DFT vs Van de Waals

thuong

Hi everybody,

Can you tell me, in general, why Density Functional Theory does not describe the Van de Waals interaction?

Many thanks,

pallidin

Perhaps this can help. An excerpt from http://arxiv.org/abs/0809.1491

"The method recently developed to include Van der Waals interactions in the Density Functional Theory by using the Maximally-Localized Wannier functions, is improved and extended to the case of atoms and fragments weakly bonded (physisorbed) to metal and semimetal surfaces, thus opening the way to realistic simulations of surface-physics processes, where Van der Waals interactions play a key role..."

kanato

My understanding is that the Van der Waals force is something that is a non-local correlation effect. The exchange-correlation potentials used in DFT are local, so they can't capture this effect. Well, GGA is semi-local but it is not good enough to model a real non-local correlations.

thuong

Thank you very much!