Materials Problems and Approaches for TEM and TEM/STEM Analyses - Sample Preparation Topic for Transmission Electron Microscopy - page 81

12 Relationship Between Sample Thickness and Analysis Type
in TEM and TEM/STEM
One of the characterization limits is sample thickness. Whereas thickness can reach
values of 200 nm for bright-field and dark-field images, in the case of high resolu-
tion, the maximum thickness is 50 nm for energy-loss analyses. As the optimum is
between 5 and 20 nm, high-voltage microscopes are used for such analyses.
However, this upper limit is not as favorable for CBED and EDS analyses which,
respectively, require a minimum sample volume to ensure 3D interactions result-
ing in the formation of HOLZ lines, as well as to have a detectable X-ray output.
Knowing the thickness of the sample determines the quantitative chemical analysis
of the interfaces. This can be determined using CBED diffraction or using plasmons.
Table 4.4 shows the ranges of sample thickness required for the various analysis
types.
For materials composed of light elements, the optimal thickness of the TEM
sample for observations in bright-field mode is between 50 and 100 nm. Overly thin
sections lack contrast, while overly thick sections lack definition. This is because
several different structures are mixed in the observation plane. For these materials, it
is also preferable to work with beam acceleration voltages between 75 and 120 keV
in order to achieve an image contrast as high as possible.
13 Assessment of TEM Analyses
The use of various observation modes (Table 4.5) in parallel on the same area to be
analyzed is very helpful. Analyses can be coupled in image and diffraction modes
and chemical analysis can be performed for the same spatial resolution (sizes of
areas analyzed at 20 Å on an analytical 200 keV microscope). Using the image
mode, local structural defects, grain boundaries, intergranular phases, interfaces,
segregations, precipitates, etc., can be characterized. Using the diffraction modes,
the local structure can be determined at the same time that chemical analysis by
EDS and/or EELS is performed. However, it is necessary to take into account that
the optimal EDS analysis conditions are different from those of EELS, in particular
with regard to the sample thickness. Note that the recommended thickness for EELS
analysis is excellent for high resolution; therefore EELS and HTREM analysis can
be combined for the same preparation.
The best compromise must always be found for the optimal use of all these
techniques, in order to respond to a problem without damaging the material.