On SEM: I tend to think that a good model for secondary electrons is (per point) (material constant) * C*(derivative of height).

It becomes much more clear in a cross sectional SEM: you see the material contrast in SE quite clearly, though it is thickness dependent to a small degree. Any surface height creates confounding signals.

The same is true on surfaces.

There is some multiplier, and it depends on the details of the detector and the settings. Quite to my surprise, images on a Hitachi CDSEM and a FEI SEM were wildly different, even for SE images, even after attempts to match them for equivalent QE. A good part of this is probably the source brilliance.

I will look up the results, as this has been a longstanding curiousity.

“The final part of my presentation was “Lessions Learned” resulting with Elmar Platzgummer in electron Multi-Beam Mask Writer MBMW-101 HVM for the 7nm node in 2017, and MBMW-201 for the 5nm node in 2019. Without Elmar Platzgummer and his team this would not have happened!” – Hans Loeschner