Binding energies in VASP

[madhu_menon2]

Dear All,
I want to calculate the binding energy of a single metal atom on graphene. I did two runs using VASP; (1) Metal atom on graphene optimization, (2) Metal atom far away from graphene and optimization. I find the energy differences to be too large. What am I doing wrong? Your help would be much appreciated. Thanks. Regards.

[alex]

Hi,

sometimes it's tricky to get the proper state of the atom. I'd suggest to check the DOS. It becomes a lot easier if you calculate the atom all alone. Destroying the symmetry (i.e., don't use a cubic box here) also helps. As well as small smearing. You need(!) to hit the proper state of the single atom.

Cheers,

alex

[SKM]

@madhu
did you resolve the issue?

i think the binding energy (the way i do generally) is to be calculated like below

Step1 : create a system having material A + Material B separated at a distance closer enough to have bonding ----> Relax the system A.B
Step2: Run a static run using CONTCAR from Step1. This Energy A.B
Step3 : consider the CONTCAR obtained from Step1. Then create two separate POSCAR files of A and B, without changing the atoms positions of A and B materials in the whole crystal lattice. Means you need to maintain same lattice dimensions as that of CONTCAR obtained in Step1.
Step4: Run static run for new POSCAR for A. This is Energy A
Step5: Run static run for new POSCAR for B. This is Energy B
Step6: Use formula for binding energy = Energy A.B - [Energy A + Energy B]

Note: In all the calculations one must keep a constant ENCUT value. And preferably EDIFF and EDIFFG also.

Anyone kindly comment if this procedure is wrong or may need modification?
Regards

[madhu_menon2]

Thank you for your suggestion.