In this project we aim to predict the morphology evolution of a porous solid due to chemical reactions between a solid and a penetrating gas phase. The oxidation of soot by diluted oxygen and NO2, passing through a porous soot layer, will serve as an example of practical relevance. This basic phenomenon during the regeneration of soot filters for cleaning the exhaust of diesel engines and is so far not sufficiently understood. It is well known that the overall oxidation rate during soot filter regeneration is a strong function of time, even under isothermal conditions. By combining experimental investigations and detailed numerical simulations we will try to show from first principles if (and to what extent) the change in the overall reaction rate primarily originates from morphological changes of the soot.

Within the project a special reactor will be constructed that allows to oxidize a soot layer stepwise under isothermal conditions. From analysis of exhaust gas, we retrieve the instantaneous overall reaction rate. After subsequent steps of partial soot oxidation, the overall BET surface will be measured without dismounting the sample. Thus, we can simultaneously obtain effective reaction rates and surface areas. Detailed information of the porous soot structure has recently become assessable by Focused Ion Beam - Scanning Electron Microscope (FIB-SEM) images. This allows to reconstruct the porous medium in microscopic scale with high resolution and provides initial conditions for numerical simulations.

A detailed 3D Lattice-Boltzmann (LB) model will be developed to simulate the soot morphology change during soot oxidation. While the reaction of multicomponent inside pores will be considered homogeneous and simulated using the discrete form of the Boltzmann kinetic equation, the heterogeneous soot morphology change during soot oxidation will be modeled as new boundary condition. As LB method can handle complex geometries efficiently, it is suited to model solid-gas phase transition. We can compare the calculated effective oxidation rate to experimental values, assuming a constant surface reaction rate at the evolving soot surface. Furthermore, FIB-SEM images obtained from soot samples after partial oxidation can be compared to simulations. Once a detailed model is available, this will serve as a basis for the development of an extended application-oriented model. For deriving anisotropic macroscopic transport parameters asymptotic homogenization will be used. The derived macroscopic model comprises the classical mass and energy balances as well as balance equations for characteristic structural properties such as BET surface.

Fig.: From left to right a sample of backscattered images, its three-dimensional reconstruction of FIB-SEM images (Hasegawa et al. 2018), as well as images pores detection and numerically reconstruction (Muntean et al. 2003).