Optical spectrum at the RPA level starting from DFT+U

[mapo]

Dear YAMBO team,

I am working on a system for which I need to include the Hubbard parameter
to correctly describe 3d electrons, otherwise I obtain a metallic behaviour.

On top of the DFT+U calculation, I would like to obtain the optical spectrum
of the system at the RPA level. Basically, my question is: can one calculate
directly the optical spectrum of a system using the KS eigenvalues and
eigenfunctions obtained from a DFT+U calculation?

Many thanks in advance,

Miguel A. Perez
University of Oxford
United Kingdom

[Davide Sangalli]

Dear Miguel,
it might work. However the dipoles will miss the contribution from the commutator of x with the non local part of the Vxc functional, i.e. the U term in this case.
It means the intensity of your peaks might be wrong. I've no idea how much the effect could be.

Best,
D.