Quantitative XPS and AES analysis of surfaces requires models to do fast calculations of the energy loss of electrons at any energy as they move in solids of general and non‐uniform composition. This is possible with the Universal cross‐section that was introduced a decade ago. A critical review is given of the Universal cross‐section in the light of the results of research carried out since then. In particular its validity is compared to that of experimental cross‐sections determined by analysis of reflected electron energy‐loss spectroscopy (REELS) spectra. It is shown that for applications in quantitative surface analysis by XPS and AES, the solids can be divided into classes according to the full width at half‐maximum of the dominating shape of the inelastic scattering cross‐section. The Universal cross‐section is quite accurate for solids with a cross‐section width 20 eV. For solids with a cross‐section width of 10–15 eV, the Universal cross‐section is still fairly good for the description of the far‐peak region (30 eV from the peak energy) but it is less accurate to account for the near‐peak region (10 eV from the peak energy). For solids with a cross‐section width 5 eV, the REELS cross‐section is always more accurate than the Universal cross‐section. A Three‐parameter Universal cross‐section was defined, which gives a good fit to narrow experimental cross‐sections. Parameters for different solids were determied and it was shown that the polymers form a separate universality class. © 1997 by John Wiley & Sons, Ltd.

No abstract

X-ray photoelectron spectroscopy (XPS) is one of the most widely used techniques for surface analysis. In the technique, which has been applied by industry for more than 30 years, atom core electrons are excited by soft X-rays and the kinetic energy of the emitted electrons is measured. Peaks occur corresponding to the binding energy of the core electrons and their position gives therefore direct identification of atoms. Since the inelastic mean free path for electrons in the energy range studied (typically ~ 0.1 to 1.5 keV) is 0.5 -3 nm, the peak intensity originates from electrons excited from atoms that are within a 2-5 nm thin surface layer. The core electron binding energy varies slightly depending on the chemical environment of the atom, and this is used to get information on the chemical bonds. Scanning XPS has been increasingly more used in the past decade driven by substantial improvements in lateral resolution. The interpretation of XPS has largely been based on simple peak intensity measurements and this strongly limits the quantitative accuracy of XPS. The depth resolution of XPS has usually been done by ion-etching the surface. In the past decade, significant advancements has however been done in the development of improved algorithms for quantitative XPS and non-destructive atom depth distribution analysis in the 0-10 nm depth range [1]. Lateral resolutionThe lateral resolution in XPS-mapping has in general been inferior to scanning techniques based on electrons or ions partly because these are easier to focus to a small spot-size and partly because of a low photoelectron yield. Improvements have been done by new designs of the X-ray source and the electron energy analyzer and the resolution has improved gradually from ~1 mm resolution 20 years ago to ~ 3 µm today. With the construction of the third generation synchrotron radiation sources, which provide very intense photon beams it is now possible to get better than 100 nm resolution and this is expected to be further reduced to better than 5 nm in the near future [2]. Depth resolution.Since the electron inelastic mean free path is ~1nm, the XPS signal attenuates strongly with the depth on the nano-meter scale. It is this property that makes XPS interesting and powerful because it gives the technique its high surface sensitivity. However this is also the source of the high uncertainty in quantitative interpretation of XPS unless it is accounted for. The fundamental problem is illustrated in fig.1 which shows spectra for different depth distributions of Cu in a Au matrix. Due to the attenuation, the Cu2p XPS-peak intensity from all four different surface morphologies is exactly identical although the surface compositions are very different. Quantification based on the peak intensity alone is thus subject to a large error. However, the energy distribution in a wider energy range below the peak depends critically on the depth distribution of atoms. These are electrons that originate from the peak but lose energy on their way out of the solid. It is th...

It is well known that the usual procedure for quantification by electron spectroscopy that is based on measured peak intensities is highly unreliable. An improved method is to take into account that the peak shape in a wide energy range, on the low kinetic energy side of the peak varies considerably with the surface morphology on the nano-meter depth scale. This observation has in recent years been applied in the formulation of a by now well known method for quantification that is based on quantitative analysis of measured peak shapes. The technique is sensitive on the 1-10 nm depth scale and it is non-destructive. The method suggested for the extraction of quantitative information from the large variation of the inelastic background with atom depth distribution will be reviewed briefly here. An example of practical application of the technique is also shown.

No abstract