Relation between quantification and crystal size L

arazbat

Hi there,

I have a problem with quantification in refinement that I am doing.

I have two compounds in my XRD pattern. I am refining lattice parameters and crystal size L in both compounds and crystal size L is around 40nm, fitting is medium (Rwp=10) and compound 1 is 85% and compound 2 is 15%. When I am refining strain L, crystal size is increasing toward 10000nm!, fitting is getting very better (Rwp=3) and compound 1 is 90% and compound 2 is 10%.

I can say when fitting is getting better compound quantification is getting near to its real value. But is it logical that refine strain L and crystal size goes to 10000nm?! I wonder can I leave this section to refine to have a better fitting while crystal size goes to very high value?

Thanks for your comments

Best

alancoelho

CS_L as described by the CS_L macro in TOPAS.INC comprises:

lor_fwhm = 0.1 Rad Lam / (Cos(Th) CeV(c,v));

CeV(c,v) is the crystallite size parameter which is being refined. When it goes to a large value such as 10000nm then the broadening is minimal and the parameter is redundant. The error in the parameter would be large and removing the CS_L from the refinement would be justified as basically its saying there's no broadening from crystallite size.

arazbat

Thanks AlanCoelho for your answer

actually there is broadening from crystallite size. As I am not refining CS_L the peak shapes are neither Lorenz nor Gaussian. But while I am refining CS_L they are getting really fitted. Problem is while i also refine strain L, fitting will get really better but CS_L goes to 10000nm!

Then, I can say I should keep strain L fix and not refine it and not care for better fitting. All means that I do not have any strain in my material. Is it a true statement?

Thanks

mfisch

No.

CS_L and Strain_L are approximations for Strain and Size effects. Just for a quantification, it may be safe to go after the best fit, so from what you've explained so far, I'd refine Strain instead of CS.

Remember that mentioned parameters are only significant if you have accurately characterised your instrument.

arazbat

Thanks Mfisch

Then as quantification is dealing with intensities and best fitting then I will care just for better fitting profile. And if I want to report crystalline size and lattice parameter for example I will go for most reasonably fitted profile.