Static linear response function real?

[Franz Fischer]

Hi,

I have a question concerning the linear response function chi, calculated at the Hartree level (i.e. f_xc = 0) and using causal Greens functions.
All static quantities (omega = 0) should in principle be real for all momenta q
So for chi(q,omega), which is the fourier transform of chi(r,t) this should also apply for omega=0 and thus also for epsilon^{-1}_{G,G'}(q, omega=0).

Now I have computed this quantity (epsilon) for a MoS2 monolayer and find some small imaginary part (for G=G' and q .neq. 0).

Why is that? Is the i0+ .neq. 0 in the cheatsheet slide "3) With local fields (RPA-LFE)"?

Thanks in advance!

Best,
Franz

[Daniele Varsano]

Dear Franz,

yes, this is the i0+ .neq. 0
You can reduce it to a very small value by lowering the value in DmRngeXd.

Best,
Daniele

[Franz Fischer]

Thanks Daniele!

Best,
Franz