Electron beam diameter, spot size, and pixel size of the image are typical parameter to describe the the scanning behaviour of an SEM. Investigating small-grain microstructures like martensite at low magnification, e.g. in the order of x2.000, by EBSD, often the hit rate is quite low which is commonly explained by the high fraction of grain boundaries. Going to higher magnifications like x10.000 and more, the hit rate increases fortunately (e.g. from 50% to 95%) without changing any focus conditions, but at the same time the statistical significance (less number of grains) decreases. Since the beam diameter for a FEG-SEM is assumed to be in the scale of a few ten nanometers and smaller, the distance between adjacent measuring points is usually clearly bigger than the assumed beam diameter. Obviously the spot size on the sample surface is bigger than expected. This can be tested applying the same step width for an EBSD scan at different magnifications. What are possible reasons? Is any SEM - and I am not talking about the different filament types - affected in a similar way, or are there differences between manufacturers?
This confirms the answer I've got from Zeiss. However, obviously this very narrow beam is not well positioned and seems to average over an area which can be much bigger than the beam diameter. Otherwise one wouldn't see twin boundaries which are perfect planar defects with atomic resolution and practically no mismatch. In so far this beam behavior is very useful for imaging, but for techniques like EBSD the beam diameter is perhaps misunderstandable.
Finally, the beam size is not the important information but the spot size on the sample surface, right?
I also think it has something to do with the deflection coils, but more likely the control signal could be the cause. Let me give you an example: Let's say we are at a view field of 100µm. The position of the beam inside this field is controlled by the signal given to the microscope from the EBSD controller in a -5V to 5V range (for Tescan microscopes at least). To hold a 5nm diameter beam on its position for a longer time, the control voltage would have to have a stability in the µV range. Otherwise the spot moves around its central position, affected by the noise of the control voltage. So the beam size is still quite small, but the "illuminated area", or spot size, is something completely different.
Have you ever tested if the hit rate is influenced by the pixel exposure time (assumed the pattern quality stays good enough)? A shorter dwell time could come with less "position noise".
Maybe a precise measurement of the control voltage could explain your findings.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science