Band Structure Problem

Message

shaka · #1 Post by **shaka** » Sun Sep 30, 2007 5:59 pm

Dear VASP masters and users:

I am doing calculation of band structure of LiNbO3 using VASP. I optimized the structure first and calculated the DOS, which is consistent with the published results. However, when I tried to calculated the band structure, I always got this error: charge density could not be read from file CHGCAR for ICHARG>10 . I have read the previous discussion about this problem. But I am still confused. It seems this message indicates there is some problem with my KPOINT ( I check the CHGCAR in the running directory). I used 4x4x4 KPOINT to optimize the structure and Line-mode to calculate band structure.

Line-mode
reciprocal
0.0000000 0.000000 0.000000 1 ! G
0.5000000 0.500000 0.500000 1 ! Z

0.5000000 0.500000 0.500000 1 ! Z
1.0000000 0.500000 0.500000 1 ! F

1.0000000 0.500000 0.500000 1 ! F
0.5000000 0.000000 0.000000 1 ! L

0.5000000 0.000000 0.000000 1 ! L
0.0000000 0.000000 0.000000 1 ! G

I guess there is some incompatibility between two sets of KPOINT. Does this mean I have to optimize the structure using Line-mode also? Or there is something else I am missing. Any reponse or suggestions will be highly appreciated !

The INCAR for optimizing the structure and calculating band structure are following:

INCAR_OPT

Starting parameters for this run:
NWRITE = 2
ISTART = 0
ICHARG = 2
INIWAV = 1

Electronic Relaxation:
PREC = high low | medium | high

NELM = 90 number of ELM steps
NELMIN = 10
EDIFF = 1E-04 energy stopping-criterion for ELM
LREAL = AUTO real-space projection (.FALSE., .TRUE., On, Auto)
ALGO = 48
LCHARG = .TRUE.
LWAVE = .TRUE.
LELF = .TRUE.

Ionic Relaxation:
NSW = 1000
IBRION = 3
EDIFFG = -0.001
ISIF = 3
ISYM = 1
POTIM = 0.1

DOS related values:
ISMEAR = 1 (-1-Fermi, 1-Methfessel/Paxton)
SIGMA = 0.02 broadening in eV

Parallelization flags:
NPAR = 4
LPLANE = .TRUE.

INCAR_BAND

Starting parameters for this run:
NWRITE = 2
ISTART = 1
ICHARG = 11
INIWAV = 1 electr: 0-lowe 1-rand 2-diag

Electronic Relaxation:
PREC = high

NELMDL = -7
NELM = 90
NELMIN = 10
EDIFF = 1E-04 energy stopping-criterion for ELM
WEIMIN = 0
LMAXMIX = 6
LREAL = AUTO real-space projection (.FALSE., .TRUE., On, Auto)
ALGO = 48
NSIM = 1 NSIM bands are optimized simultaneously (IALGO=48 only)

LCHARG = .TRUE.
LWAVE = .TRUE.
LELF = .TRUE.

Ionic Relaxation:
NSW = 0
IBRION = -1
EDIFFG = -0.001
ISIF = 3
ISYM = 1
POTIM = 0.1

DOS related values:
RWIGS = 1.376 1.894 1.03
LORBIT = -1
ISMEAR = 1 (-1-Fermi, 1-Methfessel/Paxton)
SIGMA = 0.02 broadening in eV

Parallelization flags:
NPAR = 2
LPLANE= .TRUE.

THANKS

#2 Post by **admin** » Tue Oct 02, 2007 7:34 am

probably your CHGCAR file is not available in the directory in which the job is executing, or it is not compatible, this error message has nothing to do with the KPOINTS file.

vaspman · #3 Post by **vaspman** » Tue Oct 02, 2007 10:33 am

As a new user of VASP I am not sure but your problem might be this one
Fisr you used ICHARG = 2 and then for Band structure you used ICHARG = 11

but VASP manual says we have to add +10 for non SCF.
So i think for your band calculations you will need ICHARG = 12.
Please correct me any one if I am wrong.

#4 Post by **admin** » Tue Oct 09, 2007 8:16 am

no, if you set ICHARG=12, the charge density is taken from the superposition of the atomic charges and kept fixed. This does not correspond to the self-consistent charge density of the compound.

lcyin · #5 Post by **lcyin** » Wed Mar 05, 2008 12:57 am

Hi, try to do a scf calculation after the geometry optimization, then do the DOS or Band calculations using the charge density obtained in the scf calculation.

enthusiastic · #6 Post by **enthusiastic** » Fri Apr 04, 2008 1:44 am

i am still in the trouble.
what i was told are these:
WARNING: number of atoms are different on CHGCAR file
ERROR: charge density could not be read from file CHGCAR for ICHARG>10
i don't find a probably method to solve it yet, though i have read many replies here.thanks for your help.

chenweiguang · #7 Post by **chenweiguang** » Fri Apr 25, 2008 5:07 am

You should check your POSCAR first.

giacomo giorgi · #8 Post by **giacomo giorgi** » Tue Mar 17, 2009 5:33 am

why did you set ALGO = 48 ?
Should not be IALGO =48? and then the according value of ALGO (maybe = F?)

In the way you set your input you should work with IALGO=38 and not in IALGO = 48 as I think you want to work.

I do not know, I am very new with VASP then probably my observation is stupid and not related to your problem.

My best,
Giac

#9 Post by **admin** » Mon Mar 23, 2009 10:34 am

IALGO and ALGO both set the way you choose the electronic optimization algorithm:
IALGO=38 is the same as ALGO=Normal
IALGO=48 corresponds to ALGO=Very Fast
and ALGO=Fast combines both algorithms.

An input line ALGO=48 will be ignored, you are correct. vasp will use the default Davidson algorithm then

If you define both, ALGO and IALGO in the same INCAR, your choice set in ALGO will be used (IALGO is overwritten then)

RichardBlank · #10 Post by **RichardBlank** » Fri May 22, 2009 8:24 pm

Thanks for the updates.

Richard

http://www.costaricascallcenter.com/pag ... =index.php

krackon · #11 Post by **krackon** » Tue Dec 01, 2009 9:33 am

I am having the same problem. Errors are:
ERROR: Charge density could not be read from the file CHGCAR for ICHRG>10
WARNING: Number of atoms are different on CHGCAR file

I have set NGXF, NGYF, NGZF values in the INCAR and checking CHGCAR and OUTCAR they match. As well the number of atoms in the header of the CHGCAR and POSCAR match.

Any suggestions would be appreciated.

Kevin

ZachTexas · #12 Post by **ZachTexas** » Wed Dec 08, 2010 8:16 pm

Check your CHGCAR. If the first line is blank, this error will result. Putting any non-white space characters in the first line will fix this error.

You can also add the line
SYSTEM=(non white-space characters)

To your INCAR before generating the CHGCAR. This will ensure that your CHGCAR contains the first line.

liuzr9 · #13 Post by **liuzr9** » Sat Dec 25, 2010 2:24 am

I also encountered similar problem. I found that the POSCAR and CHGCAR should be consistent.

hatdau · #14 Post by **hatdau** » Wed Apr 06, 2011 7:04 pm

It would help others to understand your problem if you can upload your INCAR, KPOINT, POSCAR and corresponding OUTCAR for each calculation (optimization, scf, band ..)

hatdau · #15 Post by **hatdau** » Wed Apr 06, 2011 7:04 pm

It would help others to understand your problem if you can upload your INCAR, KPOINT, POSCAR and corresponding OUTCAR for each calculation (optimization, scf, band ..)