Bader (iso)surfaces in VESTA interpretation

[wenbornwilliam]

Dear all,

I'm working on e-phonon calculations using ABINIT for a 2H-TaS₂ structure with 6 atoms per unit cell. I've attached a Gaussian CUBE file (converted from the output density using cut3d). I’m running into some issues with visualizing Bader surfaces in VESTA:

1) When using the ‘all_atom’ option in VESTA, the two isosurfaces for Ta atoms initially resemble Bader surfaces (in attachments). However, when I import the S atoms, their isosurfaces appear as smooth, almost spherical surfaces and only around the S atoms and not the Ta atoms.
2) Adjusting the isosurface level from 0.122142 (default isosurface value, which gives smooth surfaces) down to 0.0001 changes the display to something that resembles the unit cell volume minus the Bader volumes for Ta (also in attachments).

Given the ionic nature of my structure, I wouldn’t expect the S atoms’ charge to extend far into the vacuum, especially not between the layers. Could the issue be related to the default isosurface values in VESTA, or does it have to do fundamentally on how Bader analysis works, or something different?

I would appreciate any confirmation or correction on this!

kind regards,
William W.

[graeme]

I believe this is not related to the Bader code. There is some problem with the interpretation of your cube file. In fact, the cube file looks strange to me - the charge density values are all in a single column instead of multiple columns, as expected in a cube format.