free/2d boundary, charge, and poisson solvers

[sungen... at gmail.com]

Hello everyone, 

        I met problems on calculations of charged systems by cp2k.

        I am familiar with VASP and GAUSSIAN.  GAUSSIAN could take care of 
charged systems naturally for the atomic basis set.  But it's not easy to 
study charged systems In VASP for the plane wave basis set. Taking care of 
charged systems in VASP needs much corrections to give reasonable results, 
and the energy for charged systems is useless sometimes
. 
        I notice that, CP2K supplies several POISSON solvers to handle this 
problem. But I am not understand the usage of the poisson solvers.
        My first question is,  though I set up a  free boundary in cp2k 
(PERIODIC NONE in DFT ),   I still have to create a box for my 
calculations. This box (unit cell) is only used for representing grids for 
the plane waves and gaussians ? Am I right ? So the energy of a isolated 
systems  from cp2k should be comparable with that from  GAUSSIAN , 
especially for charged systems ( I mean, are they representing the same 
physical identity ?)

        Secondly,  what' the general rule to set the parameters for charged 
free boundary and 2d systems in cp2k ? the only thing for consideration is 
poisson solvers ? How could I  choice different solvers ?

      Thank you for any suggestions !

Geng Sun
        
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20150128/28de6ee7/attachment.htm>