My Community

Dear VASP maters,

Please help me with a question.
I know G0W0 can provide the dielectric function (inverse) as a function of wavevector q.
Can we do this just with a DFT semilocal functional (for example PBE or SCAN), without GW method?
If so, please let me know how to do it.

Appreciate your help very much!

Best,
Hong

Please try the following INCAR tags:
ALGO = CHI
LSPECTRAL = .FALSE.

Please also find a description of the process here:
wiki/index.php/GW_and_dielectric_matrix ... c_function

The method I have mentioned still involves an RPA calculation.

A colleague pointed out that ff you want to have a cheaper method to calculate epsilon^-1 vs q than please have a look at this paper:
J. Phys. Chem. Lett. 9, 2338-2345 (2018) .