Limiting factors in specimen thickness in conventional and scanning transmission electron microscopy

Abstract

Specimens of silicon containing grown-in stacking faults have been examined in conventional transmission electron microscopy (CTEM) and scanning transmission electron microscopy (STEM) at voltages between 80 and 200 kV. It is found that the use of a tungsten filament STEM severely limits the penetration because of the small current available in the (~50 Å) probe. In addition, the large beam divergence which is used generally causes the orientation of best transmission to dominate the contrast. When using a field emission (FEG) STEM this limitation is found to be largely overcome although the useful intensity is still lower in FEGSTEM than in CTEM. The factor limiting usable specimen thickness is found to be associated with a top-bottom effect in both CTEM and FEGSTEM; contrast information is lost from the top† of the specimen in CTEM and from the bottom in STEM. The maximum usable foil thickness corresponding to the loss of fringe contrast associated with this top-bottom effect is approximately the same in FEGSTEM and CTEM although larger fields of view can be more usefully observed in CTEM. These observations are contrary to earlier qualitative experimental work and to theoretical predictions which are based on the assumption that chromatic aberration sets the limit to usable foil thickness in CTEM. This assumption is not correct if the required resolution is about 30 Å–a typical requirement for the examination of metallurgical specimens.