[CP2K-user] [CP2K:19763] Ru slab optimization

[Lucas Lodeiro]

Hi,

recently I worked with the same system without problems, so I guess the
problems are in the input. As I see, there are two main errors. First an
EPS_SCF 1.0E-3 it is completely useless, your wavefunction neither the
energy are converged at all, so the computed forces are completely wrong.
As a general rule I use an EPS_SCF value equal to the square root of
EPS_DEFAULT, and usually I reduce the EPS_DEFAULT value from the default
one (1.0E-10) to 1.0E-12.
Secondly, your cell is ill defined. Ru(0001) is hexagonal shaped so there
is a 120° between A and B vectors, and your input corresponds to a square
one.

Also there are some issues in your computation: You use UKS in a non
magnetic material, which enhances the computational time a lot, use RKS
instead. The wavefunction guess is atomic for all geometries, it is better
to use the converged one from the last geometry computed to save time.
As advice, maybe you need to use more layers, I use 8 layers and
everything works pretty well. Also for converged results using a bigger
surface, you used a 4x4 surface, in my benchmark I found 6x6 is better. But
obviously this enhance the computation time... Your vacuum (around 20
Angstroem) is enough for non charged and polarized surfaces.

Regards - Lucas Lodeiro

El mar, 16 ene 2024 a las 17:52, Lakshmi Anil (<lakshmianil967 at gmail.com>)
escribió:

> Dear all,
> I was trying to do geometry optimization of the 4x4 slab model of
> Ru(0001). The job was completed within 3 hours. Unfortunately, the slab
> structure distorts. One row of atoms from the top and bottom layers is
> moved to the middle layers. I also tried to do optimization by increasing
> the vacuum but the result was in the same way. Could anyone please help me
> how to resolve this problem? I am also attaching the input and output files
> of the optimization.
>
> --
> You received this message because you are subscribed to the Google Groups
> "cp2k" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to cp2k+unsubscribe at googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/cp2k/5d803f4b-9b13-44ea-81c8-9c0d4cc9814bn%40googlegroups.com
> <https://groups.google.com/d/msgid/cp2k/5d803f4b-9b13-44ea-81c8-9c0d4cc9814bn%40googlegroups.com?utm_medium=email&utm_source=footer>
> .
>

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAOFT4PJY3WN0tEGjTLu7tAmgsmvByo6gftogjS0bjvrKnWnt2w%40mail.gmail.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20240117/a9a8fc59/attachment.htm>