Premium
The shape of the background in AES: Nonlinear features in log N ( E ) v . Log E
Author(s) -
Peacock D. C.,
Duraud J. P.
Publication year - 1986
Publication title -
surface and interface analysis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.52
H-Index - 90
eISSN - 1096-9918
pISSN - 0142-2421
DOI - 10.1002/sia.740080102
Subject(s) - spectral line , ionization , auger electron spectroscopy , linearity , logarithm , physics , electron , auger , atomic physics , secondary electrons , ionization energy , spectral shape analysis , mathematics , ion , quantum mechanics , mathematical analysis
By replotting emitted electron spectra, N ( E ), as log N ( E ) v . log E , parts of the background which obey a power law can be identified as linear segments. The importance of this linearity for Auger electron spectroscopy (AES), and the conditions under which it can be observed, are discussed. A simple expression is presented which describes the essential shape of the spectral background for analysis energies in the range 50–2000 eV and primary energies between 3 and 20 keV. It has been suggested elsewhere that such simple expressions are inadequate because ionization of inner shells by backscattered and secondary electrons can introduce broad low intensity features which disrupt the linearity of log N ( E ) v . log E . In the present work numerically calculated spectral shapes are compared with experimental AES spectra from various specimens including Al and Si. It is shown that some broad features which are often observed in EN ( E ) (constant spectrometer resolving power) spectra arise solely as a consequence of replotting power law backgrounds in this form and cannot be adduced as evidence of the ionization mechanism. Furthermore, subsurface electron sources can cause nonlinearities in logarithmic spectra which should not be ascribed to the ionization mechanism. It is concluded from a comparison of calculated spectra with those observed experimentally that the ionization mechanism causes only small deviations from linear behaviour which should not prevent the fitting of power law backgrounds to experimental AES data.