atomic chrage

[npatra2]

Hi all,

I was trying to calculate atomic charges from mulliken charges calculated by
siesta. I got some result which does not make sense ,like in h20 -- H is
negative and O is positive. Then I fond in the mailing list it depend on
basis set (in this case I used DZP for both atom).

I was trying to calculate the atomic charge by bader charge analysis. I have
the charge file rec.RHO and rec.DRHO(it says it can read the gaussian cube
file).

When I ran the bader like following,

./bader rec.RHO

or ./bader rec.DRHO

I have the following error :

GRID BASED BADER ANALYSIS (v0.26b 12/07/09)

OPEN ... rec.DRHO
GAUSSIAN-STYLE INPUT FILE
forrtl: severe (59): list-directed I/O syntax error, unit 100, file
/home/niladri/parallel_siesta/h2o/rec.DRHO
Image PC Routine Line
Source
bader 0000000000498282 Unknown Unknown Unknown
bader 0000000000497482 Unknown Unknown Unknown
bader 000000000045FE86 Unknown Unknown Unknown
bader 0000000000428311 Unknown Unknown Unknown
bader 0000000000427BFE Unknown Unknown Unknown
bader 0000000000446E07 Unknown Unknown Unknown
bader 0000000000445AD3 Unknown Unknown Unknown
bader 000000000040C72E Unknown Unknown Unknown
bader 000000000040F044 Unknown Unknown Unknown
bader 0000000000400396 Unknown Unknown Unknown
bader 00000000004002FE Unknown Unknown Unknown
bader 00000000004A2850 Unknown Unknown Unknown
bader 00000000004001E9 Unknown Unknown Unknown




And also I did not get any output file(ACF.dat, BCF.dat, AtomVolumes.dat).

Could anyone help me please?

Thanks,

Niladri Patra

[Wenjie]

Would you please send the cube file that has problem to twj916@mail.utexas.edu? I'll take a look at it.

Wenjie

[cyberfrog]

npatra2, since siesta uses pseudopotentials in order to replace core electrons, sometimes it's quite a problem to run bader. Best way to obtain consistent resutls from Bader is to allow Bader to see charge's peak on each atoms, and best way to do this it's to have the total charge as sum of core+valence electrons. RHO files in siesta instead take into account of valence electrons only. You can take a look on Vozny's blog about this topic:

http://voznyy.elinity.com/blog/2008/01/ ... ta/#more-3

you can try to run bader forgetting the core electrons to see if bader see all the atoms of your system (a charge's peak on each of them).
to do this you have to convert a *.RHO into a *.cube using the modified util rho2cube found in the vozny's blog.

pay attention: I mean RHO and not RHO.cube