Quantitative 3D Structure Studies of Supported Catalysts at the Nanoscale

Heterogeneous catalysis, where a solid catalyst facilitates a gas- or liquid-phase reaction, plays an irreplaceable role in chemical and fuel production as well as environmental control. The catalytic performance (activity, selectivity and stability) depends strongly on the details of the structure of the active sites and the morphology of the support. Understanding the structure-property relationship in catalysts and exploiting it for the development of highly selective and efficient catalysts require detailed and high-resolution characterizations of their structure, both in 2D and in 3D. In this thesis, advanced electron microscopy techniques with a focus on electron tomography and image analysis are applied to gain a deeper understanding of the distribution of the active particles and the morphology of the support for different catalysts to correlate the structural aspects with differences in catalytic performance. In particular, electron tomography is not only used as a qualitative tool to visualize nanostructures, but also pushed to provide quantitative 3D structural information at the nanoscale. As one aspect, looking into the role of the support material in catalysis, the morphology and diffusion properties of a disordered mesoporous carbon material have been quantified based on electron tomographic reconstructions. The quantitative analysis strongly depends on the fidelity of the reconstruction and the segmentation, which are affected by pore size variations, the missing wedge during tomography acquisition and the reconstruction approach. The pore structure is described in terms of geometric and topological parameters based on both SIRT and DART reconstructions. The effect of the tomography acquisition, reconstruction and segmentation approach on the pore structure description is evaluated quantitatively in order to estimate the reliability of the description. Pore variations introduced in the reconstruction and segmentation process are mainly present in a few voxel wide boundary region of the pores, slightly altering the size and connectivity of pore structure. However, the calculated diffusion properties are very sensitive to variations in the pore volume and necking between pores and are therefore noticeably influenced by the anisotropic effects of the missing wedge, which estimated in simulations. In order to address the role of the local structure of the active centers on the catalytic reaction as a second aspect for catalytic performance, electron tomography is applied to uncover differences in the distribution of Pd nanoparticles supported on mesoporous carbon (CMK-3) prepared by three synthetic methods: incipient wetness impregnation, wet impregnation and sol immobilization methods. The resultant spatial distributions of the Pd nanoparticles in the three Pd/CMK-3 catalysts are significantly different from each other and correlate with the activity and selectivity in liquid phase furfural hydrogenation. Pd nanoparticles located inside pores promote the formation of 2-methyl furan, while the selectivity for furfuryl alcohol and tetrahydrofurfuryl alcohol is suppressed correspondingly. Moreover, recycling tests indicated that leaching depends on the synthesis process and plays an important role with respect to catalytic stability. Finally, looking into metal oxides as a different type of support and especially the effect of crystallographic faceting on the catalytic performance, electron tomography in combination with HRTEM and HAADF-STEM imaging continues the 3D characterization of supported catalyst. The exposed surfaces of well faceted ceria particles were accurately determined both in 2D and in 3D. The cube shaped CeO2 nanoparticles exhibit {100} facets as main surfaces with truncated edges giving rise to some {110} facets. In contrast, the rod shaped CeO2 nanoparticles mainly exhibit two types of exposed surfaces corresponding to the {111} and {100} planes. The analysis of CO oxidation by Pt nanoparticles deposited on the cube and rod shaped CeO2 particles revealed a correlation between the oxidation activity at low temperatures and the exposed surfaces of the CeO2 nanoparticles, which is in agreement with theoretical predictions that oxygen vacancy generation on the oxide surfaces is the rate determining step in the catalytic reaction