you know that any height variation of the sample with respect to the optical center of the XRD system will shift the diffraction peaks a bit. So for a sufficiently rough or fractured surface you will get a peak broadening. This holds for metals and non-metallic material.
So as consequence any conclusions from peak broadening should be taken very carefully in such cases...
Basically when you have a higher roughness in the surface you will increase the scattering in different directions, leading to a broadening in the peaks. A small shift to lower angles can also be observed due to the Bragg principle.
Note that Gerhard Martens' comments refer to specific diffraction geometries, e.g. the most common focussing geometries, but to a much lesser extent to parallel beam geometries. Note that surface roughness also (in any case) has effects on the 2theta-dependent reflection intensities due to absorption effects.
To get a sustainable solution to your problem, please refer to the preprint article given at link DOI: 10.13140/RG.2.2.27720.65287/3 or at link https://www.researchgate.net/publication/352830671.