We have attempted to measure the O/Ni inner-shell cross-section ratios from a sample of NiO using EELS in an electron microscope. The results lay within roughly 10% of the predictions of both the hydrogenic and Hartree-Slater theories. However, the data showed a spread of around 20% which may have been caused by artefacts introduced during the specimen preparation.
I N T R O D U C T I O NElectron energy loss spectroscopy (EELS) is a technique which can be employed in electron microscopy for determining the elemental concentration ratios in small volumes within a sample. The most commonly used procedure for quantifying the energy-loss spectrum involves recourse to theoretical inner-shell atomic cross sections and there have been few attempts to determine concentration ratios using purely empirical procedures (Grande & Ahn, 1983;Malis & Titchmarsh, 1985). Such procedures have been applied for many years in X-ray analysis and include, for example, the k-factor method proposed by Cliff & Lorimer (1975). The relevant k-factors in EELS differ from those used in EDX in that they are a function of the energy integration range, the collection angle and the specimen thickness. In addition, the near-edge structure of the core-loss signal depends on the chemical environment of the atomic species of interest which, together with the previously mentioned factors, complicates the implementation of the k-factor approach as a routine method of analysis in EELS. However, in very thin samples, where the scattering parameter (the ratio of specimen thickness to inelastic mean free path) is less than 0.5, the EELS k-factor is simvly the ratio of the inner-shell vartial ionization cross sections for the A -two elements involved, provided the energy integration region is large enough to make fine structure effects negligible (Egerton, 1981a). Thus, a k-factor approach can be used to provide information on the accuracy of the predictions of theoretical atomic models.The main problem with the k-factor method lies in the need to find suitable standards of accurately known composition. We have carried out measurements on single-crystal NiO using a VG HB5 STEM equipped with post-specimen lenses (Craven & Buggy, 1981) and