vanderWaals interactions (IVDW) and consistency

[alexandrosvasi]

Dear Vasp Community,

I have a question regarding the nature of the IVDW corrections and how to be consistent when calculating properties.
I am modeling graphene sheets (slabs) with deposited alkali-ions on them and I am using IVDW = 11 to account for VdW.

To find the adsorption energy (or voltage) with respect to the metal (using as reference the bulk and not an isolated atom in a vacuum).
We need to have the Eg (energy of empty graphene sheet), Em (energy of 1 atom in the bulk metal), and Egm (energy of metal and graphene).

In order to be consistent, do I have to apply the same correction to all three calculations?
For example, the Eg and Egm need the correction but the Em does not. (based on the size and geometry)
Further, if I apply the correction to the metal I get a big shift in energy affecting the results quite significantly (while physically it shouldn't).
Example: 1 Na-metal atom without IDVW: -1.307eV and with IDVW=11: -1.469eV

Checking a lot of literature on adsorption and reproducing simple calculations of 1 atom depositions on graphene (trying to replicate their numbers) revealed that they most probably do not apply the corrections to the metal but apply it only to the Eg and Egm.

Do you have any advice in terms of consistency and what is the correct approach?
Thank you very much for your valuable help
Alexandros

[henrique_miranda]

I am moving this question to the "From Users for users" forum as it is not a question directly related to VASP but more a physics question.

I don't have a direct answer to your question.

Only a suggestion:
It seems that you are considering computing the total energies of the metal (Em) without VdW corrections because you obtain very different energies whether you set IVDW=11 or not.
Have you tried computing the same bulk metal with different IVDW settings (for example IVDW=2) and see if the energy difference is still so large?

[alexandrosvasi]

Thank you very much for your answer,
I will check

[alexandrosvasi]

well its type of corrections seems to vary the energy of Na and Li metal. For example for Na (I used the Na_pv)
pbe lda vdw12 vdw11 optpbe pbe88 1 Na atom
-1,313895 -1,45755765 -1,4758907 -1,4723043 0,9385456 0,87709605 in metal