Hi Dongxiao,
>Thank you very much for the great work of Larixite.
> Can I propose a function for creating input files of FEFF or FDMNES?
>This function may be not realistic to achieve but can be very helpful if we have.
It can already create FEFF input files, either from the web app or from the programming library (poorly documented at the moment) with
```
from larixite import cif2feffinp
# read text of CIF file
cif_text = open(“MyCIFFIle.cif’).read()
feffinp = cif2feffinp(cif_text, absorber=’Pt’, edge=’L3’, cluster_size=7.5)
```
This uses a template that can be modified if you want to change some of the other parameters in the FEFF input file
Generating FDMNES input is high on the list. It’s already started by Mauro Rovezzi, et al at ESRF/ Univ Grenoble / CNRS. There is existing code in Larch for this, but it needs a small amount
of porting work to work with the latest code and the website.
There have also been discussions about generating inputs for Ocean and other XAS/DFT calculations.
> As we know, we always do some doping to alter the physical property of materials.
> The doping are randomly distributed on the lattice site. And the kind of doping atoms can be one, two, three,...
> In the extreme condition, many kinds of doping atoms are put into the
> host and the material becomes high-entropy alloy, which is now widely
> studied by XAFS (HERFD technique, to overcome the multiple edge problem).
> Can you induce the function that can be used to edit the ATOMS part of
> the FEFF input file to introduce one or multiple doping atoms in the
> lattice and these doping can be randomly distributed?
Chris gave a nice answer for this. I’ll add that many of the structures from AMCSD have sites with mixed occupancy. These are handled by randomly selecting based on the stated occupancy.
If you edit a CIF file to have partial occupancy (using one of the existing ones as a guide, though editing CIF by hand is not for the feint of heart), you can upload and use that too. I’m
sure that could also be done completely with Pymatgen Structures, though I’ve not ever done that myself.
And, as Chris said, it is definitely OK to edit the resulting feff.inp file yourself, say to change the absorbing atom to a dopant species.
--Matt