> Dear Andrei,
>
>> However, the essence of my previous message:
>> you do not calculate any SINGLE atom,
>> you always keep TWO atoms in the unit cell,
>> but vary the interatomic distance from "very close"
>> to "infinity" (as the basis atoms of two atoms do not overlap),
>> then you'll be clean in what regards the BSSE.
>> 1. Allow the unit cell big enough for that.
>> 2. Do not forget the magnetism.
>>
> This solves some problems, but BSSE is not one of them.
> To get a BSSE-free PES,
> one would have to calculate counterpoise correction at every point. My
> experience with SIESTA is that with the default DZP basis set, BSSE can be
> fairly large.
>
> Cheers,
>
>   Herbert

Thank you Herbert -

yes, my statement might have been too optimistically formulated.

Andrei


Responder a