>
> >Very often, a simple assumption that
> >size-broadened profile has both Lorentzian and Gaussian terms yields
> >satisfactory results.
>
> Well, this assumption is easily entered in a pseudo-Voigtian or Voigtian
> profile shape model. But the corresponding size Fourier coefficients
> AnS are never checked in order to see to what size distribution
> function the final profile shape is really corresponding to. This is
> something easy to do : calculate the cosinus Fourier transform,
> then calculate the second derivative, and see how is the
> corresponding size distribution function. I have never seen
> that elementary calculation done. Hope you agree with me
> that a serious approach should not avoid such an ultimate
> checking ?
Yes, it was done (see J. Appl. Cryst. 26 (1993) 97-103) and the
column-length distribution function appears to have an "acceptable"
asymmetric shape for tested samples. It is of the form:
[(2 pi L B_G^2 + 2 B_L)^2 - 2 pi B_G^2] exp(-2 L B_C - pi L^2 B_G^2)
Here, L = na3, averaging distance perpendicular to the reflecting planes, B
integral breadth, and L and G denote Lorentzian and Gaussian components of
size-broadened Voigt function.
>
> Can you already disclose some "exact" results on the CeO2
> sample from D. Louër ? I am sure that he would not have
> distributed that sample without having realized a deep
> characterization.
>
Both Daniel and myself are of opinion that the round robin should be
anonymous. However, if you carefully read some of his recent papers, you may
find the answer there (the paper on CeO2 by his group was published in Chem.
Mater.; I don't have it handy but can find out the exact citation if
needed).
Best regards, Davor
