The pseudo-Voigt that I've applied is essentially of the form
pV = n*Lorentzian + (1 - n)*Gaussian
Some of the peaks in my diffraction spectra are "very Lorentzian" and the best fit is obtained with a value of n > 1. Does this make sense and does it have any physical meaning, or should I put an upper bound on the fitting?
From V. K. Pecharsky, P. Y. Zavalij 'Fundamentals of Powder Diffraction and Structural Characterization of Materials':
'When during refinement η becomes negative, this is usually called super-Lorentzian, and such an outcome points to an incorrect choice of the peak-shape function. Usually, Pearson VII function should be used instead of pseudo-Voigt'
Thanks Alexander, that's very helpful. Does the book have a good description of how to interpret the Pearson VII (I can get it from the library) or are there any other good articles for this? I'm particularly interested in extracting information about relative levels of disorder/paracrystallinity between samples.
This book is mainly about structure solution, so it will be probably not very helpful in your case. I looked at the recent articles on the paracrystallinity and it seems that mainly they use the approach described in Journal of Physics C, 21 (1988) pp. 4155-4170. It is based on the integral breadth of the peaks which can be applied to the Pearson VII profile without any problems.
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction