Electron probe microanalysis using oxygen x-rays: I. Mass absorption coefficients

Abstract

Electron probe microanalysis of oxygen in a range of binary and ternary oxides is described. Experimental data are presented for a range of electron accelerating voltages and for two different take-off angle instruments. The ratio of oxygen K emission from specimen and standard is found to reach a limiting value, the same for each instrument, as the applied voltage is increased. The observations are shown to support the thin-film model (Duncumb and Melford 1966) proposed for quantitative analysis of light elements and also to enable its range of applicability to be established in terms of specimen characteristics and experimental conditions. The model is used together with chemical composition data for the oxides, and a consistent set of mass absorption coefficients are calculated which are considered to be accurate within 10%. These new data are discussed and compared with previously published values.

EDS quantification of light elements using osmium surface coating

“…Some of these studies particularly focused on the quantification doi:10.2465/jmps.141126 H. Ohfuji, ohfuji@sci.ehime-u.ac.jp Corresponding author of light elements such as oxygen through the accurate evaluation of the mass absorption of X-rays by the coating layer. For instance, Love et al (1974) used a thin (<10 nm) layer of copper for the surface coating of oxides such as MgO, Cr 2 O 3 , and Fe 2 O 3 and found that the results support the thin-film model proposed for the quantitative analysis of light elements (Duncumb and Melford, 1966). Jercinovic and Williams (2005) tested a thin (10 nm) gold coating in the analysis of monazite.…”

“…In pioneering studies that focused on the quantification of light elements, attempts were also made to control and evaluate the thickness of the surface coating for accurate attenuation correction. For example, Love et al (1974) and Weisweiler (1974) used a copper coating and measured the CuKα intensity to estimate the coating thickness, and Bizouard and Charpentier (1979) and Willich and Obertop (1990) used thin coatings of nickel and gold, respectively. A clear advantage of the present osmium coating is that it is not necessary to evaluate the coating thickness every time because of the high reproducibility and reliability of the plasma CVD coating process.…”

“…The extreme thinness of the osmium coating may also be favorable for chemical quantitative analysis via electron microprobe, although the Z number of osmium is as large as those of gold and platinum (i.e., high X-ray attenuation). In fact, several pioneering studies on EPMA quantitative analysis have been conducted using ultra-thin conductive coatings prepared from gold (Willich and Obertop, 1990;Jercinovic and Williams, 2005), copper (Love et al, 1974;Weisweiler, 1974), and nickel (Bizouard and Charpentier, 1979) by wavelength dispersive spectroscopy (WDS). Some of these studies particularly focused on the quantification doi:10.2465/jmps.141126 H. Ohfuji, ohfuji@sci.ehime-u.ac.jp Corresponding author of light elements such as oxygen through the accurate evaluation of the mass absorption of X-rays by the coating layer.…”

Evaluation of an improved absorption correction based on the Gaussian ionization distribution model for quantitative electron probe microanalysis

“…The standard material for oxygen was aluminium oxide and the mass absorption coefficient of OK~ line given by Henke and Ebisu is (22<Z_<79) and Love et al 19 (8 _< Z _< 22, Z = 80) were used. The standard material for oxygen was aluminium oxide and the mass absorption coefficient of OK~ line given by Henke and Ebisu is (22<Z_<79) and Love et al 19 (8 _< Z _< 22, Z = 80) were used.…”

Electron probe microanalysis of Nicalon fibres

“…In the case of carbon, however, it was found that the X-ray intensity ratio, k, was influenced by whether measurements were made using peak heights or peak areas; in the latter case the value was 1.1 times higher. Following previous practice [13], all carbon measurements were recorded using peak areas. For the silicon measurements (PET crystal), the K~ peak shapes were essentially the same for Nicalon and the standard (silicon carbide) so that the k ratio could be established directly from peak intensities.…”