[Wien] plz help required
L-marks at northwestern.edu
Thu Mar 22 14:11:51 CET 2012
Some programs in Wien2k are flexible about the format of the input
files; others are not. Almost certainly you have made a mistake in the
format of case.inorb, for instance not enough spaces, not the right
number of lines with the "U" or something else. Please read the
Calculating an oxide surface is not so simple, and doing it with LDA+U
is more complicated. Some suggestions:
a) Do a simple bulk calculation with LDA+U first, so you understand
it. For certain you need to find the appropriate unit cell with LDA+U
(volume optimization) as this is going to be different from that for
LDA/GGA. If you do not have the appropriate parameters in the plane of
the surface your results will be wrong, not because Wien2k is wrong
but because you did the calculation wrong.
b) I strongly recommend that you use a cell with inversion symmetry.
In a very few cases (e.g. ZnO) this is not possible, but in general it
is. Calculations with inversion symmetry are much faster and can be
more accurate. It takes a bit more work to set one up, but it will pay
off in the long run.
c) For a surface you have to do an optimization of the internal
positions -- in general these will be quite different from those in
the bulk. I strongly suggest that you work through a simple
optimization using the userguide first. For a novice a recommend PORT;
for the expert MSR1a. Then transfer what you have learned to the
surface and refine first with GGA (PBE, PBEsol, WC), change the
in-plane lattice parameters if needed and refine again with LDA+U.
d) I hope you have carefully checked the literature to see what the
experimental structure is for your oxide. Oxide surfaces can be very
complicated and you can end up spending a long time dong a calculation
which is not relevant. There are a lot of irrelevant DFT calculations
of oxide surfaces in the literature.
2012/3/22 arqum hashmi <arqumhashmi at yahoo.com>:
> Dear wien2k users,
> Thanks Sir stefaan for your guidance.
> i calculated oxide material surface calculation with 50 atoms in unit cell.
> Now i want to calculate it with LDA+U and this requires two files
> case.indmc, case.inorb., i changed my case.inorb and case.indmc file
> according to my need. but when i started scf cycle program stops and it
> gives this message
> LAPW0 END
>> stop error
> and when i checked the dayfile, it shows this message
> start (Sun Nov 14 05:09:04 KST 2010) with lapw0 (200/99 to go)
> cycle 1 (Sun Nov 14 05:09:04 KST 2010) (200/99 to go)
>> lapw0 -p (05:09:04) starting parallel lapw0 at Sun Nov 14 05:09:04 KST
> -------- .machine0 : processors
> running lapw0 in single mode
> 331.288u 3.272s 5:34.95 99.8% 0+0k 0+194632io 0pf+0w
>> orb -up -p (05:14:39) 0.012u 0.004s 0:00.01 100.0% 0+0k 0+32io 0pf+0w
>> stop error
> also when i checked error files, in uporb.error file, it shows this message
> Error in Vorb.
> i cannot understand what is this error and how can i solve this.
> Please guide me, i will be very thankful to you.
> Best Regards
> Arqum Hashmi
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
Professor Laurence Marks
Department of Materials Science and Engineering
"Research is to see what everybody else has seen, and to think what
nobody else has thought"
More information about the Wien