Energy of an isolated atom for adsorption on a cluster

[vasp]

I am trying to calculate the adsorption of carbon, therefore, need to determine the energy of an isolated carbon. I followed the suggestion of VASP to construct the INCAR file:

ISTART=0
ISMEAR=0
SIGMA=0.1
AMIX=0.2
BMIX=.0001
ENCUT = 400 (got from POTCAR)

and calculated for GAMMA point,

and for POSCAR used the lattice vectors for carbon, as suggested by the VASP document, which is 8 Angstroms, which gives more than five times the size of carbon.

I tried both LDA and GGA PAW pseudopotentials. I get a negative energy of .07, which is very, very different from the energy in the POTCAR which is 146.6 ev. I expected the energy to be negative as well since it is a bound system.

What am I doing wrong? I have tried different sigmas but that the energy remained more or less the same.

atefera

[superyoyo]

I think you did not do wrong, just try to understand what is the negative energy of .07 (or .7 ). It will be definitely different from what you see in POTCAR because this is energy which is referenced with that energy in POTCAR.

You just need to check whether the convergence of energy respect to cell size, energy cutoff, number of k-points. And you may check the effect of spin polarization.
<span class='smallblacktext'>[ Edited Thu Dec 31 2009, 07:29AM ]</span>

[lcyin]

I think what you got is reasonable
I got the total energy (E0= -.13098327E+01) of an isolated C aton by using PAW-PBE
Please check the VASP manu, Section 9.5: Determining the groundstate energy of atoms