The ABINIT Users Mailing List ( CLOSED )

[matthieu verstraete <Matthieu.Verstraete@ulg.ac.be>]

Hello all,

the point was to move the discussion to the web forum (soon all
activity will be shifted over), but oh well:

1) if the irred wedge k-points that abinit determines using kptopt 1
(list all with prtvol 10) are the ones you want, check the _EIG file.

2) otherwise, set the kpoints explicitly using kptopt 0
(http://www.abinit.org/documentation/helpfiles/for-v5.8/input_variables/varbas.html#kptopt),
nkpt kpt(:,:) and wtk(:). The resulting WFK and EIG files will contain
what you want, I believe.

Re: Enrico's comments: For ZnO this must be a DFT problem, but I was
not aware of it (band ordering wrong?). And non SCF bands are cheap to
calculate. Further, I don't see why you need to do more than in the
SCF calculation. If you really do need a super fine mesh, start with
an SCF on a decent grid, then do non-SCF with kptopt 1 ngkpt 1zillion
1zillion 1zillion and it will extract the eig and wf for the
irreducible kpoints on the fine grid. At any rate, if you already have
your GW bands, you're certainly right that an EPM will be much
quicker. You could also do GW band Wannier interpolation, a la Don
Hamann.

Ciao

Matthieu


On Fri, Feb 19, 2010 at 12:18 AM, Enrico Bellotti <bellotti@bu.edu> wrote:
> Denis,
> we tried the approach you are pursuing a while ago. You have to convert your
> mesh of k-points in terms of the reciprocal lattice vectors of the crystal
> you are using. Then, with a lot of patience, you have to compute the energy
> eigenvalues of all the points with a non-self-consistent runs. We eventually
> succeeded but the result it is not worth the effort. We found that it is
> much simpler and accurate to used an EPM and reproduce the GW bands with it.
> For some materials, for example ZnO, abinit will give you the wrong valence
> bands that you can not use anyway.
> Best.
> EB
>
>
> At 10:40 PM 2/18/2010 +0100, you wrote:
>>
>> Dear Matthieu,
>>
>> thanks for posting.
>> I want to do full band Monte Carlo simulations. Therefore I need the
>> band structure in the whole irreducible wedge and not only across
>> paths of high symmetry. My irreducible wedge is meshed. I have a list
>> of k points, my mesh points, where I have to know the energy values. I
>> could translate my 3D mesh into begin and end points to use kptbounds,
>> but this is, let's say unnatural or tedious. Do you know if it's
>> possible to give abinit a bunch of k points, where it evaluates the
>> energies at these points?
>>
>>
>> Best regards
>> Denis
>>
>>
>> Quoting "matthieu verstraete" <Matthieu.Verstraete@ulg.ac.be>:
>>
>>> now answered on the forum:
>>>
>>> http://forum.abinit.org/viewtopic.php?f=8&t=65
>>>
>>> On Thu, Feb 18, 2010 at 6:32 PM, Denis Dolgos <dolgos@iis.ee.ethz.ch>
>>> wrote:
>>>>
>>>> Dear all,
>>>>
>>>> I would like to calculate the full band electronic structure of e.g.
>>>> GaAs in
>>>> the irreducible wedge of the Brillouin zone. If I remember correctly,
>>>> only
>>>> lines through the Brillouin zone were computed using kptbounds and ndivk
>>>> in
>>>> the tutorial. To get the eigenenergies in the irreducible wedge via
>>>> kptbounds is quite tedious.
>>>>
>>>> Is it possible to define the k points one by one for the eigenenergies
>>>> output?
>>>>
>>>> Thanks in advance for any advice.
>>>>
>>>>
>>>> Greeting from Zurich
>>>> Denis
>>>>
>>>
>>>
>>>
>>> --
>>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>> Prof. Matthieu Verstraete
>>>
>>> Universite de Liège
>>> Institut de Physique, Bat. B5, 3/7
>>> Allée du 6 aout, 17
>>> B- 4000 Sart Tilman, Liège
>>> Belgium
>>>
>>> Associate member European Theoretical Spectroscopy Facility (ETSF
>>> http://www.etsf.eu)
>>>
>>> Phone : +32 4 366 90 17
>>> Fax   : +32 4 366 36 29
>>>
>>> Mail : matthieu.verstraete@ulg.ac.be
>>>          matthieu.jean.verstraete@gmail.com
>>
>>
>>
>
> ||||  Enrico Bellotti, Ph.D.
> _ _   Associate Professor
> . .   Boston University
>  |    Electrical & Computer Engineering Dept.
> \_/   8 Saint Mary's Street
>      Boston, MA 02215-2421
>      USA
>      Tel. 617-358-1576
>      Fax. 617-353-1283
>      e-mail bellotti@bu.edu
>
>



-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Prof. Matthieu Verstraete

Universite de Liège
Institut de Physique, Bat. B5, 3/7
Allée du 6 aout, 17
B- 4000 Sart Tilman, Liège
Belgium

Associate member European Theoretical Spectroscopy Facility (ETSF
http://www.etsf.eu)

Phone : +32 4 366 90 17
Fax   : +32 4 366 36 29

Mail : matthieu.verstraete@ulg.ac.be
          matthieu.jean.verstraete@gmail.com