High Temperature Structure Formation and Surface Diffusion of Silver on Silicon Surfaces

The present work deals with surface diffusion and structure formation, mainly for the case of Silver on Silicon surfaces. Various techniques are combined to investigate flat and vicinal surfaces oriented in the Si(001) and Si(111) directions as well as intermediate orientations. Low energy electron microscopy (LEEM) and photoemission electron microscopy (PEEM) were used to study the growth dynamics and diffusion involved in structure formation in ultrahigh vacuum (UHV) conditions. The investigated structure formation deals with single- and multi-crystalline Ag islands and self-organized Ag wires on various Si surfaces. Ag was deposited at elevated temperatures, while the investigations were mainly carried out in-situ. The structure of the grown Ag islands and wires was investigated with either small area low energy electron diffraction (µ-LEED), spot profile analyzing-low energy electron diffraction (SPA-LEED), or scanning electron microscopy (SEM). The SEM investigations were the only investigations, where the sample was extracted from the UHV and were carried out to improve the statistical significance of the data. Ag(001) and Ag(111) islands were grown at elevated temperatures of up to 700°C. With increasing growth temperature, the shape of the islands transformed from hexagonally shaped to triangular. The relative rotation to the substrate was investigated and compared to a modified coincidence site lattice approach (CSL) which agreed very well with the experimental results. Practically all of the significant rotation angles could be explained by the CSL model. Surface diffusion fields were investigated during the decay of islands in the process of desorption. These islands are surrounded by one or several concentric adsorbate induced reconstruction zones. A simple continuum diffusion model is presented, explaining the decay mechanism. The model contains a previously presented model as a special case and was developed in collaboration with J. Krug and I. Lohmar at the University of Cologne. Several diffusion parameters are extracted from the model and are in excellent agreement with values in literature. The decay of Ag islands on vicinal Si substrates no longer yields concentric circular zones, but the zones become anisotropic, and the model can no longer be applied due to the no lack of rotational symmetry. A model from the literature is used to explain the data in combination with SPA-LEED results on multi-step formation and faceting and numerical simulations. Only a combination of all these techniques is capable of a thorough and all-embracing explanation of surface diffusion. The results are compared to the system of Indium on vicinal Si(001) surfaces. Here, in contrast to Ag on vicinal Si(001), no anisotropy is found and the drawn picture can also explain the surprising diffusion isotropy. Among the islands that were used for the diffusion investigations, on Si(001), wires form. The growth of these single crystalline wires was investigated and an attempt has been taken to clear an ongoing discussion about the cause of the wire formation. The single crystalline wires were grown on flat and vicinal Si(001) surfaces. All wires align to one of the two principal directions of the substrate. Their growth is thermally activated and surprisingly independent of the substrate vicinality. The wires align with the step edges as the sample vicinality and with it the step density is increased. The results cannot lead to a clear decision on which of the discussed phenomena diffusion anisotropy or strain are the cause for self-organized wire formation on vicinal Si(001) surfaces. We can, however, come to the conclusion, that the wire alignment is much more closely linked to the diffusion anisotropy than the formation itself. We therefore state, that the diffusion anisotropy is a possible cause for the wire alignment, restricting the wire growth to one of the possible two directions with increasing diffusion anisotropy.