CeO2-based catalysts with engineered morphologies for soot oxidation to enhance soot-catalyst contact
© Miceli et al.; licensee Springer. 2014
Received: 15 March 2014
Accepted: 3 May 2014
Published: 23 May 2014
Skip to main content
© Miceli et al.; licensee Springer. 2014
Received: 15 March 2014
Accepted: 3 May 2014
Published: 23 May 2014
As morphology plays a relevant role in solid/solid catalysis, where the number of contact points is a critical feature in this kind of reaction, three different ceria morphologies have been investigated in this work as soot oxidation catalysts: ceria nanofibers, which can become organized as a catalytic network inside diesel particulate filter channels and thus trap soot particles at several contact points but have a very low specific surface area (4 m2/g); solution combustion synthesis ceria, which has an uncontrolled morphology but a specific surface area of 31 m2/g; and three-dimensional self-assembled (SA) ceria stars, which have both high specific surface area (105 m2/g) and a high availability of contact points. A high microporous volume of 0.03 cm3/g and a finer crystallite size compared to the other morphologies suggested that self-assembled stars could improve their redox cycling capability and their soot oxidation properties. In this comparison, self-assembled stars have shown the best tendency towards soot oxidation, and the temperature of non-catalytic soot oxidation has dropped from 614°C to 403°C in tight and to 552°C in loose contact conditions, respectively. As far as the loose contact results are concerned, this condition being the most realistic and hence the most significant, self-assembled stars have exhibited the lowest T10% onset temperature of this trio (even after ageing), thus proving their higher intrinsic activity. Furthermore, the three-dimensional shape of self-assembled stars may involve more of the soot cake layer than the solution combustion synthesis or nanofibers of ceria and thus enhance the total number of contact points. The results obtained through this work have encouraged our efforts to understand soot oxidation and to transpose these results to real diesel particulate filters.
An increasing share of the automobile market has been gained by diesel engines on-board passenger cars, over the last two decades, as they are more fuel economic than gasoline vehicles. However, diesel engines entail a more challenging reduction of pollutant emissions. Particulate matter (PM) is a complex aerosol composed of nanosized carbonaceous particles (called soot) on which soluble hydrocarbons, sulphates and metals adhere through complex filtration and oxidation phenomena. These particulates have diameters that range from a few nanometers to hundreds of nanometers and beyond. This means serious problems in terms of human respiratory diseases and environmental issues[2, 3].
Driven by compulsory legislation, the reduction in PM emission is currently a technological challenge from both the engine and the catalyst points of view. In the past, many efforts were devoted to the development of catalytic diesel particulate filters (DPF), in order to achieve a cheaper and more effective solution than fuel-borne catalysts (FBC), which had proved to produce more pulmonary intrusion particles. The DPF is a ceramic filter with alternate-plugged channels, in which the flue gases enter the open channels at the inlet, cross the porous ceramic wall of the channel, where soot particles are retained, and finally exit the filter from the neighbouring channels. The soot particles deposit in the pores of the ceramic walls and progressively form a soot layer on top of the wall, which is called cake. The latter generates a drop in pressure across the filter, which becomes unsustainable for the engine; therefore, the cake periodically needs to be burned off, in order for the filter to regenerate. Regeneration is currently achieved through the post-injection of fuel from the engine[6, 7], which causes a relevant fuel penalty for modern engines.
Currently, the combination of a trap with an oxidative catalyst is commonly adopted. This involves the deposition of noble metals on carriers with a high surface area, such as zeolites or γ-alumina, or those with redox properties, like ceria (CeO2) in pure or doped form[8, 9]. It is common knowledge that rare earth metals, like ceria, are less expensive than classic noble metals and leave a lower transformation carbon footprint, which makes these materials more sustainable. Replacing noble metals with rare earth ones, or lowering the content of the former, would be a remarkable result in economic and environmental terms.
In this work, ceria-based catalysts have been investigated as active carriers to improve soot oxidation. In particular, three different morphologies have been proposed. Having redox properties, the Ce4+/Ce3+ cycle can store oxygen in lean conditions and then provide it in rich conditions to promote oxidation at the soot-catalyst interface. This ability depends to a great extent on the intrinsic activity of the catalyst and on the properties of the reaction surfaces. Redox-capable catalysts are developed by increasing oxygen mobility with vacancies, by exposing particularly active crystalline planes, or by enhancing the oxygen storage capacity (OSC) of catalysts in order to rapidly restore the oxidation state of the active metal following soot oxidation.
Another important fact is that soot oxidation is a solid-solid catalysis, and it is necessary to take into account the importance of the soot/catalyst contact conditions, which can basically be of two kinds: tight contact and loose contact. It has been demonstrated, in a real DPF, that loose contact takes place and, in these conditions, the activity of the catalyst is not the only important feature: an engineered morphology has to be designed to achieve better results.
On the basis of this evidence, new morphologies were investigated in previous works[9, 11], and in particular, a fibrous structure of the ceria-based carrier was proposed with the aim of maximizing contact between the catalyst and the soot particles. Despite their low specific surface area (SSA), these fibers in fact have a filamentous structure which enhances the number of soot-fiber contact points and, in some cases, show better performances than foamy or higher SSA nanopowders, obtained with the solution combustion synthesis (SCS) technique[9, 11]. This proves that specific surface area is not the only important factor in solid-solid catalysis and that tailored morphologies can be achieved even with low specific areas.
This concept is extremely important, given the application field of these catalysts, which have to be layered on the surface of the DPF channels. A morphology that could intercept a higher fraction of the soot cake, with a better penetration of the catalytic layer inside the soot cake, would improve the regeneration phase.
As a result, a comparison of the three different ceria morphologies, namely the nanofibers, self-assembled stars and the nanopowders obtained by SCS, has been performed in the following study.
Three different synthesis techniques were adopted in this study:
▪ The CeO2 nanofibers were synthesized by means of the precipitation/ripening method[9, 15]: starting from a 1 M aqueous solution of cerium (III) nitrate hexahydrate precursor (Sigma-Aldrich, St. Louis, MO, USA, 99%), the fibers were synthesized using a rotary evaporator and varying the NaOH/citric acid molar ratio. The residence time and conditions inside the evaporator led to different morphologies. A clear fibrous structure was obtained for a ratio of 0.8 at a constant temperature of 60°C for 6 h. One-hour drying at 110°C and calcination for 5 h in air at 600°C were performed. These processes did not cause the fibrous structure to collapse after the thermal treatment.
▪ The CeO2 self-assembled stars were prepared by mixing 0.2 M of cerium (III) chloride heptahydrate, 0.01 M of CTAB (both from Sigma-Aldrich) aqueous solutions and 80 mmol of solid urea. A hydrothermal treatment at 120°C for 12 to 24 h led to a precipitate, which was centrifuged, rinsed, filtered, dried at 60°C for 24 h and finally calcined at 600°C for 4 h. The residence time inside the reactor in hydrothermal conditions affects the size and shape of these systems, as will be shown later on.
▪ The SCS was also used for the ceria catalyst preparation in order to compare the foamy catalyst obtained with this technique with the above-mentioned alternative morphologies. In the SCS technique, a homogeneous aqueous solution of metal nitrates and urea is placed in an oven set at a constant temperature of 650°C. The solution quickly begins to boil and froth, and ignition then takes place. The exothermic reaction, due to urea combustion, provides the heat necessary for the endothermic transformation of nitrates into the desired oxide. The whole process is over in a few minutes, and the result is a foam that crumbles easily. In this case, the size and shape of the CeO2 structures were not tunable as in the other two cases, although a foamy structure and a moderate SSA were easily reached.
All the aforementioned CeO2 morphologies were characterized by means of X-ray diffraction (PW1710 Philips diffractometer, Amsterdam, The Netherlands, equipped with a Cu Kα radiation monochromator to check that the cerium oxide crystalline structure had been achieved and to estimate the average crystallite size via the Debye-Scherrer technique. A field emission scanning electron microscope (FESEM, Leo 50/50 VP Gemini column) was used to analyze the morphology of the CeO2 structures and to correlate it to its activity towards soot oxidation. A BET analysis (Micromeritics ASAP 2010 analyzer, Norcross, GA, USA) was conducted to evaluate the specific surface area of the catalysts and to perform a porosimetry analysis of the prepared catalysts. An ageing thermal treatment was performed for all three catalysts at 600°C for 5 h in order to have a better understanding of their reliability and performances under stressed conditions, namely when exposed to high temperatures for a certain period.
Temperature-programmed combustion tests (TPC) were run to establish the oxidation activity of the catalysts, both in tight contact, in order to assess their intrinsic activity, and in loose contact, in order to evaluate their behaviour in more realistic conditions.
The tight contact was prepared by ball milling the catalysts and soot for 15 min at 240 rpm; this creates a intimate contact between the two phases and is helpful to discriminate the activity of the different morphologies. Only two 1 cm diameter agate balls were used instead of standard four to prevent breaking of the delicate micrometric structures during milling, as it had been noticed during the scanning electron microscopy (SEM) analysis, that severe mechanical stress could wreck such engineered morphologies. Loose contact was obtained by gently mixing the catalyst and soot with a spatula by hand for a minute. This technique, which is quick and easy but with reproducible results, simulates the real contact conditions for soot and a catalyst inside a DPF since the cake deposits on the filter channels without any external compaction force.
TPC runs were made with a PID-regulated tubular oven, into which a U-tube quartz reactor with the catalytic bed had been inserted. The temperature rose till 750°C at 5°C/minute, while 100 ml/min of 10% O2 (obtained by dilution of air with N2) was made to flow through a fixed bed of 5 mg of Printex-U synthetic soot (Degussa, Essen, Germany), 45 mg of catalyst and 200 mg of silica, according to the standard operating procedure described in, with the only difference being an increased amount of silica in the catalytic bed, to achieve a better temperature homogeneity. The CO/CO2 concentration in the outlet gas was measured via NDIR analyzers (by ABB). Each test was repeated three times to ensure reproducibility of the obtained results. The peak temperature, Tp, in the TPC plot of the outlet CO2 concentration was taken as an index of the catalytic activity. The onset (T10%) combustion temperature, defined as the temperature at which 10% of the initial soot is converted, was also considered in order to better discriminate between the intrinsic catalytic activities of the prepared catalysts. The half conversion temperature (T50%) was also taken into account. The onset temperature is important to rank the catalysts, according to the catalytic reaction; other phenomena (such as mass transfer or diffusion limitations) may in fact influence the performances of catalysts at higher conversion stages.
The modification to the inert silica content in the bed composition led to slightly different oxidation temperatures for the materials tested in, especially as far as the onset temperature was concerned. In fact, the higher dilution heat capacity of the here adopted silica bed was relevant, especially at the reaction onset, i.e. when the heat released by soot oxidation was not able to self-sustain the reaction, and therefore had most impact on the reaction rate itself. However, the catalyst ranking in loose and tight contact conditions obtained in has here been confirmed, and it has been shown that the SA stars offer a major improvement over the other ceria morphologies developed in this work.
Crystallite sizes of the CeO 2 -based catalysts obtained by means of XRD analysis
Crystallite size [nm]
Aged SA stars
Specific surface area (SSA) of the CeO 2 -based catalysts obtained by means of BET analysis
Aged 5 h at 600°C
Recalling that soot oxidation depends on both the number of soot-catalyst contact points and on the availability of adsorbed oxygen at this contact point, it can be seen that the SA stars seem to have both features: they have the ability to maximize the contact between the soot and catalyst phase, as the fibers do, but they also have a much higher SSA, which entails a better activity at low temperatures (which depends on the oxygen coverage).
Soot combustion activity results, under close and loose contact conditions, of the SA stars, SCS nanopowders and nanofibers
Aged SA stars
Conversely, in loose contact conditions, the morphology plays a more relevant role: the nanofibers, despite the almost null SSA, exhibit an almost equivalent activity to that of the SCS powders. This behavior, which was also obtained in, is here confirmed; this is further evidence that the BET alone cannot explain the activity of the soot oxidation catalytic reaction and that the contact between soot and the catalyst should be promoted. As far as the SA stars are concerned, their performance is much better than that of the other two catalysts, especially at low temperatures: in fact, the high porosity of the catalyst provides more adsorbed oxygen to the contact points between the soot and the catalyst, which is likely to be in a sufficient amount to fully exploit this oxygen availability. As far as the aged catalyst tests are concerned, it is worth mentioning that the lower SSA penalizes T10%, but T50% still remains within the range of the other fresh catalysts.
A low temperature peak in the CO2 concentration (around 140°C) is evident in all the star-related curves. This peak is not connected to soot combustion. A tailored set of consecutive temperature-programmed desorption (TPD) runs was run to prove that the CO2 produced at low temperature is due to the desorption of CO2 from the inner nanoporosity of the self-assembled stars: in the first TPD, a fresh catalyst, previously exposed to air, was heated to 200°C in N2, and the CO2 desorption peak was recorded. The same catalyst was then cooled down in N2 and heated again in N2 to 200°C: in this case, no CO2 was noticed. The CO2 peak recorded at 140°C was therefore clearly attributable to the desorption of the CO2 formerly present in the air and was greater for the SA stars as they are characterized by the highest SSA.
Three different types of ceria catalysts have been synthetized and compared for soot oxidation using TPC runs: SCS, with an uncontrolled morphology, and two engineered design ones, nanofibers and self-assembled stars.
The purpose was to create a catalytic layer in DPF that would be able to entrap soot particles in several active points and enhance oxidation for a fast and cheap regeneration of the filter. Several TPC runs have been conducted, in both tight and loose contact mode, to investigate the contact points of all the three catalysts.
In previous works[9, 11], it was proved that engineered catalyst morphologies give better results towards soot oxidation than unstructured ones, and it was therefore decided to continue developing this idea and try and remove any drawbacks. A new morphology, with a star-like shape of micrometric size, was developed. It was deduced, from the TPC runs results, that SA stars give better results than the other catalysts, especially in loose conditions. In spite of their micrometric size, SA stars are nanostructured and have finer crystallite size: this entails a much higher BET area, greater availability of oxygen vacancies, more efficient redox cycles and, therefore, a higher oxidative capability.
Further investigations are needed to improve both the morphology and its effective deposition inside the DPF in order to improve the cake oxidation within the filter itself.
The authors declare that no one else has to be acknowledged.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.