how to plot the spin polarized charge density

[ktao]

Dear all,

Now I have a spin polarized CHGCAR. I want to plot the charge density of the minority part of the system. Could you give me some suggestion? Any comments would be appreciated.

[forsdan]

The total electron density (n_up + n_down) together with the spin density (n_up - n_down) are written to the CHGCAR. So what you can do is to write a short script that extracts the data for both densities and then solve for the minority part.

Also take a look on Danny's reply under:
http://cms.mpi.univie.ac.at/vasp-forum/ ... php?4.3865

Best regards
/Dan

[opiskelija]

I have read the reference page Forsdan gave. However I still do not understand what each of the 5 columns in my spin-polarized CHGCAR means.

CHGCAR
----------------------------------------------
(atomic coordinates)
192 224 112
192 224 112
.12542222513E+05 .12213572381E+05 .11298386482E+05 .99763507522E+04 .84629759086E+04
.69366683044E+04 .55043910496E+04 .42140592276E+04 .30916561151E+04 .21692365805E+04
.14821362350E+04 .10409059658E+04 .80534617144E+03 .68642090177E+03 .60010389775E+03
.53117179024E+03 .47509541801E+03 .42780548751E+03 .38723828847E+03 .35338109943E+03
and so on.
-----------------------------------------------------------------------

It is said that the first 3 columns are the xyz coordinates, but that seems highly unlikely. In addition, the last two columns should be (up+down) and (up-down) density, but they are too similar!

Moreover, they are even negative sometimes.

-----------------------------------------------------------------
.83381916774E-02 .70080672785E-02 .57031021424E-02 .29671948525E-02 -.22348724298E-02
-.99440048147E-02 -.19170029961E-01 -.28343486293E-01 -.35960803345E-01 -.41116494947E-01
-.43760714333E-01 -.44664180455E-01 -.45139094350E-01 -.46593663890E-01 -.50047799650E-01
-.55788268741E-01 -.63311428748E-01 -.71558593652E-01 -.79283147982E-01 -.85346826640E-01
-.88876159170E-01 -.89392545553E-01 -.87056677573E-01 -.82968370189E-01 -.79210796015E-01
-.78316696564E-01 -.82189608432E-01 -.91015853313E-01 -.10291812018E+00 -.11474812271E+00
----------------------------------------------------------------------

Does my calculation have a serious problem or something? What should I do to plot the charge density for each spin? Any comments would be appreciated.

[metosa]

copy below script to your bin directory, make it executable, run script by typing "script CHGCAR" and visualize the CHGCAR-u-d by VESTA.

#!/usr/bin/env bash
filename=`echo $1 | sed 's/PARCHG./CHGCAR-/' | sed 's/.ALLK//'`
rastgele=$RANDOM
natoms=`awk "NR==6" $1 | awk '{print ($1+$2+$3+$4+$5+$6+$7+$8+$9)}'`
ngrid=`mm $natoms+9`

grid=`awk "NR==$ngrid" $1`

nupd_b=`echo $ngrid | awk '{print ($1+1)}'`
nupd_e=`grep -n "$grid" $1 | awk "NR==2" | awk -F ":" '{print $1}'`

awk "NR<$nupd_b" $1 > title.$rastgele

cat title.$rastgele > $filename-u-d.$rastgele ; awk "NR>$nupd_e" $1 >> $filename-u-d.$rastgele
awk "NR>=$nupd_b" $filename-u-d.$rastgele > u-d1.$rastgele ; awk "NF==10" u-d1.$rastgele > u-d.$rastgele
mv $filename-u-d.$rastgele $filename-u-d

rm -rf *.$rastgele

[metosa]

sorry, you should replace

ngrid=`mm $natoms+9`

line

with

ngrid=`let ngrid=$natoms+9 : echo $ngrid`

[metosa]

sorry x 2 ;

ngrid=`let ngrid=$natoms+9 : echo $ngrid`
should be
ngrid=`let ngrid=$natoms+9 ; echo $ngrid`

[panda]

You could also use vmd, it's a free viewer and you can visualize CHGCAR there

http://www.ks.uiuc.edu/Research/vmd/

Although in the case of spin polarization I am not sure if it will help you at all, you can select spin up or spin down components for plotting, but writing your own script for plotting does always allow for the most flexibility in all cases.