Introduction Due to the rapid development and expansion of materials science into new areas such as high-temperature materiais, wear-resistant coatings, and modern ceramics, interest in quantitative electron probe microanalysis of ultralight elements, such as B, C, N, and 0, elements which are usually present in considerable quantities in such materiais, is also rapidly growing. Compared with conventional analysis of medium-tohigh Z elements (Z > 11), successful analysis of ultralight elements

more »

... equires far more stringent adherence to all of the steps involved in the complete procedure; beginning with the specimen preparation, followed by the actuaI intensity measurement, and ending with the matrix correction. As far as specimen preparation and intensity measurements are concerned, a number of specific problems have already been discussed at length Heijligers 1984a, b, 1986a-c), the major problem being the very low count rates and frequently low peak to background ratias, especially for an element like nitrogen. There is also a theoretical possibility that x-ray emission for ultralight elements may exhibit systematic differences from one compound to another. Evidence to this effect has recently been found, for example, for B- -compounds (Bastin and Heijligers 1986b, c). Additional practical problems such as a lack of electrical conductivity which can strongly affect the intensity measurements, are gradually being recognized. Regarding matrix correction, necessary in order to convert the measured intensity ratios (k-ratios) into concentrations, the corrections for ultralight elements can be an order of magnitude larger than in conventional analysis. The demands imposed upon a specific matrix correction program are correspondingly far reaching. Furthermore, it is vital that certain physical quantities such as, for example, mass absorption coefficients (MACs)