Wrong ordering of band energies in EIGENVAL file

Message

magnon963 · #1 Post by **magnon963** » Tue May 03, 2011 2:25 pm

Dear VASP users and administrator,

I'm facing very strange problem. I'm calculating adatom binding energies on graphene surface. I did SCF calculation and with the CONTCAR of SCF step and existing CHGCAR file I did band-structure calculation. Although Band-structure as well as energies look reasonable but the band energies are not in the increasing order of energies.

--------------------------------------
53 -5.777082 -6.599961
54 -5.184411 -6.828776
55 -5.987473 -5.994021
56 -5.473722 -5.216003
57 -4.844087 -4.873635
58 -4.642753 -4.076552
59 -5.123015 -5.073820
60 -5.044789 -5.112306
61 -2.834894 -4.577795
62 -1.942797 -3.167040
63 -4.076408 -3.376941
64 -2.867073 -2.721749
65 -2.083551 -4.646101
66 -2.506179 -0.895685
67 -0.964362 -2.977814
68 -4.483737 0.499811
69 -1.395876 -2.539974
70 -0.537242 -0.499614
71 -0.647712 1.512768
72 1.944858 -2.092071
73 -0.290748 -1.805019
74 1.524444 1.659817
75 2.044700 2.015725
--------------------------------------

My INCAR file for SCF step is

---------------------------------------
Start parameter for this run:
ISTART = 0 job : 0-new 1-cont 2-samecut
ICHARG = 2 charge : 1-file 2-atom 10-const
INIWAV = 1 electr : 0-lowe 1-rand

Electronic Relaxation 1
ENCUT = 550 eV
IALGO = 48 algorithm
NELM = 220
NELMIN = 0
NELMDL = -20 # of ELM STEP m
EDIFF = 1E-05 stopping cretiria for ELM
GGA = PE
LDIAG = .F.
ISPIN = 2
LORBIT = 11
MAGMOM = 8*0.5 8*0.5 16*0.5 1*2.5

LMAXMIX = 4
INIMIX = 0
AMIX = 0.1
BMIX = 0.00001
AMIX_MAG= 0.4
BMIX_MAG= 0.000001
MAXMIX = -8

Ionic Relaxation
NSW = 250 number of steps for IOM #
EDIFFG = -1.E-02 stopping-criterion for IOM
ISIF = 2
ISYM = 2
LCORR = T
IWAVPR = 1
IBRION = 2
POTIM = 0.4
VOSKOWN = 1

NPAR = 16
NSIM = 1
LPLANE = .TRUE.

DOS related values:
SIGMA = 0.02
ISMEAR = 2 broad. in eV, -4-tet -1-fermi 0 gaus

NBAND = 96

LREAL = Auto
---------------------------------------

If someone can tell me why the ordering of band-energies is not proper and can suggest me something to overcome this problem.

with best regards,
mag.

#2 Post by **admin** » Wed May 11, 2011 5:41 pm

please do a final diagonalization by setting LDIAG = .True.

superyoyo · #3 Post by **superyoyo** » Thu May 12, 2011 8:52 am

[quote=" .True.[/quote"]
Dear Admin,

I did try that for vasp5.2.2, however, it does not help.
I am sure that I put LDIAG = .True. exactly in the INCAR.
Thanks.