At reaction temperatures (typically a few 100°C), a catalyst surface becomes dynamic and adsorbates, reaction intermediates as well as catalyst particles (clusters) can move across the support. During my Ph.D., I used helium spin-echo (HeSE) spectroscopy to investigate the steady state diffusion of small molecular adsorbates on pico- to nanosecond time scales. In our group in Munich, we use scanning tunneling microscopy (STM) to monitor the cluster geometry and stability during a reaction and track mobile species on the surface. While the time resolution of STM is limited compared to HeSE, it brings the significant advantage that we can thus monitor non-equilibrium processes, co-existing species and rare events. Furthermore, using the recently developed FastSTM module, we can measure two to three orders of magnitude faster than with conventional STM instruments, reaching video-rate scan frequencies. This allows us to observe surface dynamic processes in situ with atomic resolution, even at elevated temperatures.
The surface dynamical processes that we investigate include
Our current focus lies on Pt clusters on the magnetite (001) surface. Here, we are particularly interested in the role of surface defects on catalyst reactivity and in transport limitations when studying simple test reactions.
To complement the real space information from STM with information on the chemical state of metal clutsers and support materials, we use x-ray photoelectron spectroscopy (XPS). Specifically, we use ambient pressure XPS (APXPS) to observe chemical changes in situ during a reaction. While the size and composition of catalysts can be controlled to a high degree using chemical synthesis methods, a non-zero size distribution and some ligands at the particle surface typically remain. Instead, we prepare samples with our laser evaporation source here in Munich, which produces truly monodisperse, ligand-free metal clusters. This way, we can compare the catalytic behavior of particles just one atom apart in size even with an averaging technique such as XPS. Since size-selected cluster samples intrinsically have a low coverage (to avoid coalescence into dimers, trimers, etc.), we need a highly sensitive technique. Therefore, we travel to synchrotrons (typically to the Advanced Light Source in Berkeley and recently also to the Diamond Light Source in England) to achieve the required resolution and signal intensity.
See my Google Scholar profile