I tried to check my homewritten fit programme, which uses the
Thompson-Cox-Hastings function, against a simulated diagramme created with
RIETAN. Refinements even with the known parameters failed due to severe
differences in the profile shape!
Question:
TCH give in their paper (J. Appl. Cryst. 1987, 20:79-83) the formula:
I (delta2theta) = I ( 2 etha / pi Gamma ) / (1 + 4 (delta2theta/Gamma)**2)
+ term for gaussian profile
GSAS (1994, page 129) manual gives:
L (t) = Gamma / 2pi (1/( (gamma/2)**2 + t**2))
Manual RIetan gives the same formula as TCH except from a different writing
of the first term, which looks like this:
2
Etha ------ ( 1 + 4 ( ....) )
pi Hk
Mathematically I would interpret the formula by TCH like this: 2 * Etha *
Gamma / Pi, which would be identical to the prefactor in GSAS, but the
second term is different. I assume, however, that TCH ment this: 2 * Etha /
( Pi * Gamma ), did they ?
I used in my programme the TCH with the latter interpretation. Izumi,
however used the GSAS first term and the TCH second one, but still the GSAS
and the TCH formulae differ in their second term.
For me it is clear now that the profile shapes are of course different due
to multiplying or dividing by Gamma. But who is right, who is wrong ??
Thanks in advance und Dankeschoen
Helge