Happy New Year All,
I'm trying to compute Bader charges for a paramagnetic molecule (NO2)
adsorbed on MoS2. Normally one would use the PAW wavefunction of the relaxed
system as input to pp.x with plot_num=17 to obtain the electron density, which
would then be processed with the Henkelman code
(http://theory.cm.utexas.edu/henkelman/code/bader/) to get the Bader charges.
Unfortunately, pp.x with plot_num=17 doesn't work for spin-unrestricted systems
in Quantum Espresso version 6.4.1. The latest release notes say that this has
been fixed in vers. 6.5, but I don't yet have access to this most-recent
version. I also tried plot_num=21 in vers. 6.4.1, and that doesn't work either.
Out of desperation I tried plot_num=0, which according to the pp.x
input write-up produces "electron (pseudo-)charge density". That does work, and
the results appear reasonable. The right total number of electrons is obtained,
and all the atom charges look OK (Mo = +1.18, S = -0.59, N = +0.70, O = -0.37).
The MoS2 is essentially charge-neutral and a very small negative charge appears
on the electron-acceptor NO2.
With that as background, my question is whether or not the use of
the plot_num=0 density in this way is in fact valid. I'm not sure what the term
"electron (pseudo-)charge density" actually means. Another disturbing point is
that the atom volumes found by the Bader code are slightly different for S
atoms that are related by symmetry (i.e., by reflection in a mirror plane),
even in the absence of the adsorbed molecule. In one S layer the volumes are in
the range of 833-841 bohr^3 while in the symmetrically-equivalent layer the
range is 821-828 bohr^3. I should mention that I'm using a dipole correction
layer (3 Angstroms wide) in the middle of the vacuum space (26 Angstroms wide)
because of the finite molecular dipole. I'm assuming that this won't affect the
Bader-charge calculation.
Any expert advice would be much appreciated.
Best Wishes,
Vic Bermudez
(US Naval Research Lab. - retired)
_______________________________________________
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
users mailing list [email protected]
https://lists.quantum-espresso.org/mailman/listinfo/users
