In practice JEOL shows in the delivery of a atomic resolution machine the silicon dumbbells in , see for instance

http://www.jeolusa.com/DesktopModules/Bring2mind/DMX/Download.aspx?Command=Core_Download&EntryId=594&PortalId=2&TabId=320

There is a National Standard named "GOST R 8.631-2007 - Scanning electron measuring microscopes. Method for verification." in Russian. Try this one. You need paragraph 9.4 and picture 1. They use gauges made by "Center For Surface and Vacuum Research".

http://www.complexdoc.ru/pdf/%D0%93%D0%9E%D0%A1%D0%A2%20%D0%A0%208.631-2007/gost_r_8.631-2007.pdf

http://www.nicpv.ru/

Above mentioned standard looks like an advertisement for test specimen manufacturer. I’ve never used this test specimen, so I have no comments on its usability. It’s measurements are based on a simple method of knife-edge beam diameter measurements, you can read about it here:

https://www.princeton.edu/~achaney/tmve/wiki100k/docs/Beam_diameter.html

www.science.marshall.edu/dneff/temp/KG-beamparams-Lab3-dnrev.docx

With good “knives” (test specimens) method works well for larger beam sizes. For smaller beam sizes things are rather complicated, and in most cases beam size estimations are based on resolution measurements.

I have done it here:

http://www.danilatos.com/FESEMdan/FESEM172s.html

(see pp 172-178), and discussed it again here:

http://www.sciencedirect.com/science/article/pii/S0968432812002351

There may be a problem with the very fine probe sizes of modern machines (I am not sure, really), but the above links may give you a starting idea on how to refine the techniques. I think the discussion in the first link above may lead you to the correct choice of parameters and conditions even for very fine probes.

lets say, your setup has the cathodoluminescence capability that you could spatially excite semiconductor structures and, you have semiconductor nanowires (in order of a few hunderds nm in diameter) that were grown with well-difined distance from eachother. if you look at the emission wavelength of NWs while scaning through the sample, you could probably have a rough estimation of your beamsize. I am not a expert in this field, Please correct if I am wrong.

Dear Savutjan Sidik , how can you measure diameter of electron beam (about 1 nm) with a help of nanowire of a " few hunderds nm in diameter "?

Vladimir Dusevich I was told that beam diameter is in a few hunderts nm when I visited a CL lab quite long ago. In fact, it is not the actual beam diameter rather the actual excited area in the sample surface considering the carriers diffusion.

Sorry, but you've got a very wrong information.