set rwigs one by one?

Message

phtobias · #1 Post by **phtobias** » Wed Jun 28, 2006 2:46 am

Dear vasp users,

Would it be possible to set the rwigs for each atom in the supercell?

For example, I have a supercell of 24*(TiO2). Can I set the line below in the INCAR?

rwigs = R1 R2 ... R24 R25 R26 ... R72

NB: R1~R24 are for Ti and R25~R72 are for O.

If yes, can I break the long line of rwigs into several lines with a `\` at the end of each line?

Thanks for the help.
yi-ren

forsdan · #2 Post by **forsdan** » Wed Jun 28, 2006 6:42 am

If you are about to choose all rwigs values equal for each species, can't you just write

RWIGS = 24*R_Ti 48*R_O ?

I know it's possible to do so for the MAGMOM-tag. I suppose it must work for the RWIGS-tag as well. Then you don't have to worry about line breaks.

Best regards
Dan Fors

#3 Post by **admin** » Wed Jun 28, 2006 9:48 am

by default, RWIGS is read for each atom type (as summed up in POSCAR), not for each atom explicitely. I wonder if you separated all Ti and O in POSCAR and POTCAR?
if so: yes, it is possible to read n multiple entires X as n*X in INCAR

phtobias · #4 Post by **phtobias** » Thu Jun 29, 2006 5:40 am

Head admin:
Thanks for the hint, I will try to separate POSCAR/POTCAR to see if i can assign two differnt rwigs to the same type of atoms.

For Dan:
The reason to have separate rwigs for each atom is to account for the the different geometry surounding each atom on the TiO2 surface. It might not be that criticial overall. But my goal is to get a meaningful partial charge anaylsis and want to exclude as many uncertainlties as possible. (For a site specific partial charge analysis, see, e.g, table 7 in J. Phys. Chem. B 2004, 108, 7844-7853)

yi-ren

#5 Post by **admin** » Tue Jul 11, 2006 9:38 am

if you want to do a partial charge analysis, please let me note one thing: in VASP, the partial local quantites are NOT symmetrized a posteriori, therefore please don't be confused if the partial charges (and LDOS) of symmetry-equivalent atoms differ slightly.