[CP2K-user] Geometry Optimization for Copper

[Matt W]

Your scf convergence criteria is not good enough, you should have EPS_SCF 
much smaller, maybe 10-6 to ensure energies are converged to at least micro 
hartree level.

I am not at all sure that the electric field and poisson solver work with 
k-points. I also don't have experience with the implicit psolver to comment 
on that setup. I'd remove that and check you can optimize a slab with 
periodic boundaries first then tackle the more complicated problem 
afterwards.

Matt

On Friday, June 19, 2020 at 8:20:26 AM UTC+1, yulei wang wrote:
>
> Hello all, 
>
> I am fairly new to cp2k so I believe this problem is easily resolvable.  I 
> am doing structural optimization of periodic copper surface under the 
> action of electric field,  and I fixed 3 atomic layers below ( input file 
> is cu.inp and date.coord).  I get the wrong results as the position of atom 
> 7 is too high compared to other atoms in the same layer (the result is in 
> cu-pos-1.xyz). It is undoubtedly unreasonable, and then I tried to do 
> the  geometry optimization without the action of electric field, and the 
> result is just the same. Any ideas?
>
> Thank you for your help or advice. 
>
> Best Regards, 
> Yulei Wang
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20200619/1225f3ce/attachment.htm>