how to get the information of dielectric function

[pyadav]

Dear Daniele,

Thanks for the reply.

I actually did both the calculation (for the chi_after_g0w0 case with and without QP corrections) here

chi_g0w0.zip

are the input and output files for these two cases.

The problem here is that these results (chi_after_g0w0 are very different from the chi_after_DFT (you can see from the chi data plots) at the RPA level even without adding the QP corrections.)

I would really be grateful for your help in understanding this issue.

Thanks,
Pushpendra

[Daniele Varsano]

Dear Pushpendra,
sorry but I'm confused. In all the files you sent me.
I can see two RPA calculation, one using KS energies and one using QP energies, the two differs as they should.
When you insert in your input the keyword:

Code: Select all

XfnQPdb= "E < SAVE/ndb.QP"

you are including the QP corrections.

Best,
Daniele

[pyadav]

Dear Daniele,

I'm really sorry for the confusion created here.

Here I'm now attaching the chi data (with corresponding input and report files) for two cases
1. chi after DFT (No G0W0 calculation done here) without QP corrections
2. chi after G0W0 calculation without adding the QP corrections

chi_without_qp.zip

I think these two chi should be the same as I have not included

Code: Select all

XfnQPdb= "E < SAVE/ndb.QP"

in any of these.

Thanks,
Pushpendra

[Daniele Varsano]

Dear Pushendra,
the two runs differ in the way the dipoles are computed:

Code: Select all

<   [r,Vnl] included                                 : yes
---
>   [r,Vnl] included                                 : no

I suggest you to delete the dipoles database and recalculate them from scratch.

Best,
Daniele

[pyadav]

Dear Daniele,

Thank you for the quick reply. Yes, the difference is found between the two cases.
The case (chi_after_DFT) where

Code: Select all

<   [r,Vnl] included                                 : yes

gives the result that agrees with the literature
but again when I'm doing chi_after_g0w0
using the input commands

Code: Select all

yambo  -o c -k hartree -V qp

for chi at the RPA level by setting

Code: Select all

XfnQPdb= "none"

it is again seen that

Code: Select all

<   [r,Vnl] included                                 : no

Can you please tell me how do I include this

Code: Select all

<   [r,Vnl] included                                 : yes

for the chi_after_g0w0 case to get the dielectric function in RPA without QP corrections?

Thanking you,
Pushpendra

[Daniele Varsano]

Dear Pushendra,

let me clarify, if you do not include the QP correction, the fact you did first a GW calculation is not seen by yambo.
The [r,Vnl] is included by default when calculating the dipole unless you remove or rename the ns.kb_pp file from the SAVE directory.

In your case, I can see that the dipoles are not recalculated, but read from a previous run where the [r,Vnl] was missing.

Code: Select all

 [RD./SAVE//ndb.dipoles]-

Here RD stands from READ.

Se essentially you need to
*verify you have the ns.kb_pp file in your directory
*delete your ndb.dipoles* files
*redo the calculation

I hope it is clear,

Best,

Daniele

[pyadav]

Dear Daniele,

Yes, the ns.kb_pp file was renamed earlier to fast the g0w0 calculation. Now after renaming it back and removing the ndb.dipoles* worked.
Thank you so much for the help and clarifications.

Wishes,
Pushpendra

[Daniele Varsano]

Dear Pushpendra,
if the non local part of the pseudopotential has such a large impact on the response function, it is safe to keep it also for the GW calculation.
Best,
Daniele

[pyadav]

Dear Daniele,

Yes, I get the point, I'll keep in mind this for any further calculations.

Thanks a lot.
Pushpendra