Synthetic Multilayer Crystals for EPMA of Ultra-light Elements

Abstract

Quantitative wavelength dispersive EPMA of the ultra-light elements boron, carbon, nitrogen, and oxygen with a conventional lead stearate crystal (2d spacing approximately 10 nm) is severely hampered by a number of problems. Among these the most troublesome are the problems associated with low count rates, sensitivity to peak shape alterations, and the efficient transmission of higher-order reflections of alloying elements. Acceptable count rates can be produced for elements such as boron or carbon on elemental standards, but only at the expense of increasing the beam current to rather excessive values of 100–300 nA. Even under such extreme conditions, however, the count rates for nitrogen are usually unacceptably low. This is caused by the extreme absorption of N Kα x rays in both the stearate crystal and the detector window; both contain appreciable amounts of carbon, and the wavelength of N Kα falls right into the C K-absorption edge. The situation for O Kα x rays is much more favorable in this respect. In addition, the O Kα peak is rather close to the lower mechanical limit (short wavelength limit) of most types of spectrometers making it generally impossible to have proper access to the background on that particular side of the peak. This makes a correct background determination impossible.