Skip to content

Latest commit

 

History

History
172 lines (141 loc) · 6.45 KB

README_wannier.org

File metadata and controls

172 lines (141 loc) · 6.45 KB

ecalj: Mmaximally localized Wannier with effective interaction in CRPA

SRC/wanniergw/ contains the Wannier function generator on top of ecalj resouces.

Srcipt genMLWF

>genMLWF is the script to generate the Wanneir functions. In addtion, it gives effective interaction <ij|W|kl> in CRPA. Required setting is written in the GWinput file (mkGWIN_lmf2 generate template in GWinput.tmp).

How to run samples?

We have several samples for genMLWF

Samples are: at ecaj/MATERIALS/MLWF_samples ├── CuMLWFs n1n2n3 4 4 4, rough but small sample to test it at firts. Within ~five minutes. ├── Cu ├── Fe ├── La2CuO4 ├── NiOMLWF ├── SrVO3MLWF

Run a sample at MATERIALS/MLWF_samples/CuMLWFs

Run self-consistent calculation as >lmfa cu >mpirun -np 4 lmf-MPIK cu

Then we make band plot by >job_band cu -np 4 , where 4 means number of cores. This gives correct Fermi energy, and stored into bnds.* file

Then we run main script of maxloc wannier with effective interaction W as >genMLWF -np 12 cu

(1) Look into ecalj/MATERIALS/CuMLWFs/bandplot.MLWF.cu.glt This is for interpolated band. A line “bnds.maxloc.up” u ($5):($6+de) lt 3 w l ti “Wannier” is added to usual output of bandplot.cu.isp* given by job_band.

(2) Plot psi.xsf file. This contains MaxLoc Wannier function. (not necessary; you can comment out “wanplot” line if you don’t need to plot Wannier.)

(3) grep Wan lwmat* This gives the matrix element of the Coulomb interaction and W-v. (off-diagonal elments are also calculated. Sorry, I have not yet document how to read this…)

(4) We have to specify inner window (<worb> section in GWinput) and so on related to wannier in GWinput. The template is given by mkGWIN_lmf2 from ctrl file.

OUTPUT

Wannier funciton plot

Use wan.xsf by Xcrysden

Band plot including Wannier band

See bandplot.MLWF.isp1.glt

Plot U, J parameters.

  1. Run the srcipt genMLWF; please look into this script. (originally at SRC/exec/genMLWF ).

    This is made from two stages. >argin=2; run_arg $argin $NO_MPI $nfpgw /hmaxloc lmaxloc2 (equilavent to echo 2| hmaxloc >lmaxloc2) is the end of Wannier function generation.

    At this point, you can make band plot to check whether your setting for Wannier work well or not; the model-Hilbert space by band plot. (we need syml.* file and run job_band to get original energy bands) Then plot wannier band on top of it. See MATERIALS/MLWF_samples/CuMLWFs/bandplot.MLWF.isp1.glt as an example.

` If the plot is strange, you need to choose windows for Wannier. (Repeat echo 2| hmaxloc >lmaxloc2 until you have satisfactry fitting with changing the setting wannier part in GWinput ).

  1. We get three files (see genMLWF) three files containing v and W-v information. grep “Wannier” lwmatK1 > Coulomb_v grep “Wannier” lwmatK2 > Screening_W-v grep “Wannier” lwmatK3 > Screening_W-v_crpa

These are text files <ab|W|cd> element a,b,c,d are index of Wannier functions (sorry, ask us if necessary).

Then we have Static_W.dat (RPA) and Static_U.dat (cRPA) These contains static U, U’, J, and J’ (ω = 0).

grep ’ 1 1 1 1 1’ Coulmb_v grep ’ 1 1 1 1 1 0.000000’ Screening_W-v.UP grep ’ 1 1 1 1 1 0.000000’ Screening_W-v.crpa shows Coulomb_v.UP: Wannier 1 1 0.000000 0.000000 0.000000 1 1 1 1 1 23.499183 -0.000000 Screening_W-v.UP: Wannier 1 1 0.000000 0.000000 0.000000 1 1 1 1 1 0.000000 0.000000 -20.317956 -0.000000 Screening_W-v_crpa.UP: Wannier 1 1 0.000000 0.000000 0.000000 1 1 1 1 1 0.000000 0.000000 -20.188076 -0.000000 This means <11|W|11> =23.499183-20.317956 <11|U_CRPA|11>=23.499183-20.188076 Note that this is a test example.

NOTE:

For practical calculations, we need to stop right after a line >argin=2; run_arg $argin $NO_MPI $nfpgw /hmaxloc lmaxloc2 in the genMLWF. (Need to modify genMLWF by hand; note that “run_arg” is a subrouitne of bash. This line is equivalent to >echo 2| hmaxloc >& lmaxloc2) Then we need to check the band plot, with superposing Wannier band “bnds.maxloc.up” u ($5):($6). If the Wannier band is not reasonable, change settings of inner and outer windows, and run >echo 2| hmaxloc |tee lmaxloc2 again and again until you have a reasonable Wannier band plot. Then go ahead to the next steps (you may use genMLWF2).

genMLWF I added the “grep” commands in the last part of genMLWF to make

hwmatK_MPI.F print all matrix elements of W and frequency in eV. called in genMLWF

CAUTION:

you must run job_band in advance to genMLWF! If not you may need to follow instruction of “Efermi shift” as follows;

NOTE: Efermi shift: genMLWF requies bnds.${target} to read the Fermi energy. To generate it, we need to run job_band in advance. Or run, >echo 2 | hmaxloc > lmaxloc2 (need syml*); this can be runned after genMLWF. ---- (Or need to shift Ef by hand as follows in gnuplot script.)

de = ((ef shown in “lmaxloc2”) - (ef in llmf_ef(bnds.${target}))*13.605 plot \ “bnd1.dat” u 2:3 lt 1 pt 1 not w l,\ “bnd2.dat” u 2:3 lt 1 pt 1 not w l,\ “bnd3.dat” u 2:3 lt 1 pt 1 not w l,\ “bnd4.dat” u 2:3 lt 1 pt 1 not w l,\ “bnd5.dat” u 2:3 lt 1 pt 1 not w l,\ “bnds.maxloc.up” u ($5):($6+de) lt 3 w l ti “Wannier”

Known problems:

Range of plot looks not good;

Especially, vis_wan_ubound, vis_wan_lbound should be not integer. Probably, need to improve/(bug fix) wanplot.F.

xsf files (phi.xsf, wan.xsf) are special for Xcrysden,

thus it might be inconvenient. You need to use GUI of Xcrysden. But I have not tested no other format.

development history:

Started from Apr2015, T.kotani take developments of MaxLoc Wannier, which is implemented in the ecalj by T.Miyake R.Sakakma and H.Kino. 2009; maxloc090910 (T.Miyake) 2009; Its documentation and Visualizer at Kino’s https://github.com/nim-hrkn/visualize 2014Aug; T.Kotani modified it so as to fit to latest ecalj. 2015jan: Sengwoo Jang’s modification. 2015apr: Sengwoo’s RPAWannierTable.py.

2020aug: reorganize and checke README_wannier.org