Grain size can be calculated with the help of Scherrer equation. It will give you the value of mean size of the ordered (crystalline) domains, which may be smaller or equal to the grain size.
Before using this equation, you should have XRD data analysis of your sample through which you have to find out FWHM value.
You can use XRD data to calculate the grain size of nanocrystals. You have to use Scherrer equation. Also you can get an idea about the grain size from HRTEM image analysis.
You can use XRD data to calculate the grain size of nanocrystals. You have to use Scherrer equation. When you have SEM micrograph you can use linear intercept technique. Though the former one is called the crystallite size.
Using Debye Scherrer formula you can calculate crystallite size. Using Willimason Hall plot you can calculate crystallite size as well as strain. From FESEM, you can calculate particle size, strain and agglomeration. For finding particle size from FESEM micrographs manually, you can use the software "ZEN". I have done so and it gave good results. The published paper of Ru doped ZnO nanocrystals has indicated that. Best of luck.
Be very careful with using the Scherrer equation. It gives you average crystallite size assuming that you have round crystals. You can also find that if your sample is nanocrystalline and somewhat disordered, XRD won't even show you any peaks! (And like Jorge correctly said, twinning of crystals complicates matters further!)
It depends on your sample. If it is a solid sample (ex. a metal), EBSD is an excellent technique for this (specifically, TKD for really small crystals). If it is nanoparticulate, I would do TEM. I think some more information is required about your sample though.
I suggest to use WH-plot or WA instead of Scherrer Formula only. And if u wanna go for advance analysis (domain size distributioin, dislocation, faulting, APB, GSR, etc), look fo WPPM*.
First of all u must take a pattern from the standard sample (SRM 660b, 1976b, 640d etc) to get the instrumental (optics) broadening**. If u use standard setting of BB-geometry, it's better to use point-detector instead of line-detector (depend on detector) with longer step-scan. Scan both with the same instrument setting.
Refine the SRM pattern to get the Cagliotti and Lorentz fraction and assymetry (Thompson-Cox-Hastings pseudo-Voigt convoluted with axial divergence asymmetry function). Use this parameter and keep fixed along the sample refienemt, so the broadening is only the size-strain effect (simplify). Most software come with the template for SRM, so when do refine the real sample simply insert the pattern, relase the microstructure parameter and do refinement.
*M. Leoni, T. Confente & P. Scardi, “PM2K: a flexible program implementing Whole Powder Pattern Modelling”, Z. Kristallogr. Suppl. 23 (2006) 249-254.
TEM will be more reliable than XRD when some of the grains are very small and/or highly elongated. SEM will give you a quick answer when the grains are >> 10nm, and will provide better statistics on large grains. Hence, combining SEM and TEM is usually a good idea.