RESUMO
A Nion spherical-aberration (Cs) corrector was recently installed on Lehigh University's 300-keV cold field-emission gun (FEG) Vacuum Generators HB 603 dedicated scanning transmission electron microscope (STEM), optimized for X-ray analysis of thin specimens. In this article, the impact of the Cs-corrector on X-ray analysis is theoretically evaluated, in terms of expected improvements in spatial resolution and analytical sensitivity, and the calculations are compared with initial experimental results. Finally, the possibilities of atomic-column X-ray analysis in a Cs-corrected STEM are discussed.
RESUMO
Increasing the solid angle of X-ray collection is a major factor in improving the analytical sensitivity of X-ray energy-dispersive spectrometry (XEDS) in the analytical electron microscope (AEM). A new scanning transmission electron microscope, the VG HB 603, is equipped with two XEDS detectors with the largest collection angles (0.30 and 0.17 sr) available in commercial AEMs. However, large collection angles result in a large range of take-off angles, from approximately 4 degrees to 36 degrees, and the low angles can cause strong X-ray absorption. In order to investigate possible detrimental effects of the low (and of the range of) take-off angles on quantitative microanalysis of specimens exhibiting significant absorption, a stoichiometric Ni3Al thin-film, in which the Al Kalpha line is significantly absorbed, was analysed. Furthermore, the effect of different values of the collection angle on X-ray intensities was theoretically evaluated by numerical calculations and spectral simulation. These theoretical approaches permitted correlation of changes in the X-ray take-off angle (and hence X-ray absorption) with changes in the collection angle. It is demonstrated that approximately 0.30 sr detectors, with minimum take-off angles as small as 4 degrees, only result in maximum errors of 4% in the quantification of Al in Ni3Al and, therefore, further increases in collection angle can be pursued while maintaining current levels of accuracy of quantification.