Page 1 of 1

Clarification on Si_GW example

Posted: Sun Aug 15, 2021 4:25 am
by SKM
HI
referring to the example/tutorial of Si in wiki/index.php/Bandgap_of_Si_in_GW

What is the 'extract_chi.sh', and 'extract_optics-diag.sh'.

i did optics after the DIAG run and plotted. But what will it reveal is not clear to me please.

My intention is to get GW band structure and BSE optical spectrum for optical band gap for my system. So i am testing the tutorial first.
So, will this optice from DIAG run is useful?

i understand we need to do first DFT, then GW and then BSE in sequence using results from each previous step. Right?
Regards

Re: Clarification on Si_GW example

Posted: Mon Aug 23, 2021 4:05 am
by marie-therese.huebsch
Dear SKM,

The tutorials are definitely a good starting point!

The extract_optics.sh script executes the following command:

Code: Select all

awk 'BEGIN{i=0} /<dielectricfunction>/,\
                /<\/dielectricfunction>/ \
                 {if ($1=="<r>") {a[i]=$2 ; b[i]=$3 ; c[i]=$4 ; d[i]=$5 ; i=i+1}} \
     END{for (j=0;j<i/2;j++) print a[j],b[j],b[j+i/2]}' vasprun.xml > optics.dat
So it extracts the real and imaginary part of the frequency-dependent dielectric functions. When you open vasprun.xml and search "dielectricfunction", you will find a density-density and a current-current version. I think this reference might be helpful to understand their difference: Sangalli et al., Phys. Rev. B 95, 155203.

On the other hand, the extract_chi.sh script executes the following two commands:

Code: Select all

awk 'BEGIN{i=0} /HEAD OF MICRO/,\
                /\/real/ \
                 {if ($1=="<r>") {a[i]=$2 ; b[i]=$3 ; c[i]=$4 ; d[i]=$5 ; i=i+1}} \
     END{for (j=0;j<i/2;j++) print a[j],b[j],b[j+i/2]}' vasprun.xml > chi0.dat

awk 'BEGIN{i=0} /INVERSE MACRO/,\
                /\/real/ \
                 {if ($1=="<r>") {a[i]=$2 ; b[i]=$3 ; c[i]=$4 ; d[i]=$5 ; i=i+1}} \
     END{for (j=0;j<i/2;j++) print a[j],b[j],b[j+i/2]}' vasprun.xml > chi.dat
"chi" refers to the irreducible polarizability as discussed in the context of the GW approximation of Hedin's equations. To understand the implementation, you might also want to read the publications related to the GW method that are linked in the VASP Wiki article.

Best regards,
Marie-Therese