A Low Energy Electron Diffraction Study of Surface Structures

From the low-energy electron diffraction (LEED) analysis of the (110) surface of palladium we have determined an oscillatory relaxation with respect to the bulk interlayer spacing in the two topmost layers. We used R-factors for quantitative comparison of theoretical I-V curves with the experimental data. The oscillatory relaxation of the surface Pd(110) d12 = (1.33 ± 0.05) Å, d23 = (1.44 ± 0.06) Å, d34 at its bulk value 1.38 Å. To get a minimum Pendry R-factor we had to consider atoms vibrations in the first layer in the direction parallel to the surface. Introduction To understand a lot of physical and chemical properties of solid substances the knowledge of solid structure is very important. It is important for explanation phenomena in heterogeneous catalysis, microelectronics, etc. It exists a lot of methods how to determine the surface structure. One of the most popular is LEED. In LEED we investigate elastically scattered electrons. Typical energy range in LEED is 30 500 eV. Due to the very strong interaction with matter, electrons have a mean free path in the same order as an interatomic space is. By measuring a diffraction pattern for certain energy of incident electrons a periodicity of surface lattice may be determined by kinematical approach. One of the main assumption for kinematical approach is that the incident electron scatters in the crystal just once. This assumption is often not fulfilled, but the kinematical theory describes the diffraction pattern well. A current in the direction of a diffraction maxima can be measured and the results of this measuring are the so-called I-V curves, which means that the spot intensities are measured as a function of the electron energy. For I-V curves interpretation the dynamical approach is used, where the multiple scattering is taken into account. With this method we can determine an exact positions of atoms, i.e. relaxations, adsorption positions, etc. Dynamical approach has a statistical error 0.02 Å in the determination of atoms positions in the direction perpendicular to the surface. The goal of this work is to learn how to determine a surface structure by dynamical theory and obtain information about our experimental facility on the relatively easy Pd(110) system. This is because of the determination of surface structures using I-V curves is a new method in our department. In this paper we present an overview of structural studies carried out on the Pd(110) surface. We studied a clean Pd(110)-(1x1) surface and we extract quantitative information regarding the magnitude of relaxations of the interlayer spacings in the first few layers. Experimental procedures Experiments are performed in an ultrahigh vacuum chamber (base pressure ≤ 10 mbar) equipped with three-grid LEED optics, quadrupole mass spectrometer for thermal desorption spectroscopy (TDS), sputter gun and gas exposure provisions. The sample was mounted on XYZ manipulator which permits a rotation, but does not permit a tilt. The LEED device was managed by PC equipped with PCI card with a 16-bits analog-digital/digitalanalog converter. The diffraction patterns and LEED intensities were measured with a video camera, which can measure an illumination to 2.10 lx. The Pd(110) crystal was cleaned by cycles of Ar bombardment at room temperature and annealing at 880 K as long as no further decrease of the background in the diffraction pattern was observed. Temperature was checked by the thermocouple. Whereas the sample could not be placed into the precise position for normal electron beam incidence, 10 experiments for 5 different degrees α around the expected normal impact were performed. For each degree we measured diffraction patterns for two different voltages on Wehnelt cylinder (-10 V and -5 V). Normal incidence was determined by comparing symmetry-equivalent curves. Condition for precise normal incidence is RP < 0.1 ( [1]). Measurement for the best settings for normal incidence was chosen. WDS'05 Proceedings of Contributed Papers, Part III, 580–583, 2005. ISBN 80-86732-59-2 © MATFYZPRESS