Study of Cooper Surface Oxidation by Grazing Angle X-Ray Excitation

H. J. SÁNCHEZ; C. A. PÉREZ

Goteborg

Conferencia; TXRF 2009; 2009

This work reports measurements of copper surface oxidation by XRF analysis under grazing angle excitation. The mathematical model for analyzing data is based on the usual equation for XRF intensity, corrected by the stratified model for the layer excitation.The final expressions are fitted to the data by using the Simplex algorithm. Sample preparations and measurements were carried out at the Laboratorio Nacional de Luz Sincrotron, in the Total-Reflection Station of the XRF beamline. Three samples of silicon wafers with surface layers of copper were made by using the controlled-evaporation technique under vacuum. The final thickness layers were 190 Å, 400 Å, and 800 Å. Samples were conserved in portable vacuum chambers until they were ready for measurements in order to prevent unwanted oxidation by the atmosphere. For each sample, several measurements were carried out by performing an angular scanning in the region of the critical angle for total reflection. They covered angular ranges between 0 and 20 mrad with a measuring time of 10 sec/point for approximately 60 points. Different measurements were taken immediately after evaporation, after 30 min, after 1 hour, after 4 hour, and after 30 min in oven at 65°. The results show small variations among the different spectra measured for each sample. The most significant variations are observed for the first measurement (after evaporation) and the last one (after the heating in oven). The mathematical model works correctly for a two-layer scheme but shows inconsistencies for more complex schemes. This indicates that the stratified model is not appropriate for continuous media due to the nature of the theoretical assumptions in which the stratified model is based. Our results shows also that the atmosphere oxidation is produced in the first instants of atmosphere exposition after the sputtering, and is increased only if the sample is heated, Finally, for surface layers thicker that 400 Å the layer is too thick and it behaves like a substrate.   Acknowledgements Research partially performed at LNLS - National Synchrotron Light Laboratory, Brazil.