[Pw_forum] electron_phonon="interpolated" with trans=.false. fails

Matteo Calandra matteo.calandra at impmc.jussieu.fr
Mon Jun 18 09:56:54 CEST 2012


Dear Alaska,

   if you want to calculate the electron-phonon coupling at a given 
q-point you can use the option 'simple' and this is the old calculation
without linear interpolation over q-points. In the case 'simple', 
everything work by specifying nk1, nk2, nk3 in the phonon run with 
trans=.false. .
In the case you want to carry out the linear interpolation in
arXiv:cond-mat/0504077
the situation is different and I cannot help you too much as I did not
work on this part of the routine.

All the best,

Matteo.


Le 18/06/12 09:51, pw_forum-request at pwscf.org a écrit :
> Send Pw_forum mailing list submissions to
> 	pw_forum at pwscf.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
> 	http://www.democritos.it/mailman/listinfo/pw_forum
> or, via email, send a message with subject or body 'help' to
> 	pw_forum-request at pwscf.org
>
> You can reach the person managing the list at
> 	pw_forum-owner at pwscf.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Pw_forum digest..."
>
>
> Today's Topics:
>
>     1. electron_phonon="interpolated" with trans=.false. fails
>        (Alaska Subedi)
>     2. Re: electron_phonon="interpolated" with trans=.false.	fails
>        (Alaska Subedi)
>     3. d-orbitals in fully relativistic pseudopotential for Carbon
>        -- band gap in graphene (Martin Gmitra)
>     4. Re: Regarding calculation of nmr spectra using gipaw.x
>        (Lorenzo Paulatto)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sun, 17 Jun 2012 16:36:52 +0200
> From: Alaska Subedi<asubedi at gmail.com>
> Subject: [Pw_forum] electron_phonon="interpolated" with trans=.false.
> 	fails
> To: PWSCF Forum<pw_forum at pwscf.org>
> Message-ID:
> 	<CAJdU21O_pq_chFHotQGsSxSP_r9Op83fYnhxwoo8UfpMcZo1tQ at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Dear All,
>
>
> I would like to calculate to phonon calculations on a coarser gird and
> then do electron-phonon calculation on a finer grid. However, this
> does not seem to be possible anymore in version 5.0. The program
> crashes in the elphon subroutine while calling davcio_drho.
>
>
> Any pointer to fix this error would be greatly appreciated. I have
> attached my input files below.
>
> Thanks,
> Alaska
>
> $ cat Pt-ep1.scf.in
>
>   &control
>      calculation = 'scf'
>      restart_mode='from_scratch',
>      prefix='Pt',
>      tprnfor = .true.,
>      tstress =.true.,
>      pseudo_dir = './pseudo/',
>      outdir='./tmp/'
>   /
>   &system
>      ibrav=  2, celldm(1) =7.42, nat=  1, ntyp= 1,
>      lspinorb=.true.,
>      noncolin=.true.,
>      starting_magnetization=0.0,
>      occupations='smearing',
>      degauss=0.02,
>      smearing='mp',
>      ecutwfc =30.0,
>      ecutrho =250.0,
>      la2f = .true.
>   /
>   &electrons
>      mixing_beta = 0.7,
>      conv_thr =  1.0d-8
>   /
> ATOMIC_SPECIES
> Pt  0.0    Pt.rel-pz-n-rrkjus.UPF
> ATOMIC_POSITIONS
> Pt  0.0000000   0.00000000   0.0
> K_POINTS AUTOMATIC
>    4  4  4  0  0  0
>
>
> $ cat Pt-ep1.ph.in
>
> phonons of Pt
>   &inputph
>    amass(1)=195.078,
>    prefix='Pt',
>    fildvscf='dvscf'
>    outdir='./tmp'
>    fildyn='ptdyn',
>    tr2_ph=1.0d-16,
>   /
> 1.0 0.0 0.0
>
>
> $ cat Pt-ep1.ep.in
>
> phonons of Pt
>   &inputph
>    amass(1)=195.078,
>    prefix='Pt',
>    fildvscf='dvscf'
>    outdir='./tmp'
>    fildyn='ptdyn',
>    tr2_ph=1.0d-16,
>    trans=.false.
>    electron_phonon='interpolated'
>   /
> 1.0 0.0 0.0
>
>
> ------------------------------
>
> Message: 2
> Date: Sun, 17 Jun 2012 17:52:21 +0200
> From: Alaska Subedi<asubedi at gmail.com>
> Subject: Re: [Pw_forum] electron_phonon="interpolated" with
> 	trans=.false.	fails
> To: PWSCF Forum<pw_forum at pwscf.org>
> Message-ID:
> 	<CAJdU21M_2nWD4qUwhuvTQETTVHnTjy6JL8b-30bw2Ye9MjvGiw at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
>> The program
>> crashes in the elphon subroutine while calling davcio_drho.
>
> It seems the problem is that the arrays int3 and int3_nc (and possibly
> others) do not get allocated when trans = .false. as they have now
> been removed from allocate_phq.f90.
>
>
> Alaska
>
>
> ------------------------------
>
> Message: 3
> Date: Sun, 17 Jun 2012 19:18:28 +0200
> From: Martin Gmitra<martin.gmitra at gmail.com>
> Subject: [Pw_forum] d-orbitals in fully relativistic pseudopotential
> 	for Carbon -- band gap in graphene
> To: pw_forum at pwscf.org
> Message-ID:
> 	<CAG-DMb2BfOOyvVZSNoqCfGU1i-tTO+kqNqTe9b8opTAz8LHg4g at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Dear all,
> Could you provide an advice how I can generate fully relativistic
> pseudopotential for Carbon which
> would include effect of unoccupied d-orbitals?
>
> We have learned [Phys. Rev. B 80, 235431 (2009)] using fully potential
> linearized augmented
> plane wave techniques (FPLAPW: Wien2k, FLEUR) that spin-orbit coupling
> leads to band gap
> opening in graphene at K-point of about 24 micro eV. If d-orbitals are
> not taken into account
> the band gap shrinks below 1 micro eV. The band gap I have obtained by
> using QE v5.0 and
> generating a fully relativistic pseudopotential with ld1.x for Carbon
> atom is 0.96 micro eV, which
> agrees well with FPLAPW techniques when d-orbitals are not considered.
>
> Is it possible to generate fully relativistic pseudopotential for
> Carbon atom which would include
> effects of d-orbitals for spin-orbit coupling induced band gap in graphene?
>
> Best regards,
> Martin Gmitra
>
>
> ------------------------------
>
> Message: 4
> Date: Mon, 18 Jun 2012 08:50:52 +0200
> From: Lorenzo Paulatto<lorenzo.paulatto at impmc.upmc.fr>
> Subject: Re: [Pw_forum] Regarding calculation of nmr spectra using
> 	gipaw.x
> To: PWSCF Forum<pw_forum at pwscf.org>
> Message-ID:
> 	<CAG+GtJeM7Oxi1G9JfUJCPpOLUocAskLzG-ZaPGYk48QqbHH2jw at mail.gmail.com>
> Content-Type: text/plain; charset="iso-8859-1"
>
> On Sat, Jun 16, 2012 at 8:00 AM, Prasenjit Ghosh<prasenjit.jnc at gmail.com>wrote:
>
>> while calculating the nmr spectra using gipaw, is it necessary to use the
>> paw pseudopotential for all the atoms or is it sufficient if we use the paw
>> pseudopotential for the element for which we want to calculate the spectra?
>>
>>
> Dear Prasenjit,
> I'm quite sure that the latter is true, i.e. only for the atoms whom
> spectra you wish to calculate. However, they do not need to be PAW, but
> they must include GIPAW data for wavefunction reconstruction. Even if they
> are PAW, the keyword use_paw_as_gipaw must have been specified at
> generation time in order to have the reconstruction data stored in the file.
>
> If you have some PAW dataset without reconstruction data, it is trivial to
> regenerate them with the data. Even if you don't have the ld1.x input file,
> all the necessary input variables are in the file's header.
>
> bests
>


-- 
* * * *
Matteo Calandra, Directeur de Recherche (DR2)
Institut de Mineralogie et de Physique des Milieux Condenses de Paris
Universite Pierre et Marie Curie, tour 23, 3eme etage, case 115
4 Place Jussieu, 75252 Paris Cedex 05 France
Tel: +33-1-44 27 52 16       Fax: +33-1-44 27 37 85
http://www.impmc.jussieu.fr/~calandra



More information about the users mailing list