Columns of DOSCAR file

Message

blazing · #1 Post by **blazing** » Mon Aug 10, 2009 10:55 am

Hi, everybody!

What columns of DOSCAR correspond to pz, px, py, dz2, dxz, dyz, dxy, dx2-y2 (and so on....) partial DOSs? i didnt found it in manual.

Thanks

Mikhail

forsdan · #2 Post by **forsdan** » Mon Aug 10, 2009 5:18 pm

http://cms.mpi.univie.ac.at/vasp-forum/ ... .php?3.306

See the last comment by the admin.

blazing · #3 Post by **blazing** » Tue Aug 11, 2009 4:31 am

Thanks a lot!!

blazing · #4 Post by **blazing** » Wed Aug 12, 2009 11:18 am

In relation with the thread following question. i perform static calculation of DOS (no relaxation). calculated partial Fe 3d xz and yz DOSs fairly different. But this planes of supercell is equivalent. so i would expect 3d xz and yz DOSs to be equal. why they so different?

i use following INCAR and POSCAR

SYSTEM = Si-Fe

ISTART = 0

PREC = High

IALGO = 38

LORBIT = 11

VOSKOWN = 1

ISMEAR = -5

SIGMA = 0.2

ICHARG = 2

ISPIN = 2

MAGMOM = 2*2.2 22*0

ENCUT = 500.00 eV

RWIGS = 1.45 1.32

EMIN = -12.0

EMAX = 12.0

NEDOS = 4500

NELMIN = 7

ISYM = 0

EDIFF = 1E-04

NELM = 100

LMAXMIX = 4

_________________________________________________________

Si-Fe
5.46629641000000
1.0000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 1.0000000000000000 0.0000000000000000
0.0000000000000000 0.0000000000000000 3.0000000000000004
2 22
Direct
-0.0000001665126154 0.9999994757712332 0.9999991992102141
0.4999999242219825 0.4999994044806253 0.0000008427821678
0.2502413656913599 0.2502414702442717 0.0755009127318464
0.7497579098538607 0.7497580882263163 0.0755009320541033
0.4999999705984418 0.0000001275445618 0.1612810753170716
0.0000000216216688 0.5000000754537444 0.1612810494549297
0.7499987234778644 0.2500014356763398 0.2463117951690840
0.2500013019109163 0.7499987895118092 0.2463117847534330
0.0000000311016059 0.0000000945531129 0.3308557753582138
0.5000000142102849 0.5000001140466511 0.3308574548101630
0.2500016165761108 0.2500016947636573 0.4153651401647637
0.7499984291188296 0.7499985260657783 0.4153651332107411
0.5000000305006924 0.0000001093129547 0.4999999875402176
0.0000000163364906 0.5000001018218755 0.4999999848917632
0.7499984252517670 0.2500017248790424 0.5846348729943670
0.2500016491229251 0.7499984894648687 0.5846348812144275
0.0000000366874459 0.0000001149173411 0.6691441679722935
0.5000000242707757 0.5000000879383629 0.6691425623870116
0.2500012275193796 0.2500013031597314 0.7536881966810683
0.7499988361922804 0.7499988878872008 0.7536881993653227
0.5000000096212143 0.0000000487501474 0.8387189415755849
0.0000000663320584 0.5000000988598432 0.8387189406447495
0.7497563792510331 0.2502437502701064 0.9244990589111671
0.2502441570436415 0.7497559864004311 0.9244991108052957

forsdan · #5 Post by **forsdan** » Sun Aug 16, 2009 11:27 am

If you (from symmetry arguments) are fully convinced that the Fe 3d xz and yz DOS should be the same on the specific Fe iron atom you're looking at, please first make sure that you are comparing the correct columns in the DOSCAR file. For spin-polarized calculations the format is

"energy s(spin up) s(spin down) p_y(spin up) p_y(spin down) ... " and so on according to the order in the link above.

If the difference is only minor, please try to increase the k-point sampling. Switching on the symmetrization (ISYM > 0) can also be an option.

Cheers,
/Dan Fors

<span class='smallblacktext'>[ Edited Sun Aug 16 2009, 06:47PM ]</span>

blazing · #6 Post by **blazing** » Thu Aug 20, 2009 5:05 am

you were right. i thought that the arrangement of the columns is like that:

"energy s(spin up) s(spin down) p_y(spin up) p_z(spin up) p_x(spin up) p_y(spin down) p_z(spin down) p_x(spin down)... "

Thanks for the help

Best regards

Mikhail