Moderators: Global Moderator, Moderator
VASP team,
I am wondering if it is currently possible to calculate the frequency-dependent dielectric function from a low-scaling GW calculation of a metal?
If this is possible, are there any specific input flags that need to be used? Or would it be run similar to a typical optics calculation. Thanks!
Brian
Hi,
Whatever is the type of system (a semiconductor or a metal), the low-scaling version of GW can be used only for the calculation of the quasi-particle energies and total energy. The reason is that the calculations are exclusively done on the imaginary axis of the frequency (see GW and dielectric matrix). In principle, this would be possible to calculate the dielectric function, with an interpolation method for instance, but this is not implemented.