Catalysis Today 114 (2006) 40–47

www.elsevier.com/locate/cattod

Promotional effect of rare earths and transition metals in the

combustion of diesel soot over CeO2 and CeO2–ZrO2
Eleonora Aneggi *, Carla de Leitenburg, Giuliano Dolcetti, Alessandro Trovarelli 1
Dipartimento di Scienze e Tecnologie Chimiche, via Cotonificio 108, Università di Udine, 33100 Udine, Italy

Available online 20 March 2006

Abstract
The soot combustion behavior and the textural and structural characteristics of CeO2 and a series of ceria-modified materials have been studied.
It is shown that ceria doped with transition metals (Zr and Fe) and rare earth elements (La, Pr, Sm, Tb) results in more active catalysts with
enhanced textural properties. ZrO2 enhances the thermal stability and the oxygen storage capacity of pure ceria, resulting in better performance in
soot oxidation. Remarkably, cerium–zirconium solid solution doped with rare earth does not achieve lower temperature of combustion, providing
performances comparable to CeO2 and CeO2–ZrO2. Cerium doped with Fe2O3 presents the better results as far as fresh samples are taken into
account, but suffers from a net loss of activity after calcination. TGA experiments under N2 atmosphere have confirmed the key role of oxygen
storage capacity in soot oxidation.
# 2006 Elsevier B.V. All rights reserved.

Keywords: Ceria; Ceria–zirconia; Fe2O3; Soot oxidation; Diesel; Oxygen storage; Combustion

1. Introduction Several catalyst formulations for soot oxidation have been

studied in the last years, and the most promising are those based
Among the several techniques that have been developed for on Cu/K/M/(Cl) (where M is either V, Mo or Nb) [2–4], that
reducing particulate emissions from diesel engines, filtering have been investigated in detail due to their high activity at low
followed by catalytic oxidation is one of the more promising. temperature.
The basic idea of catalytic approach consists in the use of a Regarding the intimate mechanism involved in the oxidation
catalyst to achieve the onset of regeneration at a temperature of carbon, several authors pointed out the importance of redox
comparable to that of exhaust gases. Unfortunately, this method properties of the catalyst. That is, the effectiveness of the
is affected by several drawbacks [1]: (i) catalytic filter catalyst can be related to its ability to deliver oxygen from the
regeneration is made difficult by the very variable conditions lattice to the gas phase (or better to the solid carbon reactant) in
of reaction; (ii) the process is quite slow because of the poor a wide temperature range [1]. Recently, it has been reported that
soot–catalyst contact; actually, the solid particles are immobile the use of supports based on cerium oxide confers interesting
when deposited and due to their large size can not penetrate into properties to soot combustion catalysts due to high availability
catalyst’s micropores or mesopores; (iii) the temperature in the of surface oxygen and high surface reducibility [5–9]. The
exhaust gases may vary in a wide range (indicatively from 473 success of oxygen storage systems based on ceria is due to their
to 873 K), depending upon engine load. Consequently, a useful ability to change oxidation state during operation (i.e. CeO2 to
catalyst has to operate efficiently at low temperatures and also CeO2
x) maintaining structural integrity, thus allowing oxygen
to be thermally stable. uptake and release to occur easily. State of the art material for
oxygen storage in TWC is constituted by ceria and ceria–
zirconia mixed oxides doped with small amounts of rare earth
elements, added to enhance the redox features of ceria [10].
* Corresponding author. Tel.: +39 043 255 8865; fax: +39 043 255 8803. Homogeneity of solid solutions, structural features and
E-mail address: [email protected] (E. Aneggi).
1
A member of CONCORDE, the EU-funded Coordination Action ‘‘Co-
composition are key parameters in successful redox catalyst
Ordination of Nanostructured Catalytic Oxides Research & Development in design. Since it is being reported in the literature that a strong
Europe’’. relationship exist between oxygen storage/redox capacity and
0920-5861/$ – see front matter # 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2006.02.008
 E. Aneggi et al. / Catalysis Today 114 (2006) 40–47 41

soot combustion activity [11], it is of interest to investigate here these values was estimated by checking their agreement against
if the knowledge that has been accumulated on redox systems the values of the lattice constant, assumed to comply with the
for TWC can help in developing more active soot combustion Vegard’s law [16].
catalysts.
In this study a series of ceria-based catalysts modified with 2.2. Catalytic activity
rare earth (Sm, Pr, Tb, La) and transition metal elements (Zr,
Fe) have been prepared and characterized from a structural and The catalytic activity for the combustion of soot was
textural point of view. The effect of doping in the activity of quantified by the peak-top temperature (Tm) during temperature
soot combustion is also described by putting particular attention programmed oxidation (TPO) of catalyst–soot mixtures [1,17–
in correlating soot oxidation activity with oxygen storage and 20]. Each catalyst was accurately mixed with soot in a mortar
structural properties of the materials. for 10 min in order to achieve a tight contact [21]. The kind of
contact between catalyst and soot is extremely important
2. Experimental procedures [1,22]: the tight contact conditions are poorly representative of
the real working conditions experienced by the catalyst
2.1. Catalyst preparation and characterization deposited in a catalytic trap, but they allow a rapid screen of
catalysts in reproducible experimental conditions.
Three classes of materials were prepared and characterized: A soot/catalyst weight ratio of 1:20 was adopted. During the
ceria–zirconia solid solutions, ceria–zirconia doped with rare TPO measurements 25 mg of mixture were heated at a constant
earth elements (La, Pr, Sm, Tb) and ceria and ceria–zirconia rate (10 K/min) in a quartz reactor, while the gas flow (N2 with
doped with Fe. They were prepared by coprecipitation starting 6% of O2) was kept fixed at 400 ml/min. The catalyst
from nitrates. Precipitates were dried at 393 K and calcined at temperature was checked by a chromel-alumel thermocouple,
773 K for 2 h (fresh samples). Aged samples were prepared located on the catalyst bed. The outlet composition was
following calcination in air at 1023 K for 12 h. Chemical measured by IR and paramagnetic gas analyzers (Magnos 106
composition of all samples investigated is reported in Table 1. and Uras14, ABB), by recording the percentages of CO, CO2
Textural characteristics of all fresh and aged samples were and O2 at the output of the reactor.
measured according to the B.E.T. method by nitrogen adsorption A series of tests were carried out for catalyst/soot mixtures,
at 77 K, using a Sorptomatic 1900 instrument (Carlo Erba). in order to verify the reproducibility of results; the peak
Structural features of the catalysts were characterized by X- temperatures show differences always lower than 5 K.
ray diffraction (XRD). XRD patterns were recorded on a Philips Soot oxidation activity was also tested by running TGA
X’Pert diffractometer operated at 40 kV and 40 mA using experiments (Q500, TA Instruments) either in the presence or in
nickel-filtered Cu Ka radiation. Spectra were collected using a the absence of oxygen (N2 atmosphere). In order to achieve the
step size of 0.028 and a counting time of 40 s per angular stoichiometric ratio between the amount of soot and the oxygen
abscissa in the range 20–1458. The Philips X’Pert HighScore that the catalyst can donate, we used a soot/catalyst weight ratio of
software was used for phase identification. The mean crystal- 1:115. The total flow of N2 in the furnace was 100 ml/min.
line size was estimated from the full width at the half maximum Samples were pre-treated for 1 h at 423 K to eliminate water
(FWHM) of the X-ray diffraction peak using the Scherrer absorbed, then they were heated at a constant rate (10 K/min) up
equation [12] with a correction for instrument line broadening. to 1073 K; the weight loss of the sample, after subtraction of the
Rietveld refinement [13] of XRD pattern was performed by blank, can be considered as an indication of activity of soot
means of GSAS-EXPGUI program [14,15]. The accuracy of oxidation by oxygen from the catalyst.

Table 1
Characteristics of samples used in this study
Sample Composition B.E.T. surface area (m2/g) Crystallite size (nm)a Tm (K)
Fresh Aged Fresh Aged Fresh Aged
CZ100 CeO2 57 22 7 11.2 662 700
CZ75 Ce0.75Zr0.25O2 73 44 6.3 10 660 673
CZ44 Ce0.44Zr0.56O2 96 54 4.7 13 675 674
CZ28 Ce0.28Zr0.72O2 100 58 6.2 13 695 711
CZ0 ZrO2 79 44 11.8 15.9 798 810
CF95 Ce0.95Fe0.05O1.975 92 10 5.4 44.4 639 749
CF85 Ce0.85Fe0.15O1.925 77 22 7.5 31.1 662 722
CZF95 Ce0.47Zr0.48Fe0.05O1.975 106 25 3.9 8.8 672 715
CZF85 Ce0.45Zr0.40Fe0.15O1.925 132 22 3.5 8.9 684 716
CZLa Ce0.48Zr0.50La0.02O1.99 93 54 4.0 5.6 687 685
CZPr Ce0.48Zr0.50Pr0.02O1.99 99 55 4.1 6.1 683 678
CZSm Ce0.48Zr0.50Sm0.02O1.99 92 50 4.3 6.4 684 679
CZTb Ce0.48Zr0.50Tb0.02O1.99 96 50 4.2 6.3 687 691
a
Calculated with Scherrer formula from X-ray diffraction patterns.
42 E. Aneggi et al. / Catalysis Today 114 (2006) 40–47

In order to evaluate the OSC of samples we carried out TGA This is in accordance with literature, where generally the limit
experiments in Ar/H2 (5%) flow (total flow 100 ml/min). Each for formation of a cubic (either cubic or tetragonal t00 ) solid
sample was treated in N2 atmosphere for 1 h at 553 K. Then, it solution is found at ca. 50 mol% CeO2 [25]. Thus, all our cerias
was heated at a constant rate (10 K/min) till 673 K and kept at crystallize in a cubic fluorite structure of Fm3m symmetry with
this temperature for 15 min, to eliminate the absorbed water. almost identical cell parameters. A similar situation is found
Finally Ar/H2 mixture was introduced while keeping the with fresh and aged CZ75; both crystallize in a cubic structure
temperature at 673 K for 30 min. The observed weight loss is with a smaller cell parameter than ceria, in agreement with the
due to oxygen removal by H2 to form water, and it can be introduction of the smaller Zr4+ in the lattice. The value of
associated to total oxygen storage capacity at that temperature a = 5.370 fits well with data reported previously [25–27]. No
[23]. indication of phase splitting is detected after aging, which is in
line with what found on ceria-rich samples after similar
3. Results and discussion treatments [28]. This however cannot exclude that thermal
treatment might induce some phase assessment at a nano-scale
3.1. Textural and structural characterization level, with small compositional variations, which have been
indicated as responsible for enhancement of oxygen exchange
Textural and structural characterization of all samples is in samples obtained using similar procedures [29]. In the range
reported in Tables 1 and 2. Fresh materials have surface area in of composition 25–45 mol% CeO2 good agreement between
the range 60–130 m2/g. The effect of aging strongly affects the refinement and a tetragonal structural model confirms that
pure ceria and Fe-modified catalysts, while sintering in ceria– all samples belong to the t0 tetragonal phase with space group
zirconia is less important in accordance with its better thermal P42/nmc. No peak splitting that would indicate the presence of
resistance [24]. The presence of rare earth elements within two phases could be detected, and therefore, the diffraction
ceria–zirconia lattice does not modify stability of ceria– patterns demonstrate the formation of a single solid solution-
zirconia against surface area loss. like ceria–zirconia phase. Again, this cannot exclude the
The structural features of all samples were analyzed by presence of different arrangements of oxygen sublattice or the
XRD. As shown in Table 2, XRD measurements suggest that for presence of a multi-phase system at a nanoscale level, not
binary ceria–zirconia samples with cerium content greater than detected by XRD. After calcinations, only a slight modification
50 mol% the formation of a cubic fluorite lattice is favored. of the structural parameter is observed which does not affect

Table 2
Crystallographic parameters of modified ceria samples as obtained from Rietveld refinement and Vegard law
Sample Phase Cell parameters From Vegard law
a = b = c (Å) a0 = b0 (Å) c (Å)
Fresh samples
CZ100 Cubic 5.411 5.411
CZ75 Cubic 5.370 5.341
CZ44 Tetragonal t0 5.251 5.346 5.253
CZ28 Tetragonal t0 5.177 5.261 5.207
CZ0 Monoclinic tetragonal Not refined 5.127
CF95 Cubic 5.404 5.356
CF85 Cubic 5.396 5.243
CZF95 Cubic 5.273 5.225
CZF85 Cubic 5.295 5.143
CZLa Cubic 5.300 5.285
CZPr Cubic 5.297 5.282
CZSm Cubic 5.290 5.277
CZTb Cubic 5.286 5.273
Aged samples
CZ100 Cubic 5.410 5.411
CZ75 Cubic 5.370 5.341
CZ44 Tetragonal t0 5.250 5.341 5.253
CZ28 Tetragonal t0 5.175 5.266 5.207
CZ0 Monoclinic tetragonal Not refined 5.127
CF95 Cubic 5.409 5.356
CF85 Cubic 5.408 5.243
CZF95 Cubic 5.275 5.225
CZF85 Cubic 5.292 5.143
CZLa Cubic 5.297 5.285
CZPr Cubic 5.297 5.282
CZSm Cubic 5.289 5.277
CZTb Cubic 5.285 5.273
 E. Aneggi et al. / Catalysis Today 114 (2006) 40–47 43

overall phase distribution. In summary, XRD characterization 3.2. Catalytic activity of ceria–zirconia based materials
of binary ceria–zirconia solid solutions confirms the homo-
geneity of samples (within the limits outlined above) with the Soot combustion tests were carried out either under TPO
identification of mainly cubic and tetragonal t0 phases, as conditions in a micro flow reactor, or with a TG apparatus.
generally reported for similar composition [30]. Fig. 1A shows the results of soot combustion studies carried out
In fresh samples based on CeO2 doped with Fe, XRD under air in a TG apparatus over ceria and ceria–zirconia. All
features allow to detect only the CeO2 cubic phase Fm3m the catalysts examined are active in promoting soot combustion
while Fe2O3 or other iron oxide phases are not visible. XRD in the range of temperature from ca. 600 to 800 K. As a measure
peaks are broad and the values of crystallite size obtained of activity we used the temperature at which 50% of weight loss
according to Scherrer equation are about 54 Å for sample is observed (T50, corresponding to the temperature at which
CF95 and 75 Å for sample CF85. After aging at 1023 K, peaks 50% of soot is converted under tight contact conditions).
become more intense, clear and well defined for all Fe- Similar results were also obtained by carrying out soot
containing samples. Aging is accompanied by a large drop in oxidation in a conventional flow reactor. In this case the profile
surface area and by a consequent increase of crystallite size of CO2 evolved (Fig. 1B) can give us a quantitative estimation
(444 and 311 Å, respectively, for CF95 and CF85). After of relative activity by measuring Tm value (temperature of
calcination, the XRD profile of CF samples shows peaks maximum evolution of CO2), and although absolute values are
assigned to rhombohedral Fe2O3 (hematite) with R-3c slightly different from those obtained from TG experiments, the
symmetry. The lack of peak due to iron oxide in fresh CF activity–composition curve has similar features. The values of
samples could indicate the formation of solid solution Tm are reported in Fig. 2 against composition. For fresh CZ
between Ce and Fe. However, a comparison between the samples the order of reactivity is dependent on Ce content, with
values of cell parameter with those expected if all the iron the lowest Tm value obtained with cerium-rich compositions. In
contained were dissolved in the lattice, Table 2, indicates that general, when the percentage of cerium oxide decreases, in
only a small percentage of iron is dissolved in ceria. This value fresh samples, the temperature of oxidation increases. This
slightly increases with the amount of Fe but still include a result is in agreement with a previous investigations [33], where
minority of the Fe present in the material. After aging, the the same trend is reported. The lower activity of ZrO2 (CZ0) is
increase of cell parameter indicates a segregation of the iron also in accordance with the study of Perrichon and co-workers
eventually dissolved in the lattice with formation of weak [34], in which activities in soot oxidation for several supports
signal due to crystalline Fe2O3. It is known that lower valence were investigated.
ions such as Fe3+ are extremely difficult to dissolve into the After aging there is an increase of the oxidation temperature
ceria lattice, especially when treating at high temperature which is higher with CZ100 and CZ0; as a consequence the
[31]. Mutual dissolution of Ce and Fe into Fe2O3 and CeO2 has curve assumes an inverse volcano-type profile with a minimum
been reported to exist in Fe-rich Ce/Fe mixed oxides prepared in the middle composition range, characteristic of phenomena
by coprecipitation [32]. driven by the redox features of the catalyst [23]. The main
In order to verify the formation of solid solution we difference between pure cerium and cerium–zirconium solid
recorded and analyzed XRD features of CeO2–ZrO2 solid solution is therefore related to the stability after calcination.
solution samples doped with rare earths and transition metals. Pure ceria sample decreases its activity, while for solid
XRD features of all fresh and calcined samples except CZF85 solutions containing zirconium Tm is less influenced by
show only one phase, namely a cubic phases (Fm3m) of calcination. CeO2 as support in diesel soot oxidation, has
CeO2–ZrO2 solid solution with 48 mol% of cerium. The been the subject of several studies [5–7,9,34,35]. Different
accuracy of the lattice constants retrieved from Rietveld authors agree in explaining the activity of cerium oxide with its
refinement were estimated comparing these values with oxygen storage capacity. One of the most important roles of
expectations from application of Vegard law assuming that all CeO2 in catalytic redox reactions is to provide surface sites [36]
MOx dopant (2 mol% for CZM, 5 mol% for CZF95 and and to act as an oxygen storage/transport medium by its redox
15 mol% for CZF85) were dissolved into solid solution. cycle between Ce4+ and Ce3+. Two factors can therefore explain
Table 2 compares lattice parameters retrieved from Rietveld the behavior observed: one is the availability of surface Ce4+
refinement and from Vegard law. The results show a good sites and the other the ability of the material to donate its
agreement for rare earth containing materials indicating oxygen for soot oxidation. For fresh material, the amount of
formation of ternary solid solution. The situation is different surface Ce sites is higher for CZ100 and CZ75 and
for samples containing Fe2O3. Values retrieved by Rietveld correspondingly these samples are the most active: due to
refinement are not in agreement with those computed from the high surface area, abstraction of oxygen involves mainly
Vegard’s law: the adding of a cation (Fe3+) with ionic radius surface sites, with little or no participation of the bulk in the
smaller than Ce4+ and Zr4+, should produce a decrease in cell reaction. This can be shown in Fig. 3A where an almost linear
volume in the case of a solid solution. Conversely, we observe relationship is observed by plotting the number of surface
values higher than expected indicating that Fe2O3 is deposited oxygen ions available for exchange [37] against Tm in fresh
on the surface. As in the case of pure ceria we cannot exclude samples. On the contrary, after aging, the minimum Tm is found
that a small fraction of Fe is dissolved within ceria–zirconia with CZ44 and CZ75 and no linear relationship is observed
framework. between surface oxygen ions and Tm. This indicates that, in
44 E. Aneggi et al. / Catalysis Today 114 (2006) 40–47

Fig. 1. Soot oxidation under stationary conditions carried out in a TG apparatus (A) and in a conventional flow reactor by monitoring evolution of CO2 (B): (––)
CZ100, (– – –) CZ75, (  ) CZ44, (–  –) CZ28 and (–   –) CZ0.

addition to available surface sites, other factors should oxygen is directly involved in the reaction path and therefore
contribute to the overall oxidation behavior. structural features of the material play an important role.
The role of oxygen storage in soot oxidation can be well Addition of rare earths in CZ system does not modify
evidenced by conducting a series of TG experiments in the significantly the performance. Table 1 summarizes the TPO
absence of air, under inert gas flow; a sort of temperature results of all samples, showing the temperature of maximum
programmed reduction under nitrogen and using soot as CO2 evolution during soot oxidation. Activity of rare earth
reductant. We used a soot/catalyst ratio of 1/115 in order to have modified materials is close to that of ceria–zirconia and it is not
a sufficient amount of oxygen available in the support for total influenced by calcination. This agrees with the fact that rare
oxidation of soot. The TG profiles (not reported for brevity) earth dopants are generally introduced in binary ceria–zirconia
show a continuous weight loss starting at ca. 500 K with formulation to increase thermal stability. Direct promotion of
evolution of oxygen and CO2. In order to exclude contributions OSC by rare earth doping is not straightforward and depends on
from O2 desorption/removal from the sample we carried out several variables like preparation, textural and structural
blank experiments on all catalysts by running a temperature properties [38]. This could explain lack of enhanced activity
programmed analysis in the absence of soot. The results are in fresh samples.
reported in Fig. 2B and they are compared to the behavior Let us briefly discuss the dependence of activity against
observed with the catalyst only treated under H2. It is shown surface area. It has been recently reported that B.E.T. surface
that for both high and low surface area samples the curve has a area follows a trend that is the inverse of T50 (or Tm) [35]; that is
minimum in the middle composition range. A behavior which is the surface area affects positively soot oxidation activity. This
similar to that observed in the presence of hydrogen, thus statement is true, however by plotting the difference in Tm of
pointing out that participation of lattice oxygen in soot oxidation fresh and aged samples against surface area (Fig. 4) it is shown
may contribute to overall activity, especially when surface area is that for sample having surface area greater than ca. 40 m2/g
low or in the presence of large soot particles, when access of gas (regardless of initial surface area values) the difference in Tm
phase oxygen is limited also by geometrical factors. Fig. 3B (DTm) is limited and generally less than ca. 20 K. On the
shows that there is no linear relationship in this case between the contrary, if aging results in surface area lower than 30 m2/g
number of surface oxygen ions available for exchange and the important effects in DTm are observed. Therefore we can
amount of oxygen extracted as CO2, pointing out that the lattice conclude that there is no linear relationship between the two
 E. Aneggi et al. / Catalysis Today 114 (2006) 40–47 45

Fig. 2. (A) Correlation between Tm from TPO experiments and the amount of
ceria in solid solution for both fresh (squares) and aged (circles) samples. (B) Fig. 3. Correlation between surface available oxygen and (A) Tm from TPO
Weight loss against catalyst composition measured by TGA experiments on experiments and (B) weight loss measured by TGA experiments for both fresh
soot/catalyst in N2 flow (filled symbols, left axis) and on catalyst only (squares) and aged (circles) systems.
compositions in Ar/H2 flow (open symbols, right axis) for both fresh (squares)
and aged (circles) materials.

variables: a strong dependence is observed when ending up at the XRD profile: the peaks of hematite and cubic ceria were
low surface areas, while the activity is less influenced by even narrower indicating a better crystallization of those phases
large drops in surface area, when final surface area values are (Fe2O3 phase was not visible in the fresh material). Moreover
higher than ca. 40 m2/g. the unit cell parameters calculated from the reflections of cubic
ceria moved from 5.4022 to 5.4100 Å, which is similar to that
3.3. The effect of doping with Fe of the pure ceria sample. This indicates that amorphous Fe2O3,
which was initially present on ceria, tends to crystallize and that
The results of activity of fresh and aged CF and CZF samples the small fraction of Fe that was initially dissolved in the ceria
are reported in Table 1. The fresh sample Ce0.95Fe0.05O1.975 lattice progressively segregates after calcination by forming
shows the best Tm among the samples investigated (23 K lower free, crystalline Fe2O3 deposited on ceria. As shown in Fig. 5,
than that of pure ceria), but a large drop of activity is observed the activity in soot combustion is strongly influenced by
after calcination at 1023 K. A similar situation characterizes the calcination temperature with Tm increasing by more than 100 K
other Fe-doped samples. This observation led us to investigate in the range of temperature investigated. Initially, a large drop
more in detail the behavior of CF95 after aging of the sample at in surface area (from 92 to 47 m2/g) induce only a slight
increasing temperature in the range 873–973 K. As shown in increase of Tm (639 K versus 646 K), on the contrary when
Table 3, the increase in calcination temperature induces a surface area dropped below 22 m2/g (sample calcined at 948 K)
gradual loss of surface area and a corresponding increase of Tm raised up to 685 K. These results are consistent with our
crystallite size. In addition it resulted in important changes in previous indications, in which it has been show that key
46 E. Aneggi et al. / Catalysis Today 114 (2006) 40–47

Fig. 4. Difference of Tm (DTm) in aged and fresh samples from TPO experi-
ments against surface area of the catalysts after aging. The line ideally separates Fig. 5. Correlation between surface area and Tm returned by TPO experiments
the regions in which the surface area influences (left) or not (right) the activity in for CF95 samples calcined at different temperatures (773–1023 K).
soot oxidation.

parameter is not represented by the difference in surface area key step in the development of Fe-modified cerias for
between fresh and calcined samples, but by the absolute value application in the oxidation of soot.
of surface area after calcination.
The higher activity of Fe-modified catalysts (especially CF 4. Conclusion
samples) can be correlated to a high degree of Fe–Ce
interaction which occurs through (i) the formation of cubic Ceria-based materials represent an interesting class of
ceria-like solid solution where Fe cations are dissolved within catalysts for the combustion of soot with oxygen. Our
ceria structure. In this case the interaction takes place through investigation pointed out that their activity is mainly related
the sharing of oxygen anion defined by the Fe–O–Ce bonds to two factors: oxygen storage/redox capacity and surface area.
formed in the Fe-doped CeO2 lattice [32]. (ii) The presence of An inverse dependence of Tm on surface area was observed,
amorphous Fe2O3 in close contact with ceria: apart from with the absolute value of surface area being a critical
catalytic activity deriving from CeFeO entities, the reactivity of parameter: for surface area larger than ca. 40 m2/g the variation
exposed iron site within and in close contact with solid solution in Tm is not relevant, while for lower surface areas a large
can have a significant role in the catalytic behavior observed. increase of Tm is observed. The importance of lattice (surface
The lack of these interactions due to the lower amount of ceria and bulk) oxygen in ceria has been highlighted. A simple redox
can explain the different behavior of CZFe samples. route mechanism for soot oxidation, which utilizes oxygen
Decrease of catalytic activity after aging in samples with activated from the support in a typical reduction/oxidation path
Fe2O3 has a threefold cause: segregation of Fe2O3 from solid (Mars Van Krevelen type) in which the catalyst undergoes a
solution, sintering of sample and agglomeration of free Fe2O3 partial reduction, can be put forward. Oxygen storage is
(peaks become visible in XRD profile). These factors will therefore important because it provides an alternative route for
weaken Ce–O–Fe interactions leading to a less active catalyst. the oxidation of big soot particles in contact with ceria.
Stabilization of these samples against sintering is therefore a Addition of zirconium is important to stabilize performance
after calcination and to provide more reactive oxygen ions from
Table 3 the surface and the bulk, in a manner similar to that observed in
Characteristics of CF95 after aging at different temperatures three-way catalysts.
Sample Cell parameters Tm SA Crystallite
Fe-doped materials are the most active, but they experienced
(Å)a (K) (m2/g) size (nm) strong deactivation after calcinations. It is therefore necessary
to stabilize these materials against sintering.
CF95 fresh 5.4022 (0.0004) 639 92 5.4
CF95 calc. 873 K 5.4032 (0.0002) 646 47 9.3
CF95 calc. 923 K 5.4046 (0.0001) 671 41 11.7 Acknowledgments
CF95 calc. 948 K 5.4083 (0.0001) 685 22 18.0
CF95 calc. 973 K 5.4080 (0.0001) 711 17 22.7 The authors thank financial support from MIUR (progetti
CF95 calc. 1023 K 5.4100 (0.0001) 749 10 44.4 PRIN). We are also grateful to Grace Davison (USA) for
a
Obtained from Rietveld refinement. providing samples used in this study.
 E. Aneggi et al. / Catalysis Today 114 (2006) 40–47 47

References [21] J.P.A. Neeft, M. Makkee, J.A. Moulijn, Chem. Eng. J. 64 (1996) 292.
[22] J.P.A. Neeft, O.P. van Pruissen, M. Makkee, J.A. Moulijn, Appl. Catal. B
[1] B.A.A.L. van Setten, M. Makkee, J.A. Moulijn, Catal. Rev.-Sci. Eng. 43 12 (1997) 21.
(2001) 489. [23] E. Aneggi, M. Boaro, C. de Leitenburg, G. Dolcetti, A. Trovarelli, J.
[2] G. Saracco, C. Badini, N. Russo, V. Specchia, Appl. Catal. B 21 (1999) Alloys Compd. 408–412 (2006) 1096.
[24] J. Kaspar, P. Fornasiero, in: A. Trovarelli (Ed.), Catalysis by Ceria and
233.
[3] P. Ciambelli, V. Palma, P. Russo, S. Vaccaro, J. Mol. Catal. A 204–205 Related Materials, Imperial College Press, London, 2002, pp. 217–241.
(2003) 673. [25] P. Fornasiero, G. Balducci, R. Di Monte, J. Kašpar, V. Sergo, G. Gubitosa,
[4] B.A.A.L. van Setten, J.M. Schouten, M. Makkee, J.A. Moulijin, Appl. A. Ferrero, M. Graziani, J. Catal. 164 (1996) 173.
[26] A. Trovarelli, M. Boaro, E. Rocchini, C. de Leiteburg, G. Dolcetti, J.
Catal. B 28 (2000) 253.
[5] E.E. Miro’, F. Ravelli, M.A. Ulla, L.M. Cornaglia, C.A. Querini, Catal. Alloys Compd. 323 (2001) 584.
Today 53 (1999) 631. [27] V. Sanchez Escribano, E. Fernandez Lopez, M. Panizza, C. Resini, J.M.
Gallardo Amores, G. Busca, Solid State Sci. 5 (2003) 1369.
[6] M.L. Pisarello, V. Milt, M.A. Peralta, C.A. Querini, E.E. Miro’, Catal.
[28] H. Vidal, J. Kaspar, M. Pijolat, G. Colon, S. Bernal, A. Cordon, V.
Today 75 (2002) 465.
[7] V.G. Milt, C.A. Querini, E.E. Miro’, M.A. Ulla, J. Catal. 220 (2003) 424. Perrichon, F. Fally, Appl. Catal. B: Environ. 30 (2001) 75.
[8] Grzona, C. Carrascull, D. Lick, M. Ponzi, E. Ponzi, React. Kinet. Catal. [29] E. Mamontov, R. Brezny, M. Koranne, T. Egami, J. Phys. Chem. B 107
(2003) 13007.
Lett. 75 (2002) 63.
[9] J. Setiabudi, G. Chen, M. Mul, J.A. Makkee, Moulijn, Appl. Catal. B 51 [30] J. Kašpar, P. Fornasiero, M. Graziani, Catal. Today 50 (1999) 285.
(2004) 9. [31] Z. Tianshu, P. Hing, H. Huang, J. Kilner, J. Mater. Proc. Technol. 113
[10] Trovarelli, Comments Inorg. Chem. 20 (1999) 263. (2001) 463.
[32] F.J. Perez-Alonso, M. Lopez Granados, M. Ojeda, P. terreros, S. Rojas, T.
[11] P. Ciambelli, V. Palma, P. Russo, S. Vaccaro, Catal. Today 60 (2000) 43.
[12] R. Jenkins, R. Snyder, Introduction to X-ray Powder Diffractometry, Herranza, J.L.G. Fierro, M. Gracia, J.R. Gancedo, Chem. Mater. 17 (2005)
Wiley, New York, 1996, p. 90. 2329.
[13] R.A. Young, The Rietveld Method, IUCr Oxford University Press, New [33] J.F. Lamonier, N. Sergent, J. Matta, A. Aboukaı̈s, J. Therm. Anal. Calorim.
66 (2001) 645.
York, 1993.
[14] A.C. Larson, R.B. Von Dreele, General Structure Analysis System (GSAS), [34] J. van Doorn, J. Varloud, P. Meriaudeau, V. Perrichon, Appl. Catal. B 1
Los Alamos National Laboratory Report LAUR 86-748, 2000. (1992) 117.
[35] A. Bueno-López, K. Krishna, M. Makkee, J.A. Moulijn, J. Catal. 230
[15] B.H. Toby, J. Appl. Cryst. 34 (2001) 210.
[16] D.J. Kim, J. Am. Ceram. Soc. 72 (1989) 1415. (2005) 237.
[17] J.C. Summers, S.V. Houtte, D. Psaras, Appl. Catal. B 10 (1996) 139. [36] A. Trovarelli, Catal. Rev.-Sci. Eng. 38 (1996) 439.
[18] R.J. Farrauto, R.M. Heck, Catal. Today 51 (1999) 351. [37] Y. Madier, C. Descorme, A.M. LeGovic, D. Duprez, J. Phys. Chem. B 103
(1999) 10999.
[19] M. Shelef, R.W. McCabe, Catal. Today 62 (2000) 35.
[20] G.A. Stratakis, A.M. Stamatelos, Combust. Flame 132 (2003) 157. [38] R. Di Monte, J. Kaspar, J. Mater. Chem. 15 (2005) 633.