Question on the Total Energy and the Electronic DOS

[Modey3]

Hello,

I was under the impression that VASP just calculated the electronic energies for the supercell in question: the total energy is the integral with respect to Energy of the DOS*Energy from 0 to infinity. My problem is that I'm comparing the DOS for a magnetic Fe slab to it's bulk counterpart. The DOS for the slab shifts to a lower energy with respect to its bulk counterpart. This coincides with a shift in the Fermi Energy to a lower energy. The problem is that the energy calculated by VASP gives a higher energy for the slab than the bulk, which makes sense. However, if I numerically integrate the DOS*Energy with respect to energy I get a lower energy for the slab than the bulk. In regards to the total energy that VASP calculates am I leaving anything out?

modey3

[admin]

please remind that the zero level of energy is NOT defined in any system with 3D periodic boundary conditions to within a constant offset, therefore the KS levels do not refer to a common 0!
In a slab calculation, the levels can be taken with respect to the constant vacuum level (if the vacuum is thick enough), but in the 3D bulk calculation this level is not known a priori.

[Modey3]

Thanks. I partially answered the question in another thread. Is it correct in saying that the Total Energy component, EBANDS, in the OUTCAR file corresponds to the integration of the DOS*Energy from 0 to infinity? If thats the case the EBANDS component from slab and bulk configurations cannot be compared.

Best Regards

modey3

[Modey3]

<span class='smallblacktext'>[ Edited Thu Jul 16 2009, 08:46PM ]</span>