It is widely know that physically a short beam size means better resolution because of the bean spread. This is a major argument for TKD (transmission Kikuchi diffraction) in SEM , however it was already a main argument for EBSD and it is for the maximum resolution for SEM imaging. On the other hand, typical working distances for EBSD are 18-22 mm which is quite big, and also for these investigations step widths of down to 20nm has been used years ago in order to resolve the microstructure of Cu interconnects in semiconductor devices.
Is there any paper which give a rule how the beam diameter changes with the working distance? Is it really that much different to work at 3 or 10mm with TKD or EBSD? I am sure there is a difference but how big is it?
Or is it like with the tilt of TKD samples where 20 degrees only change the beam size ration by 6% which is from my point of view negligible compared to EBSD (290%) and the practically always occurring astigmatism of the beam?
Magnetic lenses used in SEM and the overall operation may resemble that for an optical microscopes in terms of aberrations, astigmatism, etc. However, there are different factors impacting the beam size for optical microscopes vs. electron microscopes, considering factors related to the charge and mass factor as well as pressure, etc. You may find the following link useful for what you are looking for: http://www.charfac.umn.edu/sem_primer.pdf
Interesting question indeed!
Everybody knows by increasing the working distance the resolution drops a bit; that is why in demos usually microscope companies show images in small working distances ;). But how much, maybe those companies know but don't wanna reveal the answer. Shall we define a project together?
This would be really interesting if it hasn't been already done... It would surprise me...but one never knows.
@ B. Roughani: this paper is indeed very interesting. I have to check whether it answers my question completely. Thanks!
Our company uses Real time In-Flight Beam Tracing™ where we calculate beam spot at every condition you use at every microscope you are working on. It does not depend only on WD you use, but also the beam intensity matters a lot and of course it strongly depends on the type of the microscope.
We are soon going to have a brand-new TESCAN FIB/SEM. I can check Real Time In-Flight Beam Tracing more in detail ;) might help us resolving this question in detail
@Rostislav: is there any document you can give us as example? Thanks in advance...
In SEM, we know that the probe formed by the last condenser is the object for the final objective lens. Let, "object" mean the final condenser probe cross-section and "image
" mean the same of the final beam probe spot. The aim is the "de-magnify" the object size, to get a finer final beam probe.
Then, it can be inferred from simple lens-maker;s formula that for greater de-magnification value, the ratio of image distance to object distance should be as small as possible. Image distance in this case is the working distance, which is the distance between objective lens axis and the specimen surface. So, for a finer beam probe in SEM, it is necessary for the working distance to be small.
Note: A very small beam probe alone doesn't ensure better resolution, if "brightness" of the illuminating pencil isn't great enough, which means no. of electrons should be appreciably large in the beam probe, or else Signal-to-noise ratio of the imaging system would be inferior. To be brief, with the same working distance a FE-SEM would yield better image than Thermionic W-SEM.
Good, but how big :-). I guess, everybody knows the general information. I can only talk about my experience in EBSD and we are usually working with a WD of 18mm and scanning down to 30-50nm. We still can get indexable patterns but this doesn't say anything since as long as one pattern still delivers the main information, it will be indexable.
Honestly, I have limited experience with EBSD or TKD. We mostly work with the SEM instrumentation, and most of the time the only imaging mode is the default SE imaging for us. But with that we've seen the working distance has a significant effect. For example, while imaging the standard gold on carbon sample under an old JEOL T330 machine, we could resolve the individual particles only at the minimum working distance i.e. 8mm.
However I believe, for EBSD a somewhat medium value of working distance would be optimal. Because, with minimum working distance there is bound to be loss in the beam current owing to greater converging angles and for better signal one needs higher beam current. For medium values of WD, resolution would be a bit inferior but signal strength would be significant.
- Badges
- Science method