negative volumes

[afloris]

Dear users,
I am a new user of the Bader code. I have produced with Quantum-Espresso a .cub file containing the density of charge of a slab containing a molecular adlayer adsorbed on
three layers of metallic surface. Vacuum is present in the unit cell.
I am very interested in calculating the total charge associated to one particular adsorbed molecule, in presence of the surface and a Cs atom. When I run

bader -c bader 4tcnq_cs_ag100_charge.cub

I get an ACF.dat file containing *negative* atomic volumes, a negative number of electrons and clearly wrong charges associated to atoms
(e.g. -49.8 associated to the Cs atom, that should have essentially a zero valence charge associated). The ACF.dat is

# X Y Z CHARGE MIN DIST ATOMIC VOL
--------------------------------------------------------------------------------
1 32.7898 -10.9296 11.3384 -15.9763 0.5489 -452.8257
2 5.4645 5.4675 11.3384 -11.2087 0.5427 -395.6654
3 10.9296 5.4675 11.3384 -8.9665 0.5475 -352.4347
4 21.8594 -5.4645 11.3384 -20.5339 0.5424 -512.5974

.....
--------------------------------------------------------------------------------
VACUUM CHARGE: 0.0000
VACUUM VOLUME: 0.0000
NUMBER OF ELECTRONS: -1774.0010

The same analysis, but done with a free standing adlayer (i.e. without surface),
produces good results, giving me a correct charge transfer between the Cs atom and a strongly electronegative molecule.

My question is: what I am doing wrong? Are Bader analyses in presence of a metallic surface, with charge around in the adlayer/surface space, particularly "difficult" and require special care?
Can someone give an advice?
I am sorry if this issue has been raised already.
Thanks,

Andrea
--
Dr Andrea Floris
Research Associate
King's College London
Strand, London WC2R 2LS
United Kingdom
Emails: andrea.floris@kcl.ac.uk, an.floris@gmail.com

[graeme]

It looks like the file is not being read correctly. If you can upload an example, we'll try to debug the problem.

[afloris]

Thank you very much.
I uploaded the cub file plus the files produced by bader, in a tgz file.
cheers
Andrea

[afloris]

Hi,
just to upgrade, I have noticed that the volumes are negative whenever I consider an infinite system. But if I have, for example, a finite structure in a box, then the results are fine.
Perhaps someone could give me some hints about why this happens in infinite systems?
Thanks
Andrea