391 23
background, making it more difficult to discern the position
of the bremsstrahlung maximum. Nevertheless, the general
shape of the EDS bremsstrahlung continuum is a powerful
indicator of geometric effects that modify X-ray produc-
tion, both bremsstrahlung and characteristic X-rays.
. Figure 23.13 shows this effect for ideal flat polished K411
glass and several shards. The deviation in spectral shape
from the ideal case is readily apparent, and analysis of spec-
tra with such severe deviations in shape results in large
deviations in the relative accuracy of the normalized quan-
titative results.
23.5 Best Practices for Analysis of Rough
Bulk Samples
The optimum approach to the analysis of a rough specimen is
obviously to prepare a polished flat surface, but the analyst
may be confronted with a situation where no physical modi-
fication of the as-received specimen is permitted. That is, the
rough surface itself is the object of interest, so that grinding
and polishing would modify or destroy the material that is
actually necessary for the final result. How should the analyst
proceed in such a case?
The analysis of rough surfaces is inevitably going to be
compromised compared to analysis of the ideal flat polished
specimen. The analyst must seek to obtain the best possible
result under the circumstances, so the analytical strategy
must be carefully considered. Electrons that backscatter off
rough surfaces are likely to produce remote excitation of
X-rays from material(s) that are likely to differ from the
location where the beam is striking, as shown schematically
in. Fig. 23.14. It may be thought that the collimator on the
EDS will restrict the view of the EDS to just the region
directly excited by the incident beam. This is not the case.
The collimator typically permits acceptance of X-rays with
at least 50 % efficiency from an area that is 5 mm in diameter
or larger. The exact transmission response depends on the
particular EDS detector and its collimator, but the region of
transmission can be readily determined by mapping a uni-
form target, for example, an aluminum SEM mounting stub,
at the lowest magnification setting (maximum sized scanPhoton energy (keV)CountsPhoton energy (keV)Counts10000010000100008000600040002000100010010
024681012141618200
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0B_20kV10nAB_20kV10nA. Fig. 23.12 EDS spectrum of polished boron at E 0 = 20 keV
23.5 · Best Practices for Analysis of Rough Bulk Samples