In order to watch processes on the molecular level, it is necessary to develop a technique that combines the time and space domains on a microscopic scale. In the case of surface physics, the standard tool for structure determination is low-energy electron diffraction (LEED). It combines the advantages of surface sensitivity due to the inherently small mean free path of low energy electrons in solid materials and its relatively simple and robust experimental design. The trade-off is the quite elaborate data analysis, which is required to extract structural and vibrational information [1]. Currently, the development of ultra short electron pulses is a field of vital interest [2]. For time-resolved LEED experiments, an electron gun has to be developed that produces ultra short and coherent electron pulses with an expected time resolution of less than five picoseconds. The basic principle of the short pulse electron gun is the production of a photoelectron cloud by a femtosecond laser system and its subsequent acceleration and direction towards the sample. The temporal broadening of the pulse with respect to the excitation pulse is mainly governed by the energy spread of the initial photoelectrons and the field strength in the first acceleration stage. The first acceleration stage is therefore only 0.1 mm long, and a gold cathode is used for the photoemission of monochromatic electrons. The size of the electron gun is small and can be moved very closely to the sample (4 mm). This design has the advantage that no lenses have to be implemented in the gun. We are planning pump and probe experiments in order to directly study the time evolution of the molecular motion. As a first experiment the frustrated translation vibration of C60 adsorbed on Ag(100) will be investigated. This experiment will serve as proof of principle for time-resolved LEED. C60 is rather heavy, and thus the vibrations of the molecules are slow and might be resolved with a picosecond pump and probe experiment. C60 forms an Ag(100)-c(6x4)-2C60 overlayer structure. The energy of the frustrated translation mode perpendicular to the surface is about 4 meV, which translates into a root mean square (rms) displacement of about 0.2 Å at 100 K and 0.3 Å at 300 K. These changes in rms are sufficient to vary the LEED intensities significantly. The frustrated translation mode will be coherently excited by a laser pump pulse, and the relaxation is probed by a coherent low-energy electron beam whose delay with the pump pulse is scanned. The excited C60 molecules execute stronger vibrational motion about the equilibrium position than in thermal equilibrium that causes the diffracted LEED intensity to vary or even to change the LEED IV curves. The interpretation of the time-dependent experimental LEED data requires an improved approach in the LEED calculations. By varying the delay time between the pumping laser pulse and the probing electron pulse, it is possible to follow the dynamics of the vibrationally excited molecules. More specifically, we want to determine how the once excited low-frequency mode of C60 is damped (and dephases) with time (on the picosecond scale). These data may provide useful information about the coupling of the C60 molecule with the underlying substrate and neighboring C60 molecules. Later we are planning to apply this technique for the study of the migration of charge carrier from the photocatalyst to the redox catalyst (project 2). The project on time-resolved LEED will be performed in close collaboration with Dr. T. Greber, Physics Department, University of Zurich, Switzerland. A transnational research proposal within the CERC-3 framework has already been submitted.

[1] H. Over, M. Gierer, H. Bludau, and G. Ertl, Phys. Rev. B 52 (1995) 16812.

[2] J.R. Thompson, P.M. Weber, J. Estrup, SPIE Proceedings 2521 (1995) 113.