As it concerns broadening but also splitting peaks, I can offer a series of works, dealing with distortion of lattices (strain) but also distortion distribution (microstrain).
All metrical change you can quantify in terms of a symmetric strain tensor (or by 1+ epsilon, called the strech). Moreover, a part of the line broadening can be interpreted as the measure for the distribution of these strains. Sorry for a series of self references:
microstrain broadening
-Article Understanding anisotropic microstrain broadening in Rietveld...
Dilemma to distinguish strain/lattice distortion and microstrain
Ways to analyse crystal structure data in terms of metrical distortion from an ideal structure
Article The monoclinic lattice distortion of η′-Cu6Sn5
Article Crystal structures of Fe4C vs. Fe4N analysed by DFT calculat...
Note, however, as it concerns microstrain broadening: lattice parameter distributions do not necessarily map unambigeously on the 2theta scale, of there are short-range strain variations
Article Notes on the order-of-reflection dependence of microstrain broadening
the theory behind the width of the diffraction peaks is simple: The more lattice planes with identical orientation are contributing to a diffraction peak, the sharper the diffraction pattern will be. It is analogous to the diffraction of visible light from a grating or a slit system of parallel, identical slits: The more slits are illuminated, the sharper the peaks will be. You may cross-check the literature in any optics book
In my opinion, the lattice distortion could be inferred from the microstrain which could be extrapolated from XRD data using the Williamson-Hall equation. The use of the Williamson-Hall equation could be referred from the following work.
I have attached five ppt slides, try to read them and it may give an insight into lattice distortion quantification if you have any difficulty in the understanding post it in RG.
My question is...even if there is no strain in the system, still there is a peak broadening. What could be the reason for that (apart from instrumental broadening, crystallite size, etc.)?
What could be the best way to characterize/quantify the "atom not sitting at its ideal expected lattice position"?