[Wien] charged and neutral cell
pieper
pieper at ifp.tuwien.ac.at
Sat Dec 2 01:00:46 CET 2017
As I said before: when you remove an atom from case.struct, you remove
the nuclear charge AND all Z electrons. This is neutral. If your perfect
structure was neutral, then the new structure is neutral as well.
Don't try to tell Wien2k that Ba should be missing 2 electrons and O
should have them. It makes no sense: with Wien2k you have a very
sophisticated tool which has been designed by extremely clever people to
do with enormous precision exactly what you apparently want to do by
hand: distribute electrons in a solid. Throw the program a handfull of
neutral atoms, tell it their positions - thats what the struct file is
for - and let it do it's magic.
As a guidance, I really recommend the UG, and use the w2web interface.
In short (fill in the gaps from the UG):
- create a case directory for the perfect structure (perhaps using the
web interface w2web). Generate a valid case.struct file (via w2web's
structeditor, from a cif file ... whatever). Walk through
initialization. Run the scf. Inspect the file case.scf. Calculate DOS,
do a band structure spagetti plot ... I gues you have done this already.
- create a new case directory and copy the struct file of the perfect
structure to this directory. Use supercell to generate your 2x2x2 unit
cell (perhaps P, not centered). Rename it to match the directory name.
Use w2webs the structeditor to kill one Ba atom, if that is what you
want to do first. (A 2x2x2 unit cell will be a pedagogical exercise, it
is too small to accomodate the structural relaxation such a massive
defect will need). Make sure that the RMT's in the struct file are at
least 5% from touching since you definitely want to give Wien2k a chance
to relax the defect structure (see UG on structural relaxation). Walk
step by step through initialization: nn first. It will be unhappy with
the multiplicities and creat a new struct. Accept the this new struct
file. Next sgroup, which will most probably again make its own changes.
Accept them. With a little bit of luck this struct file is then ok for
the symmetry program. If the symmetry program crashes it may be due to
different rounding of exact positions in sgroup and symmetry - there was
a recent thread about in the mailing list. If symmetry is happy you are
probably in the green with the defect structure - inspect the structure
with your favorite crystal structure viewing tool, perhaps xcrysden.
Proceed with initialization. If you want run a simple scf without
structural relaxation as a reference. For anything beyond pedagogical
exercises I think scf runs with structural relaxation are mandatory.
- rinse and repeat with whatever defect structure you are interested in,
each in fresh case directory.
---
Dr. Martin Pieper
Karl-Franzens University
Institute of Physics
Universitätsplatz 5
A-8010 Graz
Austria
Tel.: +43-(0)316-380-8564
Am 30.11.2017 20:27, schrieb chin Sabsu:
> Dear Pieper Sir,
>
> Thank you for detailed reply.
>
> Okay, I understood few things now.
> But if we remove one atom from the super cell then how we can say that
> the defective cell is charge neutral?
>
> I assume that I will be convinced from your next reply that "how a
> defective supercell is considered charge neutral (A^+2 from 2x2x2 cell
> of ABO3)".
>
> Now Please guide me how I can create fully charged A or B or oxygen
> vacancies defects in ABO3 system?
>
> Thank you in advance.
>
> Sincerely
> Chin
>
> On Thursday, 30 November 2017 6:26 PM, pieper
> <pieper at ifp.tuwien.ac.at> wrote:
>
> No, if you remove atoms in the case.struct file, you remove atoms,
> that
> is neutral entities. Just don't play around with the entry Z for the
> number of electrons (which you should not do if you are not sure what
> you are doing). It defaults to the nuclear charge - use the w2w2b
> interface and leave the entry empty in structgen, it will be filled in
>
> automatically. You need not do anything special during initialization
> to
> maintain charge neutrality.
>
> As for the second question, readjust your concept of valence to align
> with how it is used within Wien2k. In a nutshell, distinction between
> core and valence electrons in Wien2k is made by their binding energy
> to
> a specific nuclear charge. If the binding energy is large enough, the
> total charge density of the electron is to high accuracy within the
> sphere of radius RMT around that nuclear charge - this is a core
> electron. If this is not the case, its a valence electron. It is much
> more difficult to be a core electron in Wien2k than it is in your
> classical thinking of Ba2+ and O2-. Leave that distinction to the
> program lstart during initialization. (There is a lot more, look into
> the introductory section of the UG and continue reading from there).
>
> If you want to calculate defect structures, start from the perfect
> structure. Perhaps begin with examples that come with the
> Wien2k-package. After a successful scf run inspect the file case.scf.
> Search (grep) for
> :NEC
> You will find two numbers, the total number of charges, nuclear and
> electrons. They should be equal within the accuracy lstart uses to
> decide wether or not charge density leaks out of the RMT-sphere.
>
> You will also find
> :NOE
> Which is the total number of not-core electrons, or
>
> :CTOnnn
> which is the total charge (in one spin channnel) in sphere nnn, and
> many
> other projections of the charge density on specific volumes, energy
> windows, spin or orbital states. Use the UG section on the scf file as
>
> an introduction. In adition you can use analysis programs like AIM
> (see
> UG, section on AIM) to calculate details.
>
> Then create a new case directory and restart there by modifying your
> file case.struct - perhaps using programs like structgen and
> supercell.
>
> Good luck
>
> ---
> Dr. Martin Pieper
> Karl-Franzens University
> Institute of Physics
> Universitätsplatz 5
> A-8010 Graz
> Austria
> Tel.: +43-(0)316-380-8564
>
> Am 30.11.2017 06:12, schrieb chin Sabsu:
>> Dear Peter Sir and other experts,
>>
>> Could you please explain how to deal charged vacancy and neutral
>> vacancy?
>>
>> If I remove an atom (say x^+2) from the unit cell then the Unit cell
>> will be charged unit cell with charge -2e.
>> If I remove an atom (say x^-2) from the unit cell then the Unit cell
>> will be charged unit cell with charge +2e.
>>
>> Are the above statements are correct?
>>
>> If above statements are correct then please suggest me how to make
> the
>> system neutral after creating vacancy.
>>
>> Below are the lines from case.in2 and case.inm where, as far as I
>> know, I need to do some changes.
>>
>> Assuming that I have xx NE in case.in2 and I remove an atom of
> vacancy
>> +2 then how I should modify these two files so that the final
>> defective cell remains charge neutral?
>>
>> case.in2
>> TOT (TOT,FOR,QTL,EFG,FERMI)
>> -13.2 xx.0 0.50 0.05 1 EMIN, NE, ESEPERMIN, ESEPER0,
>> iqtlsave
>>
>> case.inm
>> MSR1 0.0 YES (BROYD/PRATT, BG charge (-1 for core hole), norm)
>>
>> I have one another question:
>>
>> How the number NE (here I kept it xx) in case.in2 changes case by
>> case?
>>
>> In case of BaZrO3 it, NE, is 40 but I could not find any way to
>> calculate this number manually because Ba is having +2, Zr also +2
> and
>> O is -2. Total number of atoms are 5(Ba*1, Zr*1 and O*3) in the unit
>> cell and below is electronic configuration of atoms.
>>
>> Ba: _1s22s2_2p6_3s2_3p63d10_4s2_4p64d10_5S2_5P6_6S2_
>> Zr: _1s22s2_2p6_3s2_3p63d10_4S2_4P6_4D25S2_
>> _O:__1s2 2S2 2P4_
>
>>
>> _Bold are treated as F (valence) states during initialization._
>> Looking forward to hearing from any of the expert.
>>
>> Sincerely
>>
>> Chin
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>> SEARCH the MAILING-LIST at:
>>
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
More information about the Wien
mailing list