1. SE created by interaction with ingoing electrons of focused beam. They are all created at (or very near) surface, reflect topography of the surface with good resolution and bear no information about inner structure of specimen.
2. SE created by outgoing backscattered electrons when they are escaping specimens. These SE are also created near surface, but their number is proportional to backscattered electrons which have information about deeper regions. This signal has much worse resolution (about the same as BSE signal)
Image we see is created by interposition of these two signals and input of first one is much stronger, especially at lower accelerating voltage.
passage of 20nm (= 200A) of carbon is quite hard stuff for secondary electrons. Escape depths of about 30A (= 3nm) are reported:
Article Secondary Electron Emission from Thin Carbon Films
What make you confident that you have got a 20nm carbon layer (mass coverage equal to 20nm bulk carbon) and not much less in thickness?
With respect to Thomas's answer, please repeat you SEM measurement with lower voltage. So your x-ray output will be smaller as well as the maximum energy of the electrons.
1. SE created by interaction with ingoing electrons of focused beam. They are all created at (or very near) surface, reflect topography of the surface with good resolution and bear no information about inner structure of specimen.
2. SE created by outgoing backscattered electrons when they are escaping specimens. These SE are also created near surface, but their number is proportional to backscattered electrons which have information about deeper regions. This signal has much worse resolution (about the same as BSE signal)
Image we see is created by interposition of these two signals and input of first one is much stronger, especially at lower accelerating voltage.