Re: vacancies with FEFF??

[Alex Ankudinov <alex@leonardo.phys.washington.edu>]

Hi Sergio,

The CFAVERAGE card was designed for EXAFS calculations,
and that is the reason why xmu.dat file is not produced.
> I am also interested in the calculation of ELNES taking into account 
> vacancies. I am attaching the input file that I used for it. Basically I 
> removed some atoms from different coordination shells and used the 
> CFAVERAGE card. I found that the xmu.dat file is not generated.
> 
> It is the CFAVERAGE card suitable for this purposes?, or it is only 
> working for EXAFS work?.

 Yes, use CFAVERAGE only for EXAFS.
> 
> Has some of you have some experience with this type of calculations?.

For XANES calculations I would typically run code for inequivalent 
sites, obtain xmu.dat files, and add them with appropriate statistical
weights.  Thus in your case of random vacancies, I would calculate
absorption for Sn atom next to vacancy and for Sn which does not
have vacancy around. And average with weights appropriate for
vacancy concentration.

Also you may want to specify different potential type for Sn, O atoms
depending on how many vacancies around particular atom. I cannot tell
how important this may be without testing.  If this is important,
it will realy complicate the usage of CFAVERAGE for XANES.

Let me know if you have further question about averaging XANES
calculations.

Best wishes
Alexei Ankudinov