Accuracy of total energies of atoms and molecules
Posted: Tue Jul 24, 2012 1:56 am
Hi,
I would like to hear your opinions about the accuracy of total energies of free atoms and molecules calculated using VASP.
We usually compare the atomization energy of a molecule like H2 or N2 to the experimental one, and usually see an error. So, where do you think the main part of the error is associated to? Description of the atom or description of the molecule in VASP, or both? My personal intuition was molecules are more prone to errors, since there is a particular electron configuration that DFT needs to catch during bonding, while atoms are fine as long as the right electronic configuration is obtained. Well planewaves are not so useful for the purpose of investigating molecules but let's keep that aside.
I would appreciate if anyone more knowledgeable about this shares his/her point of view.
Cheers.
I would like to hear your opinions about the accuracy of total energies of free atoms and molecules calculated using VASP.
We usually compare the atomization energy of a molecule like H2 or N2 to the experimental one, and usually see an error. So, where do you think the main part of the error is associated to? Description of the atom or description of the molecule in VASP, or both? My personal intuition was molecules are more prone to errors, since there is a particular electron configuration that DFT needs to catch during bonding, while atoms are fine as long as the right electronic configuration is obtained. Well planewaves are not so useful for the purpose of investigating molecules but let's keep that aside.
I would appreciate if anyone more knowledgeable about this shares his/her point of view.
Cheers.