[CP2K-user] 2D PBC with non-orthogonal cell

[Ondrej Krejci]

Dear CP2K users and developers,

I am trying to calculate systems with only 2D periodic boundary conditions, 
because of necessity to apply electric field. As I've found out from 
attempts of my predecessors, one of the way how to do it, is to use 
"wavelet" poisson solver with periodic "XZ" conditions. Unfortunately, I've 
found out, that this seem to work only for orthogonal cells. When I am 
using the "wavelet" with non-orthogonal cells, I get completely non-sense 
geometries during optimisation. I attach a difference between standard 
periodic and wavelet run (input files, output file and output geometries).
Please is there some way around it? - 2D periodic boundary conditions with 
non-orthogonal cell?
If not - what kind of setup/keywords are available for running kind big 
systems (2000 atoms). I am trying to run the orthogonal cell at the moment, 
however it seems that with the similar setup I am limited to ~ 5000 cores 
and also restart with usage of wfn files crushes.
Thank you very much for your answer!
Kind regards,
Ondrej

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20210705/9f059ed7/attachment.htm>