Some Issues in Quantitative X-Ray Photoelectron Spectroscopy and Auger-Electron Spectroscopy

摘要

We give an overview of two issues that are important in quantitative X-ray photoelectron Spectroscopy (XPS) and Auger-electron Spectroscopy (AES). First, we review the use of XPS standard test data (STD) that consist mainly of synthesized doublets. The STD were designed to test the reliability of peak energies and peak intensities obtained from fits to measured spectra containing overlapping components (i.e., that are often associated with different chemical states). The biases and random errors in peak positions can each be 0.1 eV or more when the doublet peak separation is less than 0.5 eV. In such cases, the median biases in relative peak intensities for the doublet components could be up to a factor of four, while the random errors could be up to factor of two. Second, we discuss the effects of elastic-electron scattering in measurements of overlayer-film thickness by AES or XPS. In such measurements, the effective attenuation length (EAL) should be used instead of the inelastic mean free path (IMFP). As an example, we present calculations of practical EALs for thin films of silicon dioxide on silicon for common XPS measurement conditions. For Mg and Al Kα x rays and photoelectron emission angles less than about 60°with respect to the surface normal, the average practical EALs are between 6.5 % and 9.4 % less than the corresponding IMFPs; the actual difference depends on the x-ray source, the SiO_2 thickness, and the XPS configuration. For larger emission angles, the practical EALs can vary appreciably with SiO_2 thickness and emission angle; in these circumstances, practical EALs should be determined for the specific measurement conditions. If XPS is performed with smaller x-ray energies (i.e., with synchrotron radiation), the practical EALs can be up to about 30 % less than the IMFP and can differ for emission angles of 0°and 55°as found in recent experiments.