I couldn't understand, what is the relation between grain size or particle size and grain boundary scattering. And also i want know about, what is the role of grain boundary or impurity scattering in electrical properties.
if you are asking regarding varistors and their propertis, please check the article in attachment. If it is not giving answer to your question, dig in similar direction :-)
- each of them has its personal model (or mixed model for textured crystals)
It is difficult to explain them as one problem.
Scattering. You see, X-ray is scattered by the electron density levels, - jumps, levels, domain walls, - they can be visible and invisible boundaries - in dust or in single crystals.
The mass of such thin objects is negligible, so on the intensity of scattered X-ray is negligible too.
Intensity (shape and peak) of X-ray scattered radiation from domains depend on the size and shape domains.
In Small-angle X-ray scattering you get the size of grains, it may be more, then coherent scattering regions (CSR) of X-ray diffractometry. - In XRD you see core size, in SAXS you can see sum size: inner single crystal cluster with the shell, covering this hard core.
For example in UDD (ultra dispersed diamond) CSR is less, then SAXS getting size, as the weght of diamond core and graphit shell is equal.
In SiO_2 crystals (or in SiC, in Al_2O_3-saphir or in diamond) sometime domain walls are invisible. Wideness of X-ray peak shows that we have gnarl (knot), or crystal planes shifts are the boundaries of the continuous electron density that change the type of scattering. From the shape of the distorted peaks you can estimate the size of continuity in an imperfect crystal.