Sintering and Particle Dynamics in Supported Metal Catalysts

Sintering represents a significant route of deactivation of supported metal catalysts. Hence a fundamental understanding of the phenomenon is important when designing catalysts more resistant to deactivation by sintering. The aim of this work is to elucidate the mechanism of sintering under simulated industrial conditions. The main focus is on studying the assumptions leading to the derivation of a model for sintering of industrial steam-reforming catalysts. The validity in extrapolation of the model to the low-pressure simulated steam-reforming environment used in the in situ transmission electron microscope (in situ TEM) experiments in this study is also investigated, thus establishing a link over a wide range of pressure. In this thesis, the sintering of oxide supported metal particles were studied using in situ TEM and scanning electron microscopy. For the in situ TEM studies, a nickel-based steam-reforming catalyst was studied under conditions relevant for the steam-reforming reaction. Three sets of experiments were carried out. One set of experiments focused on the effect of atmosphere. The catalyst samples were treated in atmospheres with and without water vapor present at different temperatures. Particle diameters were measured at different instances and particle size distributions determined as a function of time. The second set of experiments focused on the effect of time. The catalyst samples were sintered in the in situ microscope for a certain period of time and the particle size distributions and mean particle diameter determined before and after treatment. The idea of these first two experiments is to create a reference to later online experiments and to compare with predictions from a model derived from experiments carried out at in a reactor at ambient pressure. In the third set of experiments the migration of the metal particles was monitored online. Samples were treated in different atmospheres and temperatures and particle migration was recorded as a function of particle size. An analysis of the three sets of experiments showed agreement with the trends found in reactor experiments under industrial conditions for the change in the mean particle diameter. Deviation from classical models were in that the smallest particles did not migrate the longest distances as classical theory predicts. Instead particles around 8-10nm are most mobile and coalesce with other particles during their migration. A population of small particles was present on the support surface even after 5 hours of sintering. This population did, however, decrease in number over the sintering period. The presence of small particles suggests that a large number of the particles do not themselves migrate and participate in the sintering only when absorbed by other particles. These small particles are somehow anchored to the support. The frequency with which metal particles were observed to sinter can account for the changes observed in the particle size distributions. Based on the observations, a model is proposed that includes one activation energy related to the release of particles from their anchoring sites and a second activation energy accounting for the migration of metal particles over the support. A series of sintering experiments on flat model catalysts were carried out and studied in the scanning electron microscope. These studies were aimed at supporting the observations and the model derived from the in situ TEM experiments. As the experiments with flat model systems produce particle size distributions with better statistics, the shape of these can be derived more convincingly. Further the flat model substrates decreases the complexity of the industrial catalyst studied in the in situ TEM by removing a spatial dimension. These experiments also showed that the smallest particles may be anchored to the support while larger particles show signs of migration.