Re: charge transfer

[Gilles Hug <gilles.hug@onera.fr>]

Hi,

Le lundi, 20 jan 2003, à 16:02 Europe/Paris, Paula a écrit :

> Hello,
>  I need to find out about bonding between atoms. When l run Ldos card, 
> the output files show the charge transfer value changes significantly 
> in value as well as from positive to negative when l alter the number 
> of atoms in the calculation, is this expected or does this indicate an 
> error in my input file?

I guess there should be only one correct charge transfer value which 
should be the one with a large cluster.
However,
 1) you should use the stoechiometry option in the potential cards 
(also it is not clear if you vary the number of atoms for the 
scf-potentials or for the FMS for ldos ? If it's for the first case you 
can get wrong answer if too little number of atoms)
2) I use also the NOHOLE card for that purpose (and just run modules 1 
& 2).

>  Am l correct in thinking that this should indicate the 
> ionicity/covalency of the bonding, e.g. zero charge transfer indicates 
> ionic bonding?

Isn't it that zero charge transfer indicates covalent bonding ?

Best,
Gilles