The influence of the fundamental physical parameters on standardless electron probe microanalysis

In the beginning of this work all methods for the standardless analysis by electron probe microanalysis are presented. The most popular method of net intensity calculation is investigated in detail concerning its fundamental parameters. Possible errors caused by these fundamental quantities as well as possible error compensation (intended or not) are discussed critically. The most reliable fundamental parameters are selected out of the huge variety of data available in literature and applied to standardless analysis of stoichiometric samples. A necessity for accurate conventional standardless analysis using simulated net count rates is the accurate knowledge of the detector efficiency. The detector response becomes rapidly worse by the time and its determination must precede any standardless analyses. A new method for determining the detector efficiency using relative count rates is presented and possible errors are estimated critically. Special interest was drawn on the relative intensity factor of Lα radiation used for the deconvolution of convoluted L peaks. This value is not a constant, as sometimes suggested in literature, but depends on many measurement parameters as acceleration voltage, take-off angle, detector efficiency and composition of the sample. Theoretical considerations concerning this subject are used to prove mathematically that the determination of mass absorption coefficients using the variable voltage method by measurements with energy dispersive spectrometers do not yield the desired mass absorption coefficient of Lα radiation but the mass absorption coefficient of a (fictitious) L line which is weighted by the respective Lα and Lβ lines. (author)