Upward Relaxation of Au Surface Atoms in C₆₀-on-Au(111) Geometry Optimization (CP2K v8.0)

[hamid zabihi]

Dear CP2K community,

I am performing a geometry optimization of a C₆₀ molecule adsorbed on an Au(111) slab, using PBE+D3 and DZVP-MOLOPT-SR-GTH basis sets.

The initial configuration had C₆₀ placed ~3 Å above the gold surface. However, after optimization, I noticed that several top-layer Au atoms were pulled significantly upward—toward the C₆₀. This distortion seems nonphysical and may indicate issues with my setup or parameters.

I would be grateful for any advice on:

• Best practices for modeling molecule-surface interactions in CP2K

• Ensuring slab stability and correct adsorption behavior

• Whether this is a known issue with this CP2K version or functional

System details:

• CP2K version: 8.0 (development)

• Functional: PBE with D3 dispersion correction

• Basis sets: DZVP-MOLOPT-SR-GTH

• Slab: 4-layer Au(111), ~15 Å vacuum in y-direction, bottom layers unconstrained

I’ve attached the following files:

• cp2k_c60_au_geoopt.inp: Input file

• out-C60_Au_GEO_OPT.log: Optimization output

• C60_Au_opt-trajectory.xyz-pos-1: Final trajectory

Thank you in advance for your help!

Best regards,
Hamid Zabihi

[hamid zabihi]

[abd el ali el omrani]

Dear Hamid,

I'm not an expert in CP2K, but based on the XYZ file, it looks like the C60 molecule has been adsorbed in a distorted configuration. I'm not sure if you relaxed the slab before performing the adsorption, but if not, I recommend doing so.

Alternatively, you could try fixing all the slab layers initially and allowing only the molecule to relax. Once the system converges, you can gradually release the top few layers of the slab (e.g., the top four layers) from the constraints. This approach may help you avoid the unexpected behavior you're observing.

This is just a simple suggestion based on my understanding—please take it as friendly advice.

--
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+uns...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/87a552d7-53be-4277-90fe-76731650704dn%40googlegroups.com.

[Marcella Iannuzzi]

Hi Hamid, 

Are you sure that the Au coordinates correspond to a Au(111) slab?

I just had a quick look, but the staking looks weird.

Regards

Marcella

[Daniele Passerone]

Hi, there are at least a couple of problems in your input files. 

The first one is the one mentioned by marcella: The termination is a 100, that is going to reconstruct to increase the lateral density (to a triangular close packed layer). 

The other one is that your cell has a lateral dimension 

    &CELL
      ABC 22.9836444855 40.0 22.9836444855
      PERIODIC XZ
    &END CELL

And since the default (see manual
https://manual.cp2k.org/trunk/CP2K_INPUT/FORCE_EVAL/SUBSYS/CELL.html

) is Angstrom, your system has a lot of empty space laterally, and this is seen in the optimization trajectory where the not-fixed atoms are behaving like a cluster and not as a slab.

HTH

Daniele

[hamid zabihi]

Dear all,

Thank you very much for your helpful insights and suggestions.

Based on your comments, I now realize that the slab I initially constructed was incorrectly terminated—more representative of a (100) surface rather than Au(111). Additionally, the excessive lateral cell dimensions likely contributed to nonphysical behavior, effectively causing the slab to behave like a cluster during optimization.

I have now reconstructed the slab with correct Au(111) termination, appropriate lattice constants, and reduced lateral vacuum. I also applied constraints to fix the bottom two layers (32 atoms) to stabilize the slab. However, despite these corrections, I still observe significant restructuring during the geometry optimization of the Au slab alone, even without the C₆₀ molecule.

To facilitate further advice, I will attach the updated input file and trajectory from this latest attempt.

Thank you again for your support.

Best regards,
Hamid Zabihi

--
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+uns...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/cp2k/361844b1-fa13-4309-8cc9-40fd6d48122bn%40googlegroups.com.