(712g) Evaluation of Soots for Diesel Particulate Filter Development Studies

Diesel soot consists of agglomerates with diameters of up to several hundred nanometers and a fine structure of spherical primary particles. Unlike the well understood formation of NO_x in internal combustion engines, formation of soot is still more ambiguous because the process is more complicated and difficult to examine. For example, the characterization of diesel particulates from a particular engine at different operation conditions may vary considerably due to variations in combustion parameters and diesel fuels. Laboratory studies often utilize diffusion flames in burners, because the soot formation processes in burner diffusion flames are thought to be fundamentally similar to those in diesel engines with sequential nucleation, growth and agglomeration steps taking place.

Until recently the impact of soot nanostructure on reactivity has not drawn much attention. Vander Wal et al. found by means of transmission electron microscopy (TEM) that the structural differences of soot/carbon at a nanoscale level (i.e. dimension and orientation of the graphitic layers) affect its oxidation reactivity. Other recent studies showed that the intrinsic oxidation rate of pyrolysis soot might vary from different fuels (such as benzene, ethanol, and acetylene) resulting in differences in soot nanostructure. Moreover, it has been speculated that peculiar nanostructures, such as amorphous and shell/core formations, are present in the soot formation process. Amorphous nanostructures are related to soot particles and have been identified mainly in less developed soot particles and condensed species. The study of soot nanostructure is important for both correlating reactivity and simulating non-catalytic/catalytic soot oxidation in exhaust systems.

Comparisons of the kinetic parameters of non-catalytic soot oxidation with those of catalytic soot oxidation have been made by many researchers to evaluate the effectiveness of soot oxidation catalysts, and the obtained kinetics information can also reveal the mechanism of soot oxidation reactions. However, the reported activation energy for non-catalytic soot oxidation ranges from 92 to 211kJ/mol since the employed experimental techniques and conditions vary from one group to another. Yezerets et al. compared the oxidative reactivity of diesel soot and carbon black, and the activation energies were found to be 92 and 117 kJ/mol, respectively. In another study of soot samples from different engine conditions, the activation energies for the soot with ash contents of 14 and 6.5% were found to be 126 and 146 kJ/mol, respectively. A recent study by Kim et al. showed that the activation energies for diesel soot and flame soot are 110 and 162 kJ/mol, respectively. Hence, it seems the kinetics parameters for soot oxidation cannot be simply compared from one study to another due to the differences in the nature of the soot and the experimental conditions of the reactors (i.e. the mass and heat transfer limitations).

In soot oxidation studies, both well defined model soots and a reliable means to simulate realistic contact conditions with catalysts are crucial. In catalytic soot oxidation studies, the “contact condition” refers to the nature of the contact between the soot and catalyst, with the contact range varying from “tight” to “loose”. The realistic contact condition obtained by in-situ collection on a DPF in an exhaust system is considered as “loose” contact. Replicating this condition in laboratory studies can be challenging. In lab-scale studies, thermogravimetric analysis (TGA) has usually been used to characterize soot oxidation behavior, permitting relatively straightforward examination of kinetics during heating. When the catalysts under study are in a powder form, loose contact can be simulated by gently mixing the soot and catalyst powders with a spatula or gently shaking the sample container with catalyst/soot mixture. However, these methods are not appropriate for study of catalysts incorporated onto complex support material shapes (e.g. porous cordierite/SiC filters).

In the present study, it was anticipated that producing flame soot with a Tiki torch oil lamp and capturing the soot directly onto a catalyst/support structure would yield a contact condition simulating realistic conditions.  The properties of this flame soot were examined by means of XRD and TEM for structure analysis, BET for surface area analysis, and TGA for reactivity and kinetics analysis. The flame soot is thought to be more uniform and controllable than similar samples of collected diesel particulates. For validation purposes, catalytic oxidation of Tiki soot using the simulated contact condition was compared with the diesel particulates collected from a real diesel engine exhaust system.