTAL
was 0.27 mol (equivalent to 100% conversion of N2O) for
the first pulse, and it decreased until the 9th pulse. After the
10th pulse it increased slowly, and finally reached a stable
state. On the other hand, O2 was produced after the 19th
pulse, and the amount also increased slowly, finally reaching
a stable state. Because no O2 was desorbed until the 19th
pulse, the amount of adsorbed oxygen (i.e., oxygen cover-
age) was increased with pulsing N2O (step [1]). Here, the
coverage of oxygen ( o) was defined as the atomic ratio of
the number of adsorbed oxygens to the number of surface
rhodiums (Rhs), which was based on the H/Rh value (0.54).
The o at each stage can be easily calculated from the data
in Fig. 1. Figure 2 shows the activity of N2O decomposition
as a function of o on the Rh/USY catalyst at 220 and 260 C,
respectively. A drastic decrease in N2O conversion was ob-
served with increasing o (up to 0.5), which may be consis-
tent with the literature (17). In practice, according to the
study on the sticking coefficient of N2O on Rh(110) (17),
the dissociative chemisorption of N2O (step [1]) stops at
o = 0.5. Note that in the case of single-crystal work, expo-
sure of the sample to N2O gas was orders of magnitude
A Rh/USY catalyst was prepared by incipient wetness
impregnation of USY (TOSOH Co., Si/Al2 = 14.6) with an
aqueous solution of Rh(NO3)3 followed by calcination in
air at 600 C for 3 h (9). The Rh loading on USY was 2 wt% .
H2 and CO chemisorption measurements were performed
with a conventional volumetric adsorption apparatus (15,
16). Total adsorption of H2 (H/Rh), as well as irreversible
adsorption of CO (CO/Rh), was measured after H2 reduc-
tion for 1 h at 500 C. In this work, for the Rh/USY catalyst,
H/Rh and CO/Rh were 0.54 and 0.51, respectively.
The reaction of N2O decomposition was performed in a
microcatalytic pulse reactor (9, 14). A quartz tube reactor
(i.d. 4mm) wascharged with 19.5mgofthe Rh/USY catalyst
(5 mm in height, Rh = 3.79 mol), and it was treated
in H2 reduction for 1 h at 500 C before measurement.
Highly purified He (99.9999% ) was used as a carrier gas
at a flow rate of 55 cm3/min. A pulse of 0.517% N2O in He
(0.27 mol/pulse) wasinjected bya jacketed switchingvalve
purged with He. No decomposition was observed on the
USY support only, even at 500 C.
Isotope-labeled 18O2 (96.5% 18O2) was obtained from
Icon Company Ltd. The 18O tracer-loaded catalyst was pre-
pared as follows: the catalyst was treated with 18O2 (80 Torr)
in an in situ closed system at 300 C for 3 h after H2 re-
duction at 500 C. The reactant gas (0.517% N2 16O in He)
and probe gases (0.208% 18O2 in He and 0.103% C16O in
He) were flushed onto the catalyst via the carrier gas. The
amount of N2O was 0.27 mol/pulse, 18O2 0.10 mol/pulse,
and CO 0.04 mol/pulse. The effluent was analyzed in an
on-line gas chromatograph system equipped with a TCD
detector (Shimadzu, GC-8A) and differentially pumped
quadrupole mass spectrometer (Balzers, QMS 200 F). To
prevent leakage of 16O2 from the atmosphere into the gas
line, the whole apparatus, which was located in a corner of
the laboratory room, was isolated from the atmosphere by
drawing curtains to make a small room in which N2 gas was
purged.
Quantitative TPD measurement in a He flow was per-
formed using 19.5 mg of the Rh/USY catalyst (3.79 mol
FIG. 1. Amount of products from each pulse of N2O over the Rh/USY
as Rh metal). The analysis equipment used was the same as catalyst at 260 C: (4) N2, (᭹) O2.

CATALYTIC DECOMPOSITION OF N2O ON Rh/USY
205
of desorbed O2 (atomic O/Rhs) was (A) 0.23 and (B) 0.85,
respectively, which are significantly lower than the dosed
values ( o = 0.5 and 1.2, respectively). This result suggests
that a large amount of oxygen was not desorbed when the
temperature was kept at 800 C for 20 min. After the N2O
flow, pretreatment C, the O2-TPD peak was similar to that
after pretreatment B, although the peak area was larger
(O/Rhs = 0.98) after the N2O flow than after the N2O pulses.
The results in Fig. 3 suggest that oxygen species with weaker
O–Rh bonds, which are desorbed below 700 C in the O2-
TPD experiment, play an important role in the high activity
as well as steady-state O2 production in the catalytic N2O
decomposition.
Similar TPD experiments were also performed using O2
pulses to elucidate the differences in the states of adsorbed
oxygen. Figure 4 shows O2-TPD spectra from the Rh/USY
catalyst after H2 reduction at 500 C followed by the pre-
treatments using O2 pulses (A and B) at 220 C. As shown
FIG. 2. Conversion of N2O to N2 over the Rh/USY catalyst as a func-
tion of oxygen coverage: (4), 220 C; (᭜), 260 C. O2 formation started at
o Լ 0.9 (indicated by the vertical arrows).
in the inset to Fig. 4, was 0.5 after 5 O2 pulses, pretreat-
o
ment A, and 1.0 after 25 O2 pulses, pretreatment B. The
amount of desorbed O2 (O/Rhs) was (A) 0.55 and (B) 0.89,
respectively, which are roughly consistent with the dosed
values ( o = 0.5 and 1.0, respectively). Therefore, it is sug-
gested that N2O pretreatments leads to the formation of
stronger O–Rh bonds than O2 pretreatments, in particular
at low o ( =0.5). At high o ( 1.0), however, both N2O and
O2 pretreatments gave oxygen species with weaker O–Rh
bonds, which were desorbed below 700 C.
In Fig. 2, it is not clear at the present stage why N2O con-
version increases at higher o coverages. From the HREEL
study of oxygen structures on Rh(110), the Rh surface
was reconstructed with increasing o, and the structural
lower. In the case of supported Rh (see Fig. 2), an increase in
activity was observed with increasing o (>0.5), O2 produc-
tion started at Լ 0.9, and finally, steady-state activity was
o
attained at high
(>1.2). This phenomenon appears not
o
to be related to a diffusion process such as oxygen spillover
from Rh particles to the support, because the same results
as in Figs. 1 and 2 were obtained when the pulse interval was
changed to 9 min. The adsorbed oxygen species with high
may play an important role in the formation of active
o
sites for O2 production.
It should be noted here that the N2/O2 ratio was 2.0 at
steady state in the flow reactor system (8, 9). In Fig. 1, how-
ever, the N2/O2 ratio was 2.4 at the 50th pulse. This means
that oxygen was still adsorbed on the Rh surface even after
the 50th pulse ( = 1.5), and the adsorbed oxygen may be
o
located in the subsurface layer. On the other hand, after a
N2O flow treatment (0.517% N2O in He) at 220 or 260 C
for 1 h, the N2/O2 ratio was constant (2.0) from the 1st N2O
pulse.
O2-TPD Studies on N2O- or O2-Treated Rh Catalyst
To elucidate the difference in the states of the adsorbed
oxygen species, O2-TPD spectra were measured at different
o after N2O or O2 pretreatments. Figure 3 shows O2-TPD
spectra from the catalyst after H2 reduction at 500 C fol-
lowed by N2O pretreatments (A, B, or C) at 220 C. In the
case of pretreatments A and B using N2O pulses, oxygen
coverage can be calculated from the data in Fig. 1, as shown
in the inset to Fig. 3 ( o = 0.5 after 10 N2O pulses and
=
o
1.2 after 50 N2O pulses). It should be recalled that the min-
imum and steady-state activities were observed after the
pretreatments A and B, respectively (Fig. 2). As shown in
Fig. 3, O2 desorption starts at 700 C after pretreatment A,
FIG. 3. Temperature-programmed desorption (TPD) spectra of O2
from the Rh/USY catalyst after N2O pretreatments: Pretreatment A:
0.517% N2O in He (10 pulses at 220 C). Pretreatment B: 0.517% N2O
in He (50 pulses at 220 C). Pretreatment C: 0.517% N2O in He (flow,
whereas it starts at 550 C after pretreatment B. The amount 30 ml/min for 1 h at 220 C).

206
TANAKA ET AL.
Mechanism of O2 Desorption Using Isotopic
Tracer Technique
The reduced Rh/USY catalyst was oxidized with 18O2
(96.5% ) gas at 300 C, and pulsed N2O decomposition was
carried out at 220 C. An isotopic equilibrium constant, Ke,
should be considered to judge incidental exchange reac-
tions that would disguise the experimental results. Taking
into account an equilibrium reaction,
16
18O2 + O2
218O16O,
[4]
Ke is generally given as
[18O16O]2
[18O2][16O2]
Ke =
.
[5]
FIG. 4. Temperature-programmed desorption (TPD) spectra of O2
from the Rh/USY catalyst after O2 pretreatments: Pretreatment A:
0.208% O2 in He (5 pulses at 220 C). Pretreatment B: 0.208% O2 in He
(25 pulses at 220 C).
If the exchange reaction equilibrates, Ke should be close
to 4 (19). The same rule applies for other exchange reac-
tions. An isotopic fraction of 18O[18 f = 18O/(16O + 18O]
on the Rh surface can be evaluated by a pulsed C16O ex-
periment. It should be noted that the amount of CO pulse
(0.04 mol) is negligible compared with that of surface Rh
atoms (2.05 mol), which was estimated by H2 chemisorp-
tion (H/Rh). Table 1 shows the 18 f and Ke for the product
molecules obtained at 220 C. The CO molecules react with
the surface oxygen to form CO2, and CO conversion was
100% in the pulse experiment at 220 C. The exchange re-
action of oxygen in CO2 is fast on metals (19). As shown in
Table 1, Ke is4.00, which suggeststhat the isotopicexchange
of oxygen in CO2 equilibrates (Table 1, Expt. 1). Therefore,
the 18 f in the product CO2 should be equal to that of the sur-
face oxygen. Since 18 f in the product CO2 was 0.69 (Table 1,
Expt. 1), the 18 f on the Rh surface after 18O2 treatment
was determined to be 0.69. As a separate experiment, the
18O2 pulse was injected onto 16O-covered catalyst at 220 C
(Table 1, Expt. 3). Comparing the 18 f value measured with-
out the Rh/USY catalyst (0.94; Table 1, Expt. 4) with 0.68,
changes led to the formation of different oxygen species
(18). Changes in the surface structures with weaker O–Rh
bonds may result in the high activity as well as steady-state
O2 production, although this is a subject for future work.
Figure 5 shows the O2-TPD spectrum from the Rh/USY
catalyst after H2 reduction at 500 C followed by O2 treat-
ment (80 Torr) at 300 C, which was the same as used in the
isotope study. The spectrum is similar in shape to spectra B
and C in Fig. 3 and spectrum B in Fig. 4. Oxygen coverage
(
o) was determined to be 1.1 from the amount of desorbed
oxygen (O/Rhs). It should be noted that no O2 desorption
takes place below 500 C.
TABLE 1
Isotopic Fraction 18O (18 f ) and Isotopic Equilibrium constant
(Ke) in the Product Molecules from 18O2, C16O, and N2 16O Pulses
at 220 C
Experiment
Pulse
C16
Surface species
Product
18 f
Ke
1
2
2
3
4
O
18O
18O
18O
16O
—
CO2
O2
N2O
O2
0.69
0.62
0.00a
0.68
0.94b
4.00
1.97
—
N2 16O
N2 16O
18O2
0.37
—
18O2
O2
FIG. 5. Temperature-programmed desorption (TPD) spectrum of O2
from the Rh/USY catalyst after O2 (80 Torr) pretreatment at 300 C.
a The isotopic abundance of 18O is 0.002.
b
18 f in the incident pulse measured without the catalyst.

CATALYTIC DECOMPOSITION OF N2O ON Rh/USY
207
the exchange coefficient of O2 (C) with surface oxygen,
18O2 + 16O(a) → 18O16O + 18O(a),
though incidental exchange reactions0between the product
O2 and the O(a) (i.e., steps [6] and [6 ]) occur, 18 f calc. does
not change: 18O fractions coming from the surface and go-
ing to the surface by the exchange reaction are canceled
(i.e., BC in Table 2), because A = B in the case of the LH
mechanism. As shown in Table 2, the experimental result
is in good agreement with the LH mechanism. In the case
of the ER mechanism (i.e., step [3]), the 18O fraction of
the product O2 should be half the value of the surface oxy-
[6]
[60]
16O2 + 18O(a) → 18O16O + 16O(a),
was estimated to be 0.28. The exchange coefficient (C) rep-
resents the isotope fraction produced during a single pass
of O2 exchanging with the surface oxygen.
After the pulsed CO experiment, an N2 16O pulse was
injected onto the catalyst at 220 C (Table 1, Expt. 2). N2O
conversion was about 15% , and the 18 f of the product O2
was 0.62, which was almost the same as the 18 f of the Rh
surface. In addition, the Ke value of oxygen produced from
N2O decomposition was 1.97, which indicates that the ex-
change of oxygen between gas phase and surface is slow
enough at low reaction temperature (220 C). Furthermore,
the exchange of oxygen in N2O with surface oxygen,
gen (i.e., 0.345). However, considering the exchange coeffi-
18
cient (C = 0.28), the
f
of the product O2 becomes 0.44
calc.
(Table 2), which is quite different from the observed 18f.
Therefore, the ER mechanism can be excluded. It should
be noted that the product O2 is formed during contact with
the catalyst during N2O reaction, whereas the C value is
measured by reaction with just the O2 pulse. This presum-
ably means that, since O2 is produced during the reaction,
the C value should be corrected for the fact that O2 is not
present in the catalyst bed the whole time, and presum-
N2 16O + 18O(a) → N2 18O + 16O(a),
can be neglected because of the very low 18 f value in the
outlet N2O (Table 1, Experiment 2).
[7]
ably be reduced by a factor of 2 (i.e., C = 0.14). This leads
18
to a corrected
f
value for the ER mechanism of 0.39,
calc.
18
which is more different from the observed
f
value. The
calc.
The mechanism of oxygen desorption is determined by
HA mechanism, which produces O2 molecules only from
the following discussion. The 18O fraction of the product O2
N2 16O, can also be excluded. In this case,
f
is 0.19
calc.
18
18
(Table 2) by considering only step [60], because the 16O2
(
f
) can be calculated based on the three mechanisms
calc.
18
that is formed can react with adsorbed 18O. If the corrected
(LH, ER, and HA). However, the
f
of the outlet O2
calc.
18
should be corrected for O2 exchange reactions using the
exchange coefficient (C), as follows:
C value (0.14) is used, the
f
value becomes 0.10, which
calc.
is quite different from the observed 18 f value. If the “hot”
18
oxygen can react with all adsorbed oxygen atoms,
f
calc.
18
f
= A(1 C) + BC.
[8]
calc.
takes the same value as in the ER mechanism. Therefore,
the present result reveals that oxygen desorption proceeds
via the LH mechanism (i.e., recombinative desorption of
O2).
Here, A is the 18O fraction of the product O2 just desorbed
from the catalyst surface, and B is the 18O fraction of the
surface (i.e., B = 0.69; Table 1, Expt. 1). In Eq. [8], BC rep-
resents the 18O fraction in the product O2 coming from the
surface, while A(1 C) represents the product O2 without
going to the surface. Table 2 shows the observed 18 f of the
product O2 produced from N2O decomposition and the cal-
culated 18 f values based on the three mechanisms (LH, ER,
The LH mechanism was also supported at a reaction tem-
perature of 240 C. It should be noted that the oxygen ex-
change coefficient (C) increased with reaction temperature
(C = 0.34 at 240 C, C = 0.71 at 300 C), while oxygen ex-
change with the USY support was negligible (C = 0.05 even
at 500 C). The isotopic experiments were possible only in a
relatively narrow range of temperatures, because at higher
temperatures (e.g., 300 C) Ke was 4.0 due to the high C
value, while at lower temperatures N2O conversions were
too low to obtain reliable results. As mentioned in the Intro-
duction, the LH mechanism (i.e., recombinative desorption
of O2) was also observed on unsupported Rh catalyst. At
present there appears to be no dependence of the support
on the reaction mechanism, because similar results have
been obtained for a Rh/SiO2 catalyst (20).
and HA). In the case of the LH mechanism (2O(a) → O2,
18
i.e., step [2]), the
f
of the product oxygen should be
calc.
the same as that of the surface oxygen (i.e., 0.69). Even
TABLE 2
Observed 18 f and Calculated 18 f Values Based on the
Three Mechanisms
18
f
calc.
18 f of O2
In the present study, all of the surface oxygens are in-
volved in the recombinative desorption of O2 (step [2]) at
low reaction temperatures (220–240 C). The TPD study in
Fig. 5 shows that desorption of O2 in He from the Rh/USY
catalyst does not take place at these temperatures. There-
fore, it should be considered that the energy of O–Rh bond
Mechanism just produced
Equation
C = 0.28 C = 0.14
LH
ER
HA
A = B
B(1 C) + BC
1/2B(1 C) + BC
BC
0.69
0.44
0.19
0.69
0.39
0.10
A = 1/2B
A = 0
18
Observed
f
= 0.62
obs.

208
TANAKA ET AL.
formation accompanied by N2O decomposition may trans- the 18O tracer study, however, O2 desorption does not pro-
fer to the surrounding adsorbed oxygen atoms on the Rh ceed via the ER mechanism (step [3]), but via the LH mech-
surface; i.e., step [2] (2O(a) → O2) proceeds via reaction- anism (step [2]) at low temperatures (220–240 C). On the
assisted desorption, as previously proposed in the case of other hand, the TPD study showed that O2 desorption from
the Rh black catalyst (14). In practice, a large bonding en- the Rh/USY catalyst (in He) after the same O2 treatment as
ergy has been reported for O–Rh (96.8 kcal/mol) (21) and in the isotopic study occurred at much higher temperatures
additional energy is released by formation of an N N bond (>500 C), which suggests reaction-assisted desorption of
(17). These exothermic processes can overcome the energy O2 during N2O decomposition.
loss caused by breakage of the NN–O bond. It should also
be noted that the overall reaction of N2O decomposition is
exothermic (4H = 19.5 kcal/mol).
REFERENCES
1. Kapteijn, F., Rodriguez-Mirasol, J., and Moulijn, J. A., Appl. Catal. B
9, 25 (1996).
There has been some discussion that metal particles
might be much hotter than support materials during an
exothermic reaction such as CO oxidation (22–24). Kember
and Sheppard (22) estimated from infrared emission mea-
surements that for a Pd/SiO2 catalyst during CO oxidation
at 300 C, there is at least 190 C temperature difference be-
tween the silica and metal components. On the other hand,
Sharma et al. (23) concluded that the metal–support tem-
perature gradients in a Pt/SiO2 catalyst are small (<7 C)
during CO oxidation. Matyi et al. (24) observed the local
temperature rise (13–19 C) of an Fe/SiO2 catalyst during
CO hydrogenation. In our pulsed experiment using a small
amount of N2O (0.27 mol), the temperature difference
2. Li, Y., and Armor, J. N., Appl. Catal. B 1, L21 (1992).
3. Oi, J., Obuchi, A., Bamwenda, G. R., Ogata, A., Yagita, H.,
Kushiyama, S., and Mizuno, K., Appl. Catal. B 12, 277 (1997).
4. Leglise, J., Petunchi, J. O., and Hall, W. K., J. Catal. 86, 392 (1984).
5. Centi, G., Dall’Olio, L., and Perathoner, S., Appl. Catal. A 194/195,
79 (2000).
6. Winter, E. R. S., J. Catal. 15, 144 (1969).
7. Lin, J., Chen, H. Y., Chen, L., Tan, K. L., and Zeng, H. C., Appl. Surf.
Sci. 103, 307 (1996).
8. Yuzaki, K., Yarimizu, T., Ito, S., and Kunimori, K., Catal. Lett. 47, 173
(1997).
9. Yuzaki, K., Yarimizu, T., Aoyagi, K., Ito, S., and Kunimori, K., Catal.
Today 45, 129 (1998).
10. Yamashita, T., and Vannice, A., J. Catal. 161, 254 (1996).
between the Rh particles and the USY support may be 11. Root, T. W., Schmidt, L. D., and Fisher, G. B., Surf. Sci. 134, 30
(1983).
small. However, temporal and/or local heat dissipated from
N2O decomposition may facilitate the recombinative des-
orption of O2. Alternatively, the hot O(a) atoms produced
12. Dandl, H., and Emig, G., Appl. Catal. A 168, 261 (1998).
13. Caratzoulas, S., Jackson, B., and Persson, M., J. Chem. Phys. 107, 6420
(1997).
by N2O decomposition may make a rapid exchange with the
surrounding O(a) already present before O2 desorption oc-
curs. However, further discussion is needed to clarify how
the energy transfer takes place on the surface.
14. Uetsuka, H., Aoyagi, K., Tanaka, S., Yuzaki, K., Ito, S., Kameoka, S.,
and Kunimori, K., Catal. Lett. 66, 87 (2000).
15. Kunimori, K., Ikeda, Y., Soma, M., and Uchijima, T., J. Catal. 79, 185
(1983).
16. Kunimori, K., Uchijima, T., Yamada, M., Matsumoto, H., Hattori, T.,
and Murakami, Y., Appl. Catal. 4, 67 (1982).
17. Li, Y., and Bowker, M., Surf. Sci. 348, 67 (1996).
18. Alfe`, D., Rudolf, P., Kiskinova, M., and Rosei, R., Chem. Phys. Lett.
211, 220 (1993).
19. Ozaki, A., “Isotopic Studies of Heterogeneous Catalysis.” Kodansha,
Tokyo, 1977.
CONCLUSIONS
Study of the dependence of oxygen coverage on pulsed
N2O decomposition revealed that activity increased with in-
creasing coverage ( o > 0.5) and that O2 production started
20. Tanaka, S., Kameoka, S., Ito, S., and Kunimori, K., to be published.
at Լ 0.9, and finally, steady-state activity was attained at 21. Pedley, J. B., and Marshall, E. M., J. Phys. Chem. Ref. Data 12, 967
o
(1984).
high
(>1.2). The O2-TPD studies after N2O or O2 pre-
o
22. Kember, D. R., and Sheppard, N., J. Chem. Soc. Faraday Trans. 2 77,
treatments suggest that oxygen species with weaker O–Rh
bonds, which are desorbed below 700 C, play an important
role in the high activity as well as the steady-state O2 pro-
1321 (1981).
23. Sharma, S., Boecker, D., Maclay, G. J., and Gonzalez, R. D., J. Catal.
110, 103 (1988).
duction in the catalytic N2O decomposition. As revealed by 24. Matyi, R. J., Butt J. B., and Schwartz, L. H., J. Catal. 91, 185 (1985).