Modification of metallic surfaces via ion implantation for plasma torches and other applications

Author(s)

Ivan Jankov

• Document type: Doctoral Thesis
• Press: São Paulo.
• Institution: Universidade de São Paulo (USP). Instituto de Física (IF/SBI)
• Defense date: 2004-12-13

Examining board members

Iuda Dawid Goldman Vel Lejbman; Jean Albert Bodinaud; Adalberto Fazzio; Maximo Siu Li; Hercilio Rodolfo Rechenberg

• Advisor: Roberto Nunes Szente

Abstract

In this work, the controlled implantation of the different elements in the metallic surfaces (particularly copper) was performed, in order to study the changes that the implantation causes on the surface structure, aiming to improve the behaviour of those metals for their use as cathodes in plasma torches or other applications, such as: catalysis, microelectronics, oxidation and corrosion of metals and others. Thin polycrystalline copper films were implanted with ions (energy 20-50keV; doses of order of \'10 POT. 15\' ions/\'cm POT. 2\') of alkaline metals (\'LI\', \'NA\', \'K\', \'RB\', and \'CS\') as well as of \'O\' and \'CL\'. Different surface analyses were performed in order to determine the changes on copper due to the ion implantation, in terms of: surface composition (Auger Electron Spectroscopy, X-ray Photoelectron Spectroscopy), topographic and surface potential structure (Kelvin Probe Force Microscopy), bulk composition (X-ray Fluorescence, Rutherford Backscattering Spectroscopy e Energy Dispersive X-ray Spectroscopy), oxygen concentration (Elastic non-Rutherford Backscattering Spectroscopy) and crystalline structure (X-ray Diffraction). Theoretical models were used to understand better the structural changes which occur on the metallic surface during the ion implantation process (Stopping and Range of Ions in Matter e Tight-Binding Linear Muffin-Tin Orbital Atomic Sphere Approximation (Coherent Potential Approximation)). The surface composition analyses of implanted and non-implanted copper substrates showed similar concentration of \'C\', \'N\', \'CL\' and \'S\' and that the only change in surface concentration, due to the ion implantation process, was the introduction of the desired ions in the surface of copper films. It was observed that even small ion doses can cause relatively large decrease of work function (2-30%) in relation to the pure copper value; the implantation of O and CL caused an increase in work function of 300m V and 900mV, respectively. The detected concentration of implanted alkali metal ions was relatively high (taking into consideration relatively small implantation doses, of the order of \'10 POT. 15\' ions/\'cm POT. 2\'), decreasing fast towards bulk of the samples; this distribution were probably caused by a migration process of implanted ions towards the surface. The EBS and KPFM results indicate that not always the largest dose produces the largest implanted ion concentration in the surface layer. It was also observed that only a part of the total ion dose is effectively implanted, due to the sputtering during the ion implantation process. Implantation of different alkali ions influences the sample oxidation process in a different way. The principal influence in the increase of oxidation is the topographic structure of the samples; however, the presence of the implanted ions on the samples surface seems to influence the initial stages of the oxidation, increasing or decreasing the oxygen adsorption. Of the cases of the O and CL implantation, the concentration of these ions seems to increase with depth, which indicates that there were no migration process involved. This is probably due to the fact that these ions create bonds with the substrate elements. The changes in the work function for two distinct cases of metal (AG and CS) deposition on CU (111) was studied with the computational programme TB-LMTO-ASA (CPA), using, in the simulations, different values for the Wigner-Seitz radius for Empty Spheres (\'WS IND. ES\'); the results on the work function changes during the deposition were approximately 20% lower in comparison with the experimental data from the literature. The effect of \'WS IND. ES\', which is, in principle, a computation artefact, on the work function value of the studied systems is normally interpreted as the non-reliability of the models based on the ASA concepts. However, the results obtained during this work indicate that there is a relation between the \'WS IND. ES\' and the surface roughness; therefore, physical meaning of the empty spheres can be understood as a \"measure\" of surface roughness. (AU)