No abstract

In order to relate an Auger signal to the atomic concentration in the surface of a solid, it is well known that corrections have to be made for matrix effects. Also, corrections may be needed to compensate for the effects of artefacts caused by properties of the sample other than the concentration of different elements. Large artefacts can be caused by the surface topography and subsurface composition variations often present in 'real' samples. This paper reports a new correction scheme that enables the quantitative analysis of samples with a structure for which no prior knowledge exists. The scheme uses the ratio of the Auger peak height to a background count rate above the energy of all significant Auger features, combined with a linear correction involving the atomic number of the substrate and the Auger backscattering factor. Inhomogeneous samples with angle of incidence variations from 0 O to 45" and an arbitrary atomic number can be analysed with an accuracy of -5 at.% using this method. Limitations exist for thin unsupported films.

Experimental results are reported for the energy-analysed angular distributions of the secondary electron cascade with energies from 200 eV up to 2450 eV using an incident beam energy of 20 keV and samples of Au and Au on Si. These. angular distributions are compared to those of the Auger electron emis9ion of the same samples. The dependence of the shape of the secondary electron cascade on the angle of incidence and the take-off angle to the analyser is examined. In the case of Au, the results are reported for a bulk and a layered specimen conoisting of a thin Au film on an Si substrate. It is concluded for these samples that the contributions of the sample geometry and atomic number to the shape of the electron spectrum can be separated within a power law description of the secondary electron cascade.

Surface analysis of practical materials often presents a challenging situation in Auger spectroscopy and imaging because the yield of Auger electrons depends not only on the composition of the surface but also upon the local surface inclination (the topography) and the composition of the material below the surface. The use of simultaneous imaging of the same area of a sample with an electron spectrometer and four quadrants of a backscattered electron (BSE) detector has been demonstrated previously to be very useful for the separate correction of either topographical contrast or sub-surface composition effects. In this work the possibility is tested of using BSE detector images to correct simultaneously acquired Auger images for the presence of both topographical and sub-surface composition effects in the same sample. Measurements on model samples with known topography and sub-surface composition are used. The samples are formed from anisotropically etched Si(100) coated in some places with a thick Au film and then overcoated everywhere with either a thin film of CuSn alloy or partially coated with a thin film of Au. The correction process is successful in reducing the artefacts everywhere and the extent of this correction is discussed.