The surface segregation in binary alloys is often studied by Auger electron spectroscopy. However, the information obtained in the Auger spectra is representative of a region thicker than one monolayer. The Auger signal must be decomposed into the contributions of the different atomic layers. On the other hand, it is now well known that matrix correction factors must he used to relate the intensity of the Auger signal to the composition of the sample.As these two parameters are seldom put together in the treatment of Auger data, we tried to combine them for the study of the surface segregation in silver-palladium alloys. We calculated the matrix correction factors for AgPd alloys in the whole range of concentrations. The attenuation length corrections and the backscattering corrections were calculated on the basis of different models (Seah and Dench, Penn, Tanuma-Powell-Peno, IchimuraShimizu-Langeron, Reuter-Ichimura, etc.). The matrix corrections were compared to the Hall and Morabito approach and with previously published results. The matrix corrections were then introduced into a layer-by-layer model of interpretation of the Auger signal. While the surface composition calculated without matrix correction reveals silver enrichment relative to the bulk of the samples, the use of matrix corrections lead to an enrichment in silver that is strongly attenuated.
INTRODUCTIONAuger electron spectroscopy is a widely used surface analysis technique. Part of the success of this method is that the Auger current for a given transition can be easily related to the atomic concentration of the emitting element. Auger electron spectroscopy is considered as a surface analysis technique because the attenuation length of the Auger electrons is small (typically a few monolayers, depending on the kinetic energy of the electrons). A lot of papers in the field of surface science were published 20 years ago, where the atomic concentration was linearly connected to the intensity of the Auger current detected by assuming that the composition calculated using the well-known formula (1) was the surface compositionwhere li is the Auger current of element i, X i is the atomic fraction of element i and Si is the sensitivity coefficient of the element i. However, two main restrictions to this simple formalism quickly appear :(1) The intensity of the Auger signal does not vary linearly with the atomic concentration of the emitting element and so some correction factors due to matrix effects must be intr~duced.'-~ (2) The Auger signal does not come only from the first atomic layer (the 'surface' of the sample) but also from the second, third, etc. layer, depending on the attenuation length of the electrons,4g5 and eqn. (1) assumes that the composition is uniform within the sampling depth of the electrons. Many studies have been published on each of these two restrictions, separately. In this paper, we examine the influence of the matrix effects combined with a layer-by-layer model in the calculation for the surface composition of a silver-pallad...