The Hole-Count Test Revisited: Effects of Test Specimen Thickness

For historical reasons the hole count, an important performance test for the Analytical Electron Microscope (AEM), is somewhat arbitrary yielding different numbers for different investigators. This was not a problem a decade ago when AEM specimens were often bathed with large fluxes of stray electrons and hard x rays. At that time the presence or absence of a thick Pt second condenser (C2) aperture could be detected by a simple comparison of the x-ray spectrum taken 'somewhere in the hole' with a spectrum collected on a 'typical thickness' of Mo or Ag foil. A high hole count of about 10-20% indicated that the electron column needed modifications; whereas a hole count of 1-2% was accepted for most AEM work. The absolute level of the hole count is a function of test specimen atomic number, overall specimen shape, and thin-foil thickness. In order that equivalent results may be obtained for any AEM in any laboratory in the world, this test must become standardized. The hole-count test we seek must be as simpl and as nonsubjective as the graphite 0.344nm lattice-line-resolution test. This lattice-resolution test spurred manufacturers to improve the image resolution of the TEM significantly in the 1970s and led to the even more stringent resolution tests of today. A similar phenomenon for AEM instruments would be welcome. The hole-count test can also indicate whether the spurious x-ray signal is generated by high-energy continuum x rays (bremsstrahlung) generated in the electron column (high K-line to L-line ratio) or uncollimated electrons passing through or around the C2 aperture (low K/L ratio).