[CP2K-user] convergence problem of Ions in implicit water (SCCS)

[Marcella Iannuzzi]

Hi ...

I don't know the basis set that you are using for Mg, however the PW cutoff 
looks far too low and only 2 valence electron for Mg might be not optimal.
I doubt that the smearing would be useful for such a system, anyway it is 
in general applied with mixing in the reciprocal space.
Alpha 0.1 is not very small, you can easily need much smaller value.

Regards
Marcella


On Saturday, February 27, 2021 at 1:22:23 PM UTC+1 bin... at gmail.com 
wrote:

> Dear CP2K users,
>
> I am trying to relax the Mg2+ solvated ions in the implicit water using 
> the self-consistent continuum solvation (SCCS) model  implemented in the 
> CP2K code. I have tested the parameters for water molecules inside the 
> implicit water according to Oliviero's paper: 
> https://aip.scitation.org/doi/10.1063/1.3676407, and found a slightly 
> more negative solvation energy (-7.77 eV) comparing to the experiment data 
> (-6.3 eV). I assume this error can be accepted and use the aforementioned 
> parameters for the relaxation of ions in implicit water.
>
> However, when I try to calculate the solvated Mg2+/Mg + 6H2O clusters 
> (anions and neutral cluster), I found the SCF can hardly get converged in 
> this case. I tried Broyden mixing method and direct linear mixing method 
> with small mixing alpha (0.1). But this can not make the calculation get 
> converged. I was wondering how can I get the SCF converged (for instance, 
> tuning the values for rho_max/min in the SCCS model)?
>
> I have attached the input file and could anyone have any experience or 
> ideas on this issue help me?
>
> I am looking forward to your reply.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20210227/303cfbc3/attachment.htm>