So, is your stress measurement instrument an optical one which measures the curvature and calculates the stress from it?

In that case what you want should be possible, but your silica sample should be decently planar. It probably won't bend that much and if it has an intrinsic inhomogeneity from the beginning, the fit that determines the stress will create nonsense because the functions it uses don't match the shape.

Edit: since you inserted that it's x-ray: Copper should have enough electrons so that an analysis should already be possible at 50nm; for carbon that wouldn't be sufficient yet. Please be aware that you will probably measure the envelope of a spread from the nucleation zone to the top layer.

I have an idea: Stressed copper has different phases than homogenously equal-phased, "relaxed" copper. Therefor conductivity of stressed copper changes non-linear during exposure to X-Ray radiation, while relaxed copper conductivity behaves almost linear.

The conductivity changes during X-Ray radiation exposure due to ionization.

Let me know if that was a good idea.

Yes. Probable instrument and method:

Article BL-02: a versatile X-ray scattering and diffraction beamline...

For this, please refer to the preprint article link http://dx.doi.org/10.13140/RG.2.2.23849.40808, titled “Determining and quantifying chemically produced stresses in (atoms of) electronic and crystalline materials”