Drude frequency in LRC

[Stephan]

Hello,

is there any possibility to insert Drude-frequency in the LRC-runlevel (optical response).
I tried

DrudeWXd= ( 0.80590 , 0.07439 ) eV # [Xd] Drude plasmon

and

DrudeWXs= ( 0.80590 , 0.07439 ) eV # [Xd] Drude plasmon

Both inputs were not recognized.

Thanks and Regards

Stephan

[Daniele Varsano]

Dear Stephan,

DrudeWXd= ( 0.80590 , 0.07439 ) eV # [Xd] Drude plasmon

should be the right way to include the Drude contribution in LRC.
Strange it is not working, of course if your system is a metal.
Can you post report/input/output? and we will have a look.

Which version of the code are you using?
There was a problem some time ago, but I think it was solved.
Just to be sure:

You can also try to add this variable in your input:

Code: Select all

DrudeWBS=( 0.80590 , 0.07439 ) eV

please report if it works in this case.
If this is the case there is a small correction is needed in the code.

Best,
Daniele

[Stephan]

Hello Daniele Versano,

I meanwhile recognized, that DrudeWXd was read by the program as you will see in the attacher report and output file, but there is no Drde contribution in the resulting plot of Im(eps).
DrudeWBS did not work.

I use yambo-3.4.1-rev61

I have a further question:
In the SiF tutorial there is
NGsBlkXd=1 when LRC is used.
Is this senseful or shall I choose the same value as in the RPA calculation?

Thanks and regards

StephanLudwig

[Daniele Varsano]

Dear Stephan
I had a look to your report file and it looks that your Q grid it is not uniform:

Code: Select all

[RD./SAVE//db.kindx]----------------------------------------
  Polarization last K   : 244
  QP states             :   1  244
  X grid is uniform     :no
  BS scattering         :yes
 - S/N 008162 --------------------------- v.03.04.01 r.3187 -

 IBZ Q-points : 1
 BZ  Q-points : 1

 Q [00001] : 0.00      0.00      0.00     (iku) * weight  1.000000

 Q [00001] : 0.00      0.00      0.00     (cc ) * weight  1.000000

and this can cause problems. Check if you have non symmorphic symmetries in quantum espresso. You can avoid them by using the kyeword:
force_symmorphic=.true.

About the NGsBlkXd, yes you need to converge them, the tutorials are meant as example and they are not converged calculations.
About the Drude, we will check if there is something wrong.

Best,
Daniele

[Stephan]

Hi Daniele,

thank you. I will repeat the input calculation with a uniform grid..

Thanks and Reagards

Stephan

[Stephan]

Hi Daniele,
I meanwhile use uniform kgrids, and I still have the problem that there does not occur any Drude-contributions in the LRC-calculation although the report proclaims that it was recognized. In the attachment you'll fin the report and the eps-file.
It would be great if you could help.

Thanks and regards S

Stephan

[Daniele Varsano]

Dear Stephan,
I do not know the detail of your system, but why you say that the Drude term is not included?
I can see divergences in the non-interacting epsilon and a peak at 0.66eV in the eps2.
Best,

Daniele

[Stephan]

Dear Daniele,
Do you mean that I have to bild the sum of the non-interacting eps0/im and the LRC result eps/Im to obtain the full imaginary part of the dielectric function on the LRC runlevel?

Thanks and regards

Stephan

[Daniele Varsano]

No, you do not need to sum the two responses.
Anyway you are right, the spectrum should have the diverging behaviour also in the TDDFT response.
We need to check .
Best,

Daniele

[Stephan]

Thank you very much

Best regards

Stephan