Hjhy 6dcvb GG
Hjhy 6dcvb GG
Hjhy 6dcvb GG
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
a r t i c l e i n f o a b s t r a c t
Article history: A CeO2–Ag catalyst with a ‘rice-ball’ morphology, consisting of Ag particles in the center surrounded by
Received 21 January 2011 fine CeO2 particles, exhibits exceptional catalytic performance for soot oxidation by O2 below 300 °C. The
Revised 30 May 2011 reaction mechanism over this catalyst was studied by O2 temperature-programmed desorption (O2-TPD),
Accepted 2 July 2011 18
O/16O isotopic exchange (IE) reaction, and electron spin resonance (ESR) techniques. It is speculated
Available online 10 August 2011
that adsorbed oxygen species on the Ag surface migrate to the CeO2 surface via the Ag/CeO2 interface
to form Ox
n species (at least partly O2 ) and further migrate onto the soot particles. Due to morphological
Keywords:
compatibility of the moderately large Ag particles (ca. 30–40 nm) and the extremely large interfacial area
Soot oxidation
Silver
with the CeO2 particles, the formation and migration rates of the oxygen species on the CeO2–Ag catalyst
Ceria are efficiently promoted, resulting in its distinguished catalytic performance and relative insensitivity to
Morphology the contact mode of the soot–catalyst mixture.
Reaction mechanism Ó 2011 Elsevier Inc. All rights reserved.
Diesel emission control
0021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.07.001
290 K. Yamazaki et al. / Journal of Catalysis 282 (2011) 289–298
However, these catalysts have drawbacks in some practical appli- we tried to elucidate the mechanism involved in catalytic soot oxi-
cations that are related to either thermal degradation or selective dation, especially over the CeO2–Ag catalyst with unique ‘‘rice-
leaching in condensed water during the soot oxidation process ball’’ morphology, which exhibits an overwhelmingly high cata-
[15]. lytic performance among the prepared catalysts.
Recently, many researchers have reported that CeO2-based oxi-
des have good activity for soot oxidation at lower temperature 2. Experimental
[16–23]. Aneggi et al. studied the effect of Ag addition on the soot
oxidation activity of various metal oxides [24]; the addition of Ag 2.1. Catalyst preparation
to ZrO2 and Al2O3 resulted in very active catalysts, while addition
to CeO2 had little benefit. On the other hand, Machida et al. and A CeO2–Ag catalyst was synthesized by a coprecipitation-based
Shimizu et al. reported that Ag loading onto CeO2 enhanced the method. One hundred and fifty milliliters aqueous solution of
catalytic activity for soot oxidation [25,26]. However, the design AgNO3 (Toyo Chemical Industrial, 29.63 g) and Ce(NO3)36H2O
concept of these catalysts, i.e., morphology control based on the (Wako Pure Chemical Industries, 50.49 g) was added to a diluted
functions of Ag and CeO2, has not been devoted to the problem ammonia solution (35.6 g of 25% ammonia solution, diluted by
of catalyst–soot particle contact. In addition, silver-supporting cat- 100 mL of water) instantly (in less than a second) with a rotary
alysts may suffer from performance degradation due to Ag sinter- stirrer (350 rpm) at room temperature. The mixture was stirred
ing under oxidative conditions at temperatures around 400 °C [27]. for 1 min, and the coprecipitate was heated uniformly by steam
Considering this background, we have developed an innovative in an autoclave at 120 °C for 10 min. The gained coprecipitate
CeO2–Ag catalyst, which exhibits exceptional performance for soot was separated by centrifugation and calcined at 500 °C for 5 h in
oxidation with gaseous oxygen at temperatures below 300 °C [28]. air. The Ag content of the obtained CeO2–Ag catalyst was 39
This catalyst has a unique agglomeration morphology, as shown in wt.%. A more detailed description of the catalyst preparation is gi-
Fig. 1, in which the scanning electron microscopy (SEM) image (a) ven elsewhere [28].
shows that this catalyst consists of numerous spherical ‘balls’ with Ag(x)/CeO2, Ag(x)+CeO2, Ag(x)/Al2O3, and CeO2 catalysts, where
diameters of about 100 nm and the transmission electron micros- ‘x’ denotes the Ag content in wt.%, were used as reference catalysts.
copy (TEM) image of the cut plane of a single ball (b) shows that The CeO2 catalyst was obtained by calcination of a commercial
the center Ag metal is surrounded by fine CeO2 particles. The mor- CeO2 powder (Rhodia, BET surface area 150 m2/g) at 600 °C for
phology of this catalyst is schematically illustrated in Fig. 2, in 50 h, with a BET surface area of 78 m2/g. The Ag(x)/CeO2 catalysts
which the unique agglomeration nanostructure is similar to a Jap- were prepared by impregnation of the CeO2 catalyst with aqueous
anese ‘rice-ball’, consisting of a center composed of Ag particles AgNO3 solution and calcination at 500 °C for 5 h in air. The Ag(x)+
(rice-ball filling) surrounded by fine CeO2 particles (grains of rice). CeO2 catalysts were prepared by physical mixing of the CeO2 cata-
This morphology was designed to increase the Ag/CeO2 interface lyst and a commercial Ag powder (Nisshin Engineering, BET surface
area per unit surface area of Ag particles and to inhibit Ag sintering, area 9 m2/g) with a magnetically driven mortar for 20 min fol-
because thermally stable CeO2 particles act as barriers to sintering. lowed by calcination at 300 °C for 5 h in air. The Ag(x)/Al2O3 cata-
A CeO2–Ag catalyst with such morphology was synthesized by a lysts were prepared by impregnation of a commercial Al2O3
novel nanofabrication method based on precipitation between powder (Showa Denko, UA-5205 with BET surface area 25 m2/g)
aqueous solutions of nitrates and ammonia [28]. with aqueous AgNO3 solution followed by calcination at 500 °C
In this study, temperature-programmed oxidation (TPO) over for 5 h in air.
the soot/catalyst mixture under tight and loose contact modes
was used to evaluate the catalytic performance. To reveal the dif- 2.2. Catalytic performance evaluation
ferent aspects involved in catalytic soot oxidation, oxygen temper-
ature-programmed desorption (O2-TPD), 18O/16O isotopic The catalytic performance for soot oxidation was evaluated by
exchange (IE) reaction, and electron spin resonance (ESR) tech- temperature-programmed oxidation (TPO) of soot–catalyst mix-
niques were employed to characterize the active property of oxy- tures so as to compare the CeO2–Ag catalyst with other reference
gen species, oxygen mobility or migration rates and identification catalysts. Two kinds of carbon black powder (Degussa AG, Prin-
of oxygen species, respectively. Through discussing these charac- tex-V with BET surface area 85 m2/g, and Degussa AG, Printex-U
terization results and correlating with their catalytic performances, with BET surface area 92 m2/g) were used as the model soot in this
500nm
Ag
100nm
Fig. 1. (a) SEM image of CeO2–Ag catalyst particles and (b) TEM image of the cut plane of a single spherical agglomerate of CeO2–Ag catalyst [28].
K. Yamazaki et al. / Journal of Catalysis 282 (2011) 289–298 291
CeO 2
Cross section
Ag
100 nm
: Ag : CeO 2
study. The tight contact and loose contact modes were adopted for X-ray diffraction (XRD; Rigaku, RINT-TTR) patterns of the pow-
the soot–catalyst mixture [1,6]. The tight contact was attained by der samples were recorded using Cu Ka radiation and a fixed
grinding catalyst with Printex-V using a magnetically driven mor- power source (50 kV and 300 mA). The data were obtained be-
tar for 5 min. The loose contact was achieved by mixing catalyst tween 20° and 60° (2h) with a step of 0.02° and for 1 s per step.
with Printex-U loosely using a spatula for 10 min in a reproducible The average particle sizes of all the phases present in the catalysts
way. Because the grain size of the Printex-U is smaller than that of were estimated using Scherrer’s equation.
the Printex-V, for a better contact between the soot and catalyst, Oxygen temperature-programmed desorption (O2-TPD) was
the Printex-U was used for the samples in the loose contact mode. performed in a fixed-bed reactor. For O2 pre-adsorption, 160 mg
The weight ratios of the soot to catalyst were both 1:19. of the catalyst was heated in a 20 mL/min O2 flow at 600 °C for
In TPO experiments, the heat and mass transfer were con- 30 min and cooled down to room temperature. After purging the
cerned. Against heat and mass transfer limitations, two measures catalyst with Ar for 30 min at room temperature, the catalyst
were taken as follows: for heat release from soot combustion, was heated to 600 °C at 20 °C/min under a 20 mL/min Ar flow.
quartz wool (Tosoh, Fine grade) was used to dilute samples of cat- The corresponding O2 desorption spectrum was detected using
alyst–soot mixtures. For mass transfer limitation, the gas of 10% the Q-MS.
O2/He was flowed at a relative high flow rate (50 mL/min), Isotopic exchange (IE) reaction experiments were carried out in
although just 40 mg of the soot–catalyst mixture was put in a a re-circulated reactor (ca. 100 cm3) coupled to the Q-MS. Masses
fixed-bed flow reactor. The two measures ensured that there was of 32, 34, 36 (oxygen isotopomers, 16O2, 18O16O, and 18O2), and
no apparent temperature rises caused by soot oxidation, and mass 28 (to detect a possible leak) were continuously monitored. The
transfer limitation could be ignored. So, we can say that the reac- vacuum connection to the mass spectrometer was thermoregulat-
tion regime was mostly controlled by the chemical kinetics, in- ed to maintain a constant pressure of 104 Pa, while the pressure in
stead of by mass and heat transfer. A thermocouple was inserted the reactor loop was 5650 Pa. Thirty milligrams of the catalyst was
into the soot–catalyst mixtures in order to monitor the reaction placed into a quartz reactor for in situ experimentation. After oxi-
temperature. TPO spectra were recorded at a heating rate of dation with oxygen at 500 °C for 30 min and evacuation at 500 °C
20 °C/min. The concentration of CO2 and CO in the effluent was for 10 min, the sample was cooled to 300 °C; 5650 Pa of pure
18
analyzed online using a quadrupole mass spectrometer (Q-MS; Ul- O2 (99.2% of 18O, ICON) was then introduced at 300 °C, and the
vac, RG-102). Soot oxidation was regarded as the sum of the partial pressure variation of the oxygen isotopomers, P36 (18O2),
amount of CO2 and CO formed during TPO; however, the amount P34 (18O16O), and P32 (16O2), was continuously recorded. The total
of CO formed was much less than that of CO2 in the presence of pressure (P36, P34, and P32) remained virtually constant. A more de-
all of the catalysts examined. CO was only observed in the absence tailed description of this experiment is given elsewhere [29,30].
of the catalyst. The catalytic performance in this study is indicated From the partial pressure values, the 18O atomic fraction ag(t) in
by the temperature corresponding to the maximum soot oxidation the gas phase at a time t can be defined as:
rate (Tmax) derived from the TPO spectra, with reference to the
ag ðtÞ ¼ ½ð1=2ÞP34 ðtÞ þ P 36 ðtÞ=½P36 ðtÞ þ P34 ðtÞ þ P32 ðtÞ ð1Þ
method by Moulijn et al. [1,6]. A lower Tmax value indicates higher
catalytic performance for soot oxidation. The IE reaction at a time t, IE(t), corresponds to the disappearance of
18
O from the gas phase; therefore, it is possible to determine IE(t)
2.3. Catalyst characterization (in moles of oxygen molecules per catalyst weight) using the fol-
lowing equation [29]:
The BET surface area was measured by single-point N2 adsorp- IEðtÞ ¼ ng ½1 ag ðtÞ=wc
tion at 196 °C using an automatic surface area analyzer (Microda-
¼ ðPT =RÞðV r =T r þ V c =T c Þ½1 ag ðtÞ=wc ð2Þ
ta, MS4232II) after pretreatment at 200 °C for 20 min.
292 K. Yamazaki et al. / Journal of Catalysis 282 (2011) 289–298
where ng is the total number of moles of oxygen molecules in the The particle sizes of the Ag and CeO2 phases and lattice spacings
gas phase, wc is the catalyst weight, PT is the total pressure, R is of Ag(1 1 1) and CeO2(1 1 1) calculated from the XRD patterns of
the ideal gas constant, Vr and Vc are the volumes of the heated the catalysts are summarized in Table 1. The lattice spacing of
and unheated zones of the reactor, respectively, and Tr and Tc are Ag(1 1 1) and CeO2(1 1 1) in the CeO2–Ag catalyst is consistent
the temperatures of the heated and unheated zones of the reactor, with that of Ag(1 1 1) in the Ag powder and that of CeO2(1 1 1)
respectively. The rate of IE reaction (in moles of oxygen molecules in the CeO2 catalyst, respectively. Thus, no solid reaction occurred
per catalyst weight and per time unit), RIE, can be calculated using between the Ag and CeO2 phases in this catalyst. In a similar fash-
the following equation: ion, no solid reaction occurred in the Ag(x)/CeO2 and Ag(x)+CeO2
catalysts, irrespective of the Ag content. The particle sizes of the
RIE ¼ ðng =wc Þdag ðtÞ=dt Ag (36 nm) and CeO2 (16 nm) phases from the XRD patterns of
¼ ðPT =Rwc ÞðV r =T r þ V c =T c Þdag ðtÞ=dt ð3Þ the CeO2–Ag catalyst are consistent with the TEM image (Fig. 1b).
Table 1
Morphology and structural properties of catalysts; BET surface area, particle size, and lattice spacing.
Catalyst BET surface area (m2/g) Particle size (nm) Lattice spacing (nm) Peak intensity ratio (–) Ag(1 1 1)/CeO2(1 1 1)
Ag CeO2 Ag(1 1 1) CeO2(1 1 1)
CeO2–Ag 14.7 36 16 0.2358 0.3121 0.73
Ag(39)/CeO2 30.1 89 21 0.2359 0.3124 0.37
Ag(10)/CeO2 52.0 60 20 0.2359 0.3121 0.11
Ag(3.2)/CeO2 59.2 28 20 0.2359 0.3122 0.03
Ag(1.9)/CeO2 70.0 20 20 0.2356 0.3121 0.02
Ag(0.95)/CeO2 78.1 n.d.a 20 – 0.3120 –
Ag(39)+CeO2 49.9 47 20 0.2358 0.3120 0.85
Ag(10)+CeO2 72.0 35 20 0.2358 0.3122 0.14
Ag(3.2)+CeO2 77.2 31 20 0.2358 0.3122 0.05
Ag(1.9)+CeO2 77.7 32 20 0.2359 0.3122 0.04
Ag(0.95)+CeO2 78.0 34 20 0.2355 0.3120 0.02
Ag(39)/Al2O3 12.7 48 – 0.2359 – –
CeO2 77.8 – 20 – 0.3120 –
Ag powder 9.6 34 – 0.2359 – –
a
n.d.: not detected.
CO 2 + CO concentration (%)
2: Ag(39)/CeO2 2
(a) 3 2 4
3: Ag(39)+CeO2 (a) 4 2: Ag(39)/CeO 2
3: Ag(39)+CeO 2
5 4: Ag(39)/Al2O3 4: Ag(39)/Al2O 3
1 1 3 5
5: CeO2 2 5: CeO 2
1' 6: non-catalyst 6: non-catalyst
1 1 1'
6 6
0 0
200 300 400 500 600 700 200 300 400 500 600 700
CO2 + CO concentration (%)
2
CO 2 + CO concentration (%)
2 2: Ag(39)/CeO 2
(b) 8
2: Ag(39)/CeO2 (b) 7: Ag(10)/CeO 2
7: Ag(10)/CeO2 8: Ag(3.2)/CeO 2
7 2
8: Ag(3.2)/CeO2 2 9: Ag(1.9)/CeO 2
10
9 9: Ag(1.9)/CeO2 10: Ag(0.95)/CeO 2
1 1
10: Ag(0.95)/CeO2 8 9
10 7
0 0
200 300 400 500 600 700 200 300 400 500 600 700
CO2 + CO concentration (%)
2
CO 2 + CO concentration (%)
2 3: Ag(39)+CeO 2
(c) 3 11 3: Ag(39)+CeO2 (c) 11: Ag(10)+CeO 2
12 12: Ag(3.2)+CeO 2
11: Ag(10)+CeO2
3 13: Ag(1.9)+CeO 2
12: Ag(3.2)+CeO2 11
13 13: Ag(1.9)+CeO2 12 14: Ag(0.95)+CeO 2
1 1
14: Ag(0.95)+CeO2 13 14
14
0 0
200 300 400 500 600 700 200 300 400 500 600 700
Temperature (oC) Temperature (oC)
Fig. 4. (a) TPO profiles for soot oxidation in the presence and absence of CeO2–Ag, Fig. 5. (a) TPO profiles for soot oxidation in the presence and absence of CeO2–Ag,
Ag(39)/CeO2, Ag(39) + CeO2, Ag(39)/Al2O3, CeO2 catalysts in tight contact mode. (b Ag(39)/CeO2, Ag(39) + CeO2, Ag(39)/Al2O3, CeO2 catalysts in loose contact mode. (b
and c) TPO profiles for soot oxidation in the presence of Ag(x)/CeO2 and Ag(x) + CeO2 and c) TPO profiles for soot oxidation in the presence of Ag(x)/CeO2 and Ag(x) + CeO2
catalysts in tight contact mode. catalysts in loose contact mode.
loose contact modes, the TPO profile with the repeated experiment slightly shifted to higher temperatures in all cases. This indicated
(curve 10 ) exhibits almost the same catalytic performance to that that the CeO2–Ag catalyst somewhat deteriorated after experi-
with the first TPO run in both contact modes. This result indicates enced the first test. It is assumed that the rice-ball morphology
that the soot oxidation activity of the CeO2–Ag catalyst is almost of the CeO2–Ag subtly collapses during the soot oxidation, or some
reproducible, and the CeO2–Ag catalyst has the potential for future unburned fine soot particles and ash generated in the first run
actual application. It was also noted that the second spectrum caused bad contact between the used catalyst and fresh soot.
294 K. Yamazaki et al. / Journal of Catalysis 282 (2011) 289–298
Table 2
18
Catalytic performance for soot oxidation and parameters from O2-TPD, O/16O IE reaction, and ESR experiments.
Catalyst Tmax (°C) DTmaxa (°C) O2 desorption (lmol/g) IE reaction rate (lmol/g min) Spin density (lmol/g)
Tight contact Loose contact
CeO2–Ag 315 376 61 33.5 22.6 0.830
Ag(39)/CeO2 381 563 182 9.1 9.1 0.232
Ag(10)/CeO2 362 526 164 17.0 12.9 0.465
Ag(3.2)/CeO2 371 550 179 9.3 7.4 –
Ag(1.9)/CeO2 414 596 182 2.1 3.9 –
Ag(0.95)/CeO2 466 610 144 0.5 – –
Ag(39) + CeO2 351 461 110 23.5 – –
Ag(10) + CeO2 342 443 101 25.0 – –
Ag(3.2) + CeO2 355 522 167 18.0 – –
Ag(1.9) + CeO2 384 548 164 5.0 6.4 –
Ag(0.95) + CeO2 433 597 164 1.9 – –
Ag(39)/Al2O3 480 575 95 2.0 2.7 0.091
CeO2 462 590 128 0.0 0.6 0.002
None (soot only) 660b 640c – – –
a
DTmax = (Tmax in loose contact mode) (Tmax in tight contact mode).
b
Tmax of Printex-V.
c
Tmax of Printex-U.
0.2 650
1
(a) 1'
O 2 concentration (%)
1: CeO2-Ag
1': CeO2-Ag (2nd run) 600
3 2: Ag(39)/CeO2
3: Ag(39)+CeO2
0.1
2 4: Ag(39)/Al2O3 550 loose contact
4 5: CeO2
5
500
Tmax (oC)
0
100 200 300 400 500 600
450
0.2
(b)
O 2 concentration (%)
2: Ag(39)/CeO2 400
6: Ag(10)/CeO2
7: Ag(3.2)/CeO2
0.1 8: Ag(1.9)/CeO2 350
2 9: Ag(0.95)/CeO2 tight contact
6 7
8 9
300
0 10 20 30 40
0 O2 desorption (µmol/g)
100 200 300 400 500 600
0.2 Fig. 7. Tmax value from TPO experiments in tight and loose contact modes as a
(c) function of the amount of O2 desorption determined from O2-TPD experiments.
O 2 concentration (%)
3: Ag(39)+CeO2
10 10: Ag(10)+CeO2
3 11: Ag(3.2)+CeO2
11 12: Ag(1.9)+CeO2 200 and 450 °C, corresponding to the light-off of soot oxidation be-
0.1 12 13: Ag(0.95)+CeO2 low 300 °C over the CeO2–Ag catalyst (Figs. 4a and 5a). O2 desorp-
13 tion is also observed from the Ag(x)/CeO2, Ag(x)+CeO2, and Ag(39)/
Al2O3 catalysts in the same temperature range, but not for the CeO2
catalyst.
0 The amounts of O2 desorption from the catalysts between 200
100 200 300 400 500 600
and 450 °C are summarized in Table 2. The CeO2–Ag catalyst exhib-
Temperature (oC)
its the largest O2 desorption among all the catalysts. Fig. 7 reveals
Fig. 6. (a) O2-TPD profiles of CeO2–Ag, Ag(39)/CeO2, Ag(39) + CeO2, Ag(39)/Al2O3, an excellent correlation between the Tmax values for TPO in tight
CeO2 catalysts. (b and c) O2-TPD profiles of Ag(x)/CeO2 and Ag(x) + CeO2 catalysts. and loose contact modes, respectively, as a function of the amounts
of O2 desorption from O2-TPD over all the catalysts. These results
Further investigations are necessary to clarify the involved reasons strongly suggest that active oxygen species for soot oxidation cor-
for this deactivation, as well as more efforts to improve its respond to the adsorbed oxygen species on the catalyst from gas-
thermostability. eous O2, which is desorbed in the temperature range between
200 and 450 °C.
3.3. O2-TPD The O2-TPD experiment was also repeated with the used CeO2–
Ag catalyst. As indicated in Fig. 6a, almost the same O2-TPD spectra
Fig. 6a–c show O2-TPD profiles of the CeO2–Ag and reference of the CeO2–Ag catalyst were gained in first and second runs. This
catalysts. The CeO2–Ag catalyst exhibits O2 desorption between result demonstrates that the gaseous oxygen could be adsorbed
K. Yamazaki et al. / Journal of Catalysis 282 (2011) 289–298 295
15
The IE reaction as a informative technique was employed to
characterize oxygen mobility over the catalysts [29,30] in this
study. As shown in Fig. 8a, the partial pressures of three oxygen 10
molecules, P36 (18O2), P34 (18O16O), and P32 (16O2), evolve during
the 18O/16O IE reaction over the CeO2–Ag catalyst at 300 °C. The 5
18
O2 molecules in the gas phase adsorb and decompose on the cat-
alyst surface and then exchange and desorb to 18O16O and 16O2
0
molecules into the gas phase. Fig. 8b and c show IE reaction profiles 0 10 20 30 40
against reaction time over the CeO2–Ag and reference catalysts de- O 2 desorption (µmol/g)
rived by Eq. (2). At 300 °C, the IE reaction proceeds the fastest over
the CeO2–Ag catalyst, apparently occurs over the Ag(x)/CeO2, Fig. 9. IE reaction rates for the 18O/16O IE reaction versus the amount of O2
desorption determined from O2-TPD experiments.
Ag(x)+CeO2, and Ag(39)/Al2O3 catalysts, but is detectable over the
CeO2 catalyst, indicating that the co-existence or inter-particle
contact of Ag particle and CeO2 is truly necessary for the IE reaction exchangeable with the oxygen in CeO2, bridged by the interface be-
or oxygen migration. tween the Ag and the CeO2 particles of the catalysts. As the result,
The initial IE reaction rates by Eq. (3) are summarized in Table 2. the CeO2–Ag catalyst with the unique ‘‘rice-ball’’ morphology
The CeO2–Ag catalyst shows higher IE reaction rate than the Ag(x)/ exhibits the fastest oxygen migration via the interface between
CeO2, Ag(x)+CeO2, Ag(39)/Al2O3, and CeO2 catalysts. The IE reaction the Ag and the CeO2 particles among all the catalyst samples.
rates or migration rate closely associates with oxygen species Fig. 9 shows a good correlation between the IE reaction rates for
adsorbing and migrating on Ag surface, the interface between Ag the 18O/16O IE reaction and the amounts of O2 desorption from O2-
and CeO2, and the surface and bulk of CeO2. These results above re- TPD measurements.
veal that the active oxygen species are potentially mobile from the
gaseous O2 onto the Ag surface through adsorption and actively 3.5. ESR
5000 to room temperature under the same gas flow and exposure to
He flow for 30 min at room temperature), no ESR signals of oxygen
4000
: P36 species were observed. However, when exposed to 3% H2/N2 flow
3000 : P34 at 200 °C for 2 min after the same pretreatment, ESR signals with
2000 : P32 g// = 2.046, g// = 2.037, and g\ = 2.010 were observed, as shown in
Fig. 10. These signals are all attributed to superoxide (O2 ) species
1000
bonded to Ce cations [31,32]. ESR analyses of other catalysts using
0 the same procedure also provide signals with the same g values.
0 20 40 60 80 Among all of the catalysts shown in Fig. 9, the CeO2–Ag catalyst
1000 : CeO2-Ag shows the highest ESR signal intensity.
(b) : The spin densities of O 2 species over the catalysts are listed
IE reaction (µmol/g)
Ag(39)/CeO2
800 : Ag/Al2O3 in Table 2. Fig. 11 shows a linear correlation between the spin
: CeO2
600
g =2.037
400
g =2.046 g =2.010
200
0 CeO 2-Ag
0 20 40 60 80
Intensity (arb. unit)
1000 : Ag(39)/CeO2
(c) : Ag(10)/CeO2
IE reaction (µmol/g)
800 : Ag(3.2)/CeO2
: Ag(1.9)/CeO2 Ag(39)/CeO 2
600 : Ag(1.9)+CeO2
Ag(10)/CeO 2
400
200 Ag(39)/Al2O 3
0
0 20 40 60 80 CeO 2
Time (min)
3100 3300 3500 3700
Fig. 8. (a) Partial pressure evolution of 18O2, 18O16O, and 16O2 molecules for the IE Magnetic field (gauss)
reaction over the CeO2–Ag catalyst at 300 °C. (b) IE reaction profiles of CeO2–Ag,
Ag(39)/CeO2, Ag(39)/Al2O3, CeO2 catalysts. (c) IE reaction profiles of Ag(x)/CeO2 and Fig. 10. ESR spectra of CeO2–Ag, Ag(39)/CeO2, Ag(10)/CeO2, Ag(39)/Al2O3, CeO2
Ag(x) + CeO2 catalysts. catalysts.
296 K. Yamazaki et al. / Journal of Catalysis 282 (2011) 289–298
erately large Ag particles (ca. 30–40 nm) and an extremely large of oxygen species. In this case, transformation of the active oxygen
interface area with CeO2 particles. species is required at the interface. Therefore, migration via the
interface should be a rate determining step for the complete migra-
4.2. Effect of rice-ball morphology on the migration of active oxygen tion of the active oxygen species during soot oxidation. 18O/16O IE
species for soot oxidation reactions reveal that the CeO2–Ag catalyst exhibits the fastest oxy-
gen migration via the interface (Fig. 8b and c). The interface area
The soot used in this study cannot physically contact the Ag between the Ag and CeO2 particles in the present CeO2–Ag catalyst
particles in the center of the CeO2–Ag catalyst particles, because is extremely large; therefore, it can be concluded that the rice-ball
the particle size (ca. 100 nm) is too large, and soot oxidation pro- morphology is also suitable for the fast migration of active oxygen
ceeds efficiently on the CeO2 surface surrounding the spherical species for soot oxidation. Two functions of the interface, faster
agglomerate. Therefore, it is supposed that the active oxygen spe- migration and increased formation of the active oxygen species,
cies generated on the Ag surface from gaseous O2 migrates to soot were confirmed by the good correlation in Fig. 9.
particles on the CeO2 surface during soot oxidation. In this section,
we investigate the migration of the active oxygen species for soot 4.3. Mechanism of soot oxidation over CeO2–Ag catalyst
oxidation over the CeO2–Ag catalyst.
ESR analyses show that some atomic oxygen species adsorbed We propose a possible mechanism for soot oxidation over the
on the Ag surface from gaseous O2 under oxidative conditions are CeO2–Ag catalyst, depicted in Fig. 12, that illustrates the morphol-
transformed into O2– species on the CeO2 surface under weak ogy of the catalyst. First, gaseous O2 is adsorbed on the surface of
reductive conditions (Fig. 11). It is known that the catalyst is ex- Ag particles through a synergistic effect with CeO2 particles to form
posed to local reductive conditions during soot oxidation [20]. atomic oxygen species, which are the first active oxygen species for
Thus, it is likely that the active oxygen species formed on the Ag soot oxidation. The extremely large interface between Ag and CeO2
surface of the CeO2–Ag catalyst migrate via the interface between particles and the moderately large size of the Ag particles contrib-
Ag and CeO2 particles to the CeO2 surface, of which a certain frac- ute to the formation of the species. The atomic oxygen species on
tion are transformed into the O 2 species during soot oxidation. the Ag surface then migrate to the surface of CeO2 particles via
Surface mobility phenomena of active oxygen species on CeO2- the interface, transforming into Ox n species (at least partly O2 ),
containing oxides have been invoked by many authors and consid- which is the second active oxygen species. The extremely large
ered as an elementary step of the reaction mechanisms interface contributes to the fast migration of the species. The first
[17,29,30,44–46]. Martin et al. reported that the reoxidation of and second active oxygen species exist in equilibrium during soot
pre-reduced CeO2 proceeds with very mobile oxygen species and oxidation. Finally, the Ox
n species on the CeO2 surface, which is a
leads to an excess of oxygen uptake due, in particular, to the pres- mobile active oxygen species, migrates onto the soot particle sur-
ence of different oxygen species, which could be superoxide (O 2) faces through the contact surface between CeO2 and soot particles,
and peroxide (O2 2 ) species [46]. Krishna et al. reported that CeO 2 oxidizes the soot completely to CO2, and finally releases into the
and rare-earth-modified CeO2 catalysts function to increase active gas phase. The mobile active oxygen species abundantly formed
oxygen transfer to the soot surface, followed by chemisorption of in the interior portion of catalyst, migrate fast out to the external
the active oxygen to form surface oxygen complexes during soot surface, and efficiently access to soot particles. This results in its
oxidation [17]. Therefore, it is very likely that active Ox n species outstanding catalytic performance for soot oxidation, but insensi-
(n = 1 or 2, x = 1 or 2) on the CeO2 surface of the CeO2–Ag catalyst tivity to the contact between catalyst and soot.
also migrate onto the soot particles for soot oxidation. Most of the Good correlations between the catalytic performance for soot
Ox
n species would be O2 , although there remains the possibility for oxidation and all parameters concerning the active oxygen species
the presence of other oxygen species. of catalysts in this study (Figs. 7, 9 and 11) demonstrate that the
In general, the interface area between Ag and CeO2 is smaller mechanism for soot oxidation over Ag(x)/CeO2, conventional sup-
than the Ag and CeO2 surface areas in catalysts composed of Ag ported catalysts, and Ag(x)+CeO2, particle mixed catalysts should
and CeO2 particles, which result in a bottleneck for the migration be the same as that over the CeO2–Ag catalyst, assuming that the
gas phase
CO 2
x-
On soot
O2 (gaseous)
CO 2 gas phase
x-
O n x- x- On
On
O n x- O n x-
CeO 2 O n x- O n x-
x- x-
O n x- On On O n x-
O* O* O* O* x
O n x- O* O* On - x
On -
O* O*
O* O*
O n x- O* O*
x
On -
O* Ag O*
O n x- O* O* O n x-
Fig. 12. Schematic mechanism for soot oxidation over the CeO2–Ag catalyst.
298 K. Yamazaki et al. / Journal of Catalysis 282 (2011) 289–298
Ag and CeO2 particles and their interface have the same functions. is concluded that the rice-ball morphology of the CeO2–Ag catalyst
Some researchers have suggested mechanisms for soot oxidation is optimal for the formation and migration of active oxygen species
over Ag(x)/CeO2 catalysts; however, they require further informa- for soot oxidation.
tion, such as the mobility of the active oxygen species, the respec-
tive role of Ag and CeO2 particles, and the factors influencing the Acknowledgments
activity for soot oxidation. By employing the rice-ball morphology
in a CeO2–Ag catalyst, the configuration of Ag, CeO2, and soot par- The authors thank K. Domen at the Department of Chemical
ticles can be set, which assists in understanding the mechanism of System Engineering, The University of Tokyo, for helpful
soot oxidation over the Ag(x)/CeO2 and CeO2–Ag catalysts. discussions.
On the other hand, the CeO2 catalyst does not desorb O2, as
determined from O2-TPD experiments; however, it does have some References
catalytic performance for soot oxidation (Figs. 4a and 5a). Several
researchers have studied CeO2 and rare-earth-modified CeO2 cata- [1] B.A.A.L. van Setten, M. Makkee, J.A. Moulijn, Catal. Rev. Sci. Eng. 43 (2001) 489.
[2] K. Hinot, H. Burtscher, A.P. Weber, G. Kasper, Appl. Catal. B 71 (2007) 271.
lysts for soot oxidation and have proposed mechanisms for soot [3] B.J. Cooper, J.E. Thoss, SAE Paper 890404, 1989.
oxidation over these catalysts that are based on their oxygen stor- [4] K.N. Pattas, C.C. Michalopoulou, SAE Paper 920362, 1992.
age and release capacities [16–21] and should therefore be differ- [5] G. Lepperhoff, H. Luders, P. Barthe, J. Lemaire, SAE Paper 950369, 1995.
[6] J.P.A. Neeft, M. Makkee, J.A. Moulijn, Appl. Catal. B 8 (1996) 57.
ent from that over the CeO2–Ag catalyst, although such [7] C. Badini, G. Saracco, V. Serra, Appl. Catal. B 11 (1997) 307.
mechanisms may be involved in the case of the CeO2–Ag catalyst. [8] G. Saracco, C. Badini, N. Russo, V. Specchia, Appl. Catal. B 21 (1999) 233.
[9] G. Neri, G. Rizzo, S. Galvagno, A. Donato, M.G. Musolino, R. Pietropaolo, Appl.
Catal. B 42 (2003) 381.
5. Conclusion [10] H. An, C. Kilroy, P.J. McGinn, Catal. Today 98 (2004) 423.
[11] C.A. Querini, L.M. Cornaglia, M.A. Ulla, E.E. Miro, Appl. Catal. B 20 (1999) 165.
An innovative CeO2–Ag catalyst with a unique agglomeration [12] M.L. Pisarello, V. Milt, M.A. Peralta, C.A. Querini, E.E. Miro, Catal. Today 75
(2002) 465.
morphology similar to a rice-ball, which consists of a central Ag [13] R. Jimenez, X. Garcia, C. Cellier, P. Ruiz, A.L. Gordon, Appl. Catal. A 314 (2006)
particle agglomerate surrounded by fine CeO2 particles, exhibits 81.
outstanding performance for soot oxidation with gaseous O2 below [14] D. Fino, N. Russo, G. Saracco, V. Specchia, J. Catal. 217 (2003) 367.
[15] B.A.A.L. van Setten, C.G.M. Spitters, J. Bremmer, A.M.M. Mulders, M. Makkee,
300 °C in tight and loose contact modes. The catalytic performance J.A. Moulijn, Appl. Catal. B 42 (2003) 337.
is higher than those of conventional Ag/CeO2 supported catalysts [16] A. Bueno-Lopez, K. Krishna, M. Makkee, J.A. Moulijn, J. Catal. 230 (2005) 237.
and of Ag+CeO2 mixed particle catalysts irrespective of the Ag con- [17] K. Krishna, A. Bueno-Lopez, M. Makkee, J.A. Moulijn, Appl. Catal. B 75 (2007)
189.
tent, in addition to those of Ag/Al2O3 and CeO2 catalysts, in both [18] K. Krishna, A. Bueno-Lopez, M. Makkee, J.A. Moulijn, Appl. Catal. B 75 (2007)
contact modes. Moreover, the CeO2–Ag catalyst is less sensitive 210.
to contact between catalyst and soot than the other reference [19] E. Aneggi, M. Boaro, C. de Leitenburg, G. Dolcetti, A. Trovarelli, Catal. Today 112
(2006) 94.
catalysts. [20] E. Aneggi, M. Boaro, C. de Leitenburg, G. Dolcetti, A. Trovarelli, J. Alloys Compd.
The reaction mechanism and high performance for soot oxida- 408-412 (2006) 1096.
tion over the CeO2–Ag catalyst were investigated. O2-TPD experi- [21] E. Aneggi, C. de Leitenburg, G. Dolcetti, A. Trovarelli, Catal. Today 114 (2006)
40.
ments show that atomic oxygen species adsorbed weakly on the
[22] P. Palmisano, N. Russo, P. Fino, D. Fino, C. Badini, Appl. Catal. B 69 (2006) 85.
Ag surface through a synergistic effect with CeO2 particles from [23] T. Masui, K. Minami, K. Koyabu, N. Imanaka, Catal. Today 117 (2006) 187.
gaseous O2, which are desorbed in the temperature range between [24] E. Aneggi, J. Llorca, C. de Leitenburg, G. Dolcetti, A. Trovarelli, Appl. Catal. B 91
200 and 450 °C, function as active oxygen species for soot oxida- (2009) 489.
[25] M. Machida, Y. Murata, K. Kishikawa, D. Zhang, K. Ikeue, Chem. Mater. 20
tion. 18O/16O IE reaction experiments revealed that atomic oxygen (2008) 4489.
species on the Ag surface can migrate to the CeO2 particles via the [26] K. Shimizu, H. Kawachi, A. Satsuma, Appl. Catal. B 96 (2010) 169.
Ag/CeO2 interface. ESR analysis indicated that some of the atomic [27] S.R. Seyedmonir, D.E. Strohmayer, G.J. Guskey, G.L. Geoffroy, M.A. Vannice, J.
Catal. 93 (1985) 288.
oxygen species on the Ag surface migrate to the CeO2 surface and [28] T. Kayama, K. Yamazaki, H. Shinjoh, J. Am. Chem. Soc. 132 (2010) 13154.
transform into Ox n species (n = 1 or 2, x = 1 or 2, at least partly [29] F. Dong, A. Suda, T. Tanabe, Y. Nagai, H. Sobukawa, H. Shinjoh, M. Sugiura, C.
O2 species) during soot oxidation. These results suggest that the
Descorme, D. Duprez, Catal. Today 90 (2004) 223.
[30] F. Dong, A. Suda, T. Tanabe, Y. Nagai, H. Sobukawa, H. Shinjoh, M. Sugiura, C.
O2 species on the CeO2 surface also function as the active oxygen Descorme, D. Duprez, Catal. Today 93-95 (2004) 827.
species for soot oxidation. A possible mechanism for soot oxidation [31] X. Zang, K.J. Klabunde, Inorg. Chem. 31 (1992) 1706.
was proposed, where the active oxygen species formed on the Ag [32] X. Li, A. Vannice, J. Catal. 151 (1995) 87.
[33] L. Gang, B.G. Anderson, J. van Grondelle, R.A. van Santen, Appl. Catal. B 40
surface from gaseous O2 migrate to the CeO2 surface via the inter- (2003) 101.
face, transform into Oxn species, and then further migrate onto soot [34] C.T. Campbell, Surf. Sci. 157 (1985) 43.
particles where oxidation occurs. The abundantly formed Ox n spe-
[35] S.N. Trukhan, V.P. Ivanov, B.S. Bal’zhinimaev, Kinet. Catal. 38 (1997) 565.
[36] G.W. Busser, O. Hinrichsen, M. Muhler, Catal. Lett. 79 (2002) 49.
cies migrate fast out to the external surface and efficiently access
[37] M.J. Lippits, A.C. Gluhoi, B.E. Nieuwenhuys, Catal. Today 137 (2008) 446.
to soot particles. This is the key factor for its distinguished soot oxi- [38] B.S. Bal’zhinimaev, Kinet. Catal. 40 (1999) 795.
dation performance and insensitivity to the contact between cata- [39] A. Takahashi, N. Hamakawa, I. Nakamura, T. Fujitani, Appl. Catal. A 294 (2005)
lyst and soot. 34.
[40] G.I.N. Waterhouse, G.A. Bowmaker, J.B. Metson, Appl. Catal. A 265 (2004) 85.
The CeO2–Ag catalyst shows the abundant formation of active [41] L. Kundakovic, M. Flytzani-Stephanopoulos, Appl. Catal. A 183 (1999) 35.
oxygen species due to the compatibility of the moderately large [42] J.C. Wu, P. Harriott, J. Catal. 39 (1975) 395.
Ag particles (ca. 30–40 nm) and the extremely large interface area [43] X.E. Verykios, F.P. Stein, R.W. Coughlin, J. Catal. 66 (1980) 368.
[44] C. Li, Y. Song, Y. Chen, Q. Xin, X. Han, W. Li, Stud. Surf. Sci. Catal. 112 (1997)
between the Ag and the CeO2 particles due to the rice-ball mor- 439.
phology. This catalyst also exhibits fast migration of the active oxy- [45] D. Duprez, Stud. Surf. Sci. Catal. 112 (1997) 13.
gen species due to the extremely large interface area. Therefore, it [46] D. Martin, D. Duprez, J. Phys. Chem. 100 (1996) 9429.