HF convergence

[fatimazahra]

Hi all
I want study the convergence of the Hartree-Fock self energy for Bulk Silicon system, please see the attached files
I ran
1) P2y
2) yambo
3) yambo -x
4) yambo -F INPUTS/01HF.in -J HF_XRy
I repeat points 4) for EXXRLvcs=3,6,7 and 15 Ry
But, I found the same value of direct gap and indirect gap for all value EXXRLvcs

4.4.4.tar.gz

[Daniele Varsano]

Dear Fatima,
something is going wrong with your runs:
From what I can view from the report you are not running HF calculations but just repeating the setup.
If you look at the end of the report, the input file is mirrored, and you have:

Code: Select all

 .-Input file : (none)

So you are still looking at the DFT band gap. I have no idea why the input it is not correctly found. In any case, in order to make things simpler just rename your input file 01HF.in in yambo.in and run the executable without the F option: yambo will look for the yambo.in as input. Next, please note that if you want to increase the EXXRLvcs 3,6,7 and 15 Ry you have to enter the units in the input: i.e.

Code: Select all

EXXRLvcs= 15         Ha

instead of

Code: Select all

EXXRLvcs= 15         RL

RL means Reciprocal Lattice and in this way you are including only 15 G vectors.
Hope it helps,
Daniele

[fatimazahra]

Dr. Daniele
Thank you very much for your help
I did all the changes as you told me but I get the same problems,
you find in attachment the new files

SIC.tar.gz

[Daniele Varsano]

Dear Fatima,
please note that the problem it is not the same: Now you are calculating HF matrix element. Have a deep look at the report and you can see

Code: Select all

XC HF and DFT [eV] @ K [1] (iku): 0.00      0.00      0.00
 <1|HF|1> = -16.98063 -.3073E-7 <1|DFT|1> = -10.50084 -.5556E-8
 <2|HF|2> = -12.04911 0.2742E-8 <2|DFT|2> = -11.32918 0.1649E-7
................

so you have for each k-point and band you asked the HF and Vxc matrix elements from where you can calculate the new gap. The two number refers to real and imaginary part.

Now the two calculations are the same, but from the mirror of the input file it looks that you run the same input having 1 Ha in EXXRLvcs.

Code: Select all

EXXRLvcs= 1            Ha    # [XX] Exchange RL components

Best,

Daniele

[fatimazahra]

Dr Daniele
I ran the calculation again, with different values of EXXRLvcs, at the raport, i got same resultat, although different value of EXXRLvcs (in the raport)
please see new attechement.

resultats.zip

[Daniele Varsano]

Dear Fatima,
as I told you in the last post, please read the report carefully.
The results are different:

If you look at:

Code: Select all

 Indirect Gaps      [ev]: 0.636567  4.625871
  Direct Gaps        [ev]: 2.576077  8.242179

of course they are the same. As reported the come after reading the s.db1 database which contain the DFT electronic structure and it is the same for both calculations.

Next if you look at the HF corrections:
you have for instance the first band first k-point:

Code: Select all

<1|HF|1> = -14.81218

for the first run and:

Code: Select all

<1|HF|1> = -17.08972

for the second run.
If you are interested only in the gap and not in the entire band structure,you can calculate only for the points you are interested in (check the k-point and bands corresponding to the gap for the DFT calculation, it is also written in the report in this section):

Code: Select all

 [02.05] Energies [ev] & Occupations

and select them in the input using this variable:

Code: Select all

%QPkrange                    # [GW] QP generalized Kpoint/Band indices
  1|  8|  4| 5|
%

here for example is meant for calculating 2 bands (index 4 and 5 ) for 8 kpoints.
next your non self-consisteng QP gap will be:
E_DFT+<HF> - <Vxc>
If you instead perform a GW run, you will have these results collected in a more practical output file, named o.qp

Hope it helps,
Daniele

[fatimazahra]

Dr Daniele
Thank you very much, you're really helped me, thanks a lot.
I redid the calculate who found in this link :
http://www.yambo-code.org/tutorials/GW/index.php
I think that the values ​​of direct and indirect gap for plot the curves taking of this section:

Code: Select all

 Indirect Gaps      [ev]: 0.636567  4.625871
  Direct Gaps        [ev]: 2.576077  8.242179

thank you again