About the output

[huangj3]

Dear esteemed yambo experts,

I am learning to use yambo. I have a few questions about the output and please forgive me if any of these questions are nonsense.

(1) In the o.eps_q* and the o.eel_q* output files, the 2nd/3rd columns (eps/Im[2] and eps/Re[3]) appear to be the results with local field effects whereas the 4th/5th columns (eps0/Im[4] and eps0/Re[5]) are the results without local field effects. Is that right? Please confirm. Additionally, could you please tell me about the 6th/7th columns if they exist (eps`/Im[6] and eps`/Re[7])? From my calculations, they are equal to eps/Im[2] and eps/Re[3].

(2) Also regard to o.eps_q* and o.eel_q*, a comment in 0_driver.F says that "In TDDFT with BS based kernel epso is non-sense". Does that mean for such calculations all the 2nd-4th columns include local field effects?

(3) I noticed that only with $yambo -o c or $yambo -o c -t a, the program gives o.eps_q* and o.eel_q* for various momentum transfer. With other methods, such as $yambo -o b -y h or $yambo -o b -t a -y h or $yambo -b -o b -y h, the program only gives the results for q-->0. Does that mean if I want to calculate plasmon dispersion, I can only use $yambo -o c or $yambo -o c -t a?

(4) Should the macroscopic dielectric constant be read from the real part of the dielectric function at q=0 and w=0, i.e. zero momentum transfer and zero energy transfer?

Thanks a lot for your help and answers.

Sinicerely,

Jingsong Huang
Computer Sciences and Mathematics
Center for Nanophase Materials Sciences
Oak Ridge National Laboratory
Oak Ridge, TN 37831, USA

[Daniele Varsano]

Dear Jingsong Huang,

In the o.eps_q* and the o.eel_q* output files, the 2nd/3rd columns (eps/Im[2] and eps/Re[3]) appear to be the results with local field effects whereas the 4th/5th columns (eps0/Im[4] and eps0/Re[5]) are the results without local field effects. Is that right? Please confirm. .

Confirmed

Additionally, could you please tell me about the 6th/7th columns if they exist (eps`/Im[6] and eps`/Re[7])? From my calculations, they are equal to eps/Im[2] and eps/Re[3].

When diagonalizing with iterative procedures (as Haydock), eps`/Im[6] and eps`/Re[7]) are the results at the N-1 step of the procedure. Once converged, they should be equal to eps0/Im[4] and eps0/Re[5]).

Also regard to o.eps_q* and o.eel_q*, a comment in 0_driver.F says that "In TDDFT with BS based kernel epso is non-sense". Does that mean for such calculations all the 2nd-4th columns include local field effects?

Are you intentioned to use the TDDFT, BS based kernel? This is the kernel developed here. Here eps0 is nonsense as QP energies enter in the definition of \chi_0.

Does that mean if I want to calculate plasmon dispersion, I can only use $yambo -o c or $yambo -o c -t a?

Yes, you can use linear response with LF effects, or TDDT-ALDA. BSE for finite q, it was coded for an old version of the code but it is not released anymore.

Should the macroscopic dielectric constant be read from the real part of the dielectric function at q=0 and w=0, i.e. zero momentum transfer and zero energy transfer?

Yes, exactly.

Hope it helps,

Cheers,

Daniele

[huangj3]

Thank you so much, Daniele. This is very helpful.

Best regards,

Jingsong Huang
Computer Sciences and Mathematics
Center for Nanophase Materials Sciences
Oak Ridge National Laboratory
Oak Ridge, TN 37831, USA