[Ifeffit] A feff question

[Scott Calvin]

Hi all,

I know how to handle doping in feff/ifeffit/artemis, but I'm not so 
confident with crystal structures that have a large fraction of 
vacancies. Suppose a site has a SOF of 0.50...and no other atom that 
substitutes. I could run atoms with the site filled, and then run 
feff, and then use Artemis to cut the amplitude of paths involving 
that scatterer in half...but will the feff calculation be anywhere 
near correct? It seems to me like this is a little like running a 6 
angstrom cluster and then taking paths out to 6 angstroms--in that 
case, those last paths aren't very accurate because feff treats those 
atoms as if they are missing some neighbors. In this case, we have 
the opposite problem, with feff treating atoms as having more 
neighbors than they do.

I guess one solution would be to go into the feff.inp file and remove 
roughly the right fraction of the atoms in question. Is that the 
standard way on handling this?

--Scott Calvin
Sarah Lawrence College

P.S. I'll look forward to seeing many of you in a couple of days!