On Nov 22, 2007, at 11:30 AM, Telepeni Irvin wrote:
Thanks for your reply, that must be the explanation, but I am not
quite sure of how to filter these sites from those of interest
(since I dont know where they are)? I could just try all of them
but I am sure that's not the ideal method.
When you compute a Fourier map in GSAS, it is best to compute it over
only the minimum volume needed to allow computation of the entire
unit cell (aka asymmetric unit). If you do that FORSRH will not find
multiple symmetry-related peaks.
My experience is that only the first few Fourier peaks are meaningful
in a powder difference Fourier, so I will put trial atoms at the
first few positions found in a map -- if they make chemical sense
(you might want to save then and then use DISAGL to look at distances
and angles to test that). My first step after including them is to
refine occupancies (with fixed positions and Uisos) of the trial
atoms to find if putting scattering density at the site actually
improves the model. If the occupancy refines very small, then it is
hard to believe in the site.
Irvin, your previous questions implied one of two things might be
true: that FORSRH is not giving you peaks where the map is showing
positive density or that you do not see peaks in the map when plotted
in FOX corresponding to the peak locations found in FORSRH. If either
is true, you may have found a bug that needs to be documented if it
will be tracked down and fixed. With luck, the software is fine.
Brian
