[QE-users] Problem with PBE0 calculation
Rameswar Bhattacharjee
rb1820 at georgetown.edu
Fri Jan 13 17:35:22 CET 2023
Dear Giuseppe and M.V Kondrin,
Thank you very much for your suggestion and such a fruitful discussion.
Looks like there is no direct way to plot the band structure using a hybrid
functional. However, as you suggested this can be done with some tricks and
techniques. I never heard about the scissor operator which I need to learn
how to use. But as suggested by Mr. MV Konrin, I could get the DOS using
PBE0 which is also very useful for the band gap. Thanks a lot for such
wonderful guidance.
Rameswar
On Thu, Jan 12, 2023 at 5:20 AM mkondrin <mkondrin at hppi.troitsk.ru> wrote:
> Dear Giuseppe,
>
> Thank you for very pertinent comment. Of course, I have meant
> stoichiometric materials, not the defect structures. Indeed, in the
> latter case band structure is not well defined notion, but probably DOS
> for some supercell approximation of defect structure is more robust.
>
> Sincerely yours,
> M. V. Kondrin
>
> On 01/12/2023 01:06 PM, Giuseppe Mattioli wrote:
> >
> > Dear M. V. Kondrin
> > The "scissors operator" strategy can be appropriate in the case of
> > stoichiometric materials, but it may be not when you investigate
> > defects or dopants in a host matrix, where the band structure (and the
> > DOS) can depend on the localization of electronic charge on defects
> > that can be seen as "deep" or "shallow" by different GGA or EXX
> > functionals. The best method to use depend on Rameswar's purposes.
> > Best
> > Giuseppe
> >
> > Quoting mkondrin <mkondrin at hppi.troitsk.ru>:
> >
> >> I also performed calculations using HSE06 functional for obtaining
> >> electron DOS (using projwfc.x). It is compatible with hybrid
> >> functionals and usually provide correct values of electron bandgaps.
> >> After that I just used "scissors" operation on electron band
> >> structure obtained using standard PBE to match bandgap obtained
> >> previously.
> >>
> >> Approximately in the same words I described this operation in the
> >> recent paper which is under review now.
> >>
> >> Hope this helps.
> >>
> >> Sincerely yours,
> >> M. V. Kondrin
> >>
> >>
> >>
> >> On 01/12/2023 11:03 AM, Giuseppe Mattioli wrote:
> >>>
> >>> AFAIK, nscf calculations are not compatible with EXX functionals.
> >>> There are a few workarounds, like manually adding the band path as
> >>> "fake" k-points with weight=0 to the regular grid obtained by the
> >>> automatic distribution used in the scf calculation.
> >>> I generally perform EXX calculations of isolated or very large
> >>> systems @Gamma. More experienced users may add more accurate
> >>> information.
> >>> HTH
> >>> Giuseppe
> >>>
> >>> Quoting Rameswar Bhattacharjee <rb1820 at georgetown.edu>:
> >>>
> >>>> Hi Giuseppe,
> >>>> Thank you very much for your detailed reply. I followed the
> >>>> instructions
> >>>> and was able to complete the SCF calculation. Also, I did an NSCF
> >>>> calculation using the same combination you suggested as a pre-step to
> >>>> perform the band calculation. My ultimate aim is to get the band
> >>>> gap using
> >>>> a hybrid functional such as PBE0.
> >>>>
> >>>> However, I am not able to run the band calculation with this
> >>>> combination
> >>>> got an error like below:
> >>>>
> >>>>
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>
> >>>> Error in routine exx_grid_init (1):
> >>>>
> >>>> wrong EXX q grid
> >>>>
> >>>>
> >>>>
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>
> >>>> It seems the k-points grid I am using is not suitable for EXX
> >>>> calculation
> >>>> although I used the same grid for PBE and it was perfect. I am sure
> >>>> someone
> >>>> in the forum had performed band structure using hybrid functional
> >>>> and would
> >>>> be able to help me resolve this. Any help would be highly appreciated.
> >>>>
> >>>> Here is my input for band calculation using PBE0
> >>>>
> >>>> ------------------------------------------
> >>>>
> >>>> &control
> >>>>
> >>>> restart_mode='from_scratch'
> >>>>
> >>>> calculation='bands'
> >>>>
> >>>> prefix='Ethyne-Anthracene',
> >>>>
> >>>> pseudo_dir = '/home/rb1820/QE-PP/NC',
> >>>>
> >>>> outdir='./scratch'
> >>>>
> >>>> /
> >>>>
> >>>> &system
> >>>>
> >>>> ibrav = 0,
> >>>>
> >>>> tot_charge=0,
> >>>>
> >>>> nat=36,
> >>>>
> >>>> ntyp=2,
> >>>>
> >>>> ecutwfc = 90,
> >>>>
> >>>> ecutfock= 150,
> >>>>
> >>>> input_dft='pbe0'
> >>>>
> >>>> vdw_corr='grimme-d2',
> >>>>
> >>>> nbnd = 60
> >>>>
> >>>> /
> >>>>
> >>>> &electrons
> >>>>
> >>>> electron_maxstep = 1000
> >>>>
> >>>> mixing_beta=0.7
> >>>>
> >>>> diagonalization='david'
> >>>>
> >>>> diago_david_ndim = 2
> >>>>
> >>>> conv_thr=1.0d-8
> >>>>
> >>>> /
> >>>>
> >>>> &ions
> >>>>
> >>>> /
> >>>>
> >>>> &CELL
> >>>>
> >>>> cell_dynamics='bfgs',
> >>>>
> >>>> press = 0.0
> >>>>
> >>>> cell_dofree = 'x',
> >>>>
> >>>> /
> >>>>
> >>>> ATOMIC_SPECIES
> >>>>
> >>>> H 1.008 H_ONCV_PBE-1.2.upf
> >>>>
> >>>> C 12.0107 C_ONCV_PBE-1.2.upf
> >>>>
> >>>>
> >>>> ATOMIC_POSITIONS {angstrom}
> >>>>
> >>>> K_POINTS { crystal }
> >>>>
> >>>> 11
> >>>>
> >>>> -0.5 0.0 0.0 1.0
> >>>>
> >>>> -0.4 0.0 0.0 1.0
> >>>>
> >>>> -0.3 0.0 0.0 1.0
> >>>>
> >>>> -0.2 0.0 0.0 1.0
> >>>>
> >>>> -0.1 0.0 0.0 1.0
> >>>>
> >>>> 0.0 0.0 0.0 1.0
> >>>>
> >>>> 0.1 0.0 0.0 1.0
> >>>>
> >>>> 0.2 0.0 0.0 1.0
> >>>>
> >>>> 0.3 0.0 0.0 1.0
> >>>>
> >>>> 0.4 0.0 0.0 1.0
> >>>>
> >>>> 0.5 0.0 0.0 1.0
> >>>>
> >>>>
> >>>> CELL_PARAMETERS {angstrom}
> >>>>
> >>>> 6.933462797 0.000000000 0.000000000
> >>>>
> >>>> 0.000000000 20.000000000 0.000000000
> >>>>
> >>>> 0.000000000 0.000000000 20.000000000
> >>>> ---------------------------------------------------------------
> >>>>
> >>>>
> >>>> Thanks
> >>>>
> >>>> Rameswar Bhattacharjee
> >>>> Georgetown University
> >>>>
> >>>> On Wed, Jan 11, 2023 at 3:21 AM Giuseppe Mattioli <
> >>>> giuseppe.mattioli at ism.cnr.it> wrote:
> >>>>
> >>>>>
> >>>>> Dear Rameswar Bhattacharjee
> >>>>> I don't know if this is the actual source of error, but I strongly
> >>>>> recommend to use norm-conserving pseudopotentials when performing EXX
> >>>>> calculations. You have generally no speed-up with US because ecutrho
> >>>>> is more or less often the same, and in your case you are using
> >>>>> tremendously high (and likely useless) values for ecutwfc and
> >>>>> ecutrho.
> >>>>> Moreover, using NC you can finely tune the density basis set used to
> >>>>> build Fock integrals with ecutfock, otherwise stuck to the ecutrho
> >>>>> value. This combination of values is probably responsible for a very
> >>>>> high memory request (the code prints an estimate at the beginning of
> >>>>> the output). For example, *using C and H ONCV pseudopotentials*, you
> >>>>> should be on the reasonably accurate side using
> >>>>>
> >>>>> ecutwfc=90.0
> >>>>> ecutfock=135.0~180.0
> >>>>> ecutrho is automatically fixed at ecutwfc*4=360.0 in the case of NC
> >>>>>
> >>>>> HTH
> >>>>> Giuseppe
> >>>>>
> >>>>> Quoting Rameswar Bhattacharjee <rb1820 at georgetown.edu>:
> >>>>>
> >>>>>> Hi Everyone,
> >>>>>> I am trying to get a band structure using a hybrid functional. As
> >>>>> suggested
> >>>>>> previously in this forum I choose PBE0 functional with pbe-ultrasoft
> >>>>>> pseudopotential. The calculation starts fine and scf iteration
> >>>>>> converges
> >>>>>> smoothly. But the job is stuck and does not complete with the
> >>>>>> last line
> >>>>> "ACE
> >>>>>> projected onto 60 (nbndproj) and applied to 60 (nbnd)
> >>>>>> bands". I am
> >>>>>> not sure what mistake I am making here or if it is just a memory
> >>>>>> issue as
> >>>>>> no error message is printing. Any suggestion would be highly
> >>>>> appreciated. I
> >>>>>> am providing last few lines of the output and my input str for
> >>>>>> reference.
> >>>>>> Thank you.
> >>>>>>
> >>>>>> Last few lines of the scf output
> >>>>>> -------------------------------
> >>>>>>
> >>>>>> highest occupied, lowest unoccupied level (ev): -3.5627
> >>>>>> -3.2294
> >>>>>>
> >>>>>>
> >>>>>> ! total energy = -312.15802972 Ry
> >>>>>>
> >>>>>> estimated scf accuracy < 3.2E-09 Ry
> >>>>>>
> >>>>>>
> >>>>>> convergence has been achieved in 12 iterations
> >>>>>>
> >>>>>>
> >>>>>> Using ACE for calculation of exact exchange
> >>>>>>
> >>>>>>
> >>>>>> EXX grid: 4188907 G-vectors FFT dimensions: ( 100, 288,
> >>>>>> 288)
> >>>>>>
> >>>>>> ACE projected onto 60 (nbndproj) and applied to 60 (nbnd)
> >>>>> bands
> >>>>>> ----------------------------------
> >>>>>>
> >>>>>> Input str
> >>>>>> ------------------------------
> >>>>>>
> >>>>>> &control
> >>>>>>
> >>>>>> restart_mode='from_scratch'
> >>>>>>
> >>>>>> calculation='scf'
> >>>>>>
> >>>>>> prefix='Test',
> >>>>>>
> >>>>>> pseudo_dir = '/home/QE-PP',
> >>>>>>
> >>>>>> outdir='./scratch'
> >>>>>>
> >>>>>> /
> >>>>>>
> >>>>>> &system
> >>>>>>
> >>>>>> ibrav = 0,
> >>>>>>
> >>>>>> tot_charge=0,
> >>>>>>
> >>>>>> nat=36,
> >>>>>>
> >>>>>> ntyp=2,
> >>>>>>
> >>>>>> ecutwfc = 140,
> >>>>>>
> >>>>>> ecutrho = 1400.0,
> >>>>>>
> >>>>>> input_dft='pbe0'
> >>>>>>
> >>>>>> vdw_corr='grimme-d2',
> >>>>>>
> >>>>>> nbnd = 60
> >>>>>>
> >>>>>> /
> >>>>>>
> >>>>>> &electrons
> >>>>>>
> >>>>>> electron_maxstep = 1000
> >>>>>>
> >>>>>> mixing_beta=0.7
> >>>>>>
> >>>>>> diagonalization='david'
> >>>>>>
> >>>>>> diago_david_ndim = 2
> >>>>>>
> >>>>>> conv_thr=1.0d-8
> >>>>>>
> >>>>>> /
> >>>>>>
> >>>>>> &ions
> >>>>>>
> >>>>>> /
> >>>>>>
> >>>>>> &CELL
> >>>>>>
> >>>>>> cell_dynamics='bfgs',
> >>>>>>
> >>>>>> press = 0.0
> >>>>>>
> >>>>>> cell_dofree = 'x',
> >>>>>>
> >>>>>> /
> >>>>>>
> >>>>>> ATOMIC_SPECIES
> >>>>>>
> >>>>>> H 1.008 H.pbe-rrkjus_psl.1.0.0.UPF
> >>>>>>
> >>>>>> C 12.0107 C.pbe-n-rrkjus_psl.1.0.0.UPF
> >>>>>>
> >>>>>>
> >>>>>> ATOMIC_POSITIONS {angstrom}
> >>>>>>
> >>>>>>
> >>>>>> K_POINTS automatic
> >>>>>>
> >>>>>> 12 1 1 0 0 0
> >>>>>>
> >>>>>>
> >>>>>> CELL_PARAMETERS {angstrom}
> >>>>>>
> >>>>>> 6.933462797 0.000000000 0.000000000
> >>>>>>
> >>>>>> 0.000000000 20.000000000 0.000000000
> >>>>>>
> >>>>>> 0.000000000 0.000000000 20.000000000
> >>>>>>
> >>>>>> ----------------
> >>>>>>
> >>>>>>
> >>>>>> Rameswar Bhattacharjee
> >>>>>>
> >>>>>> Georgetown University
> >>>>>
> >>>>>
> >>>>>
> >>>>> GIUSEPPE MATTIOLI
> >>>>> CNR - ISTITUTO DI STRUTTURA DELLA MATERIA
> >>>>> Via Salaria Km 29,300 - C.P. 10
> >>>>> I-00015 - Monterotondo Scalo (RM)
> >>>>> Mob (*preferred*) +39 373 7305625
> >>>>> Tel + 39 06 90672342 - Fax +39 06 90672316
> >>>>> E-mail: <giuseppe.mattioli at ism.cnr.it>
> >>>>>
> >>>>> _______________________________________________
> >>>>> The Quantum ESPRESSO community stands by the Ukrainian
> >>>>> people and expresses its concerns about the devastating
> >>>>> effects that the Russian military offensive has on their
> >>>>> country and on the free and peaceful scientific, cultural,
> >>>>> and economic cooperation amongst peoples
> >>>>> _______________________________________________
> >>>>> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> >>>>> users mailing list users at lists.quantum-espresso.org
> >>>>> https://lists.quantum-espresso.org/mailman/listinfo/users
> >>>>>
> >>>
> >>>
> >>>
> >>> GIUSEPPE MATTIOLI
> >>> CNR - ISTITUTO DI STRUTTURA DELLA MATERIA
> >>> Via Salaria Km 29,300 - C.P. 10
> >>> I-00015 - Monterotondo Scalo (RM)
> >>> Mob (*preferred*) +39 373 7305625
> >>> Tel + 39 06 90672342 - Fax +39 06 90672316
> >>> E-mail: <giuseppe.mattioli at ism.cnr.it>
> >>>
> >>> _______________________________________________
> >>> The Quantum ESPRESSO community stands by the Ukrainian
> >>> people and expresses its concerns about the devastating
> >>> effects that the Russian military offensive has on their
> >>> country and on the free and peaceful scientific, cultural,
> >>> and economic cooperation amongst peoples
> >>> _______________________________________________
> >>> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> >>> users mailing list users at lists.quantum-espresso.org
> >>> https://lists.quantum-espresso.org/mailman/listinfo/users
> >>
> >>
> >> _______________________________________________
> >> The Quantum ESPRESSO community stands by the Ukrainian
> >> people and expresses its concerns about the devastating
> >> effects that the Russian military offensive has on their
> >> country and on the free and peaceful scientific, cultural,
> >> and economic cooperation amongst peoples
> >> _______________________________________________
> >> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> >> users mailing list users at lists.quantum-espresso.org
> >> https://lists.quantum-espresso.org/mailman/listinfo/users
> >
> >
> >
> > GIUSEPPE MATTIOLI
> > CNR - ISTITUTO DI STRUTTURA DELLA MATERIA
> > Via Salaria Km 29,300 - C.P. 10
> > I-00015 - Monterotondo Scalo (RM)
> > Mob (*preferred*) +39 373 7305625
> > Tel + 39 06 90672342 - Fax +39 06 90672316
> > E-mail: <giuseppe.mattioli at ism.cnr.it>
> >
> > _______________________________________________
> > The Quantum ESPRESSO community stands by the Ukrainian
> > people and expresses its concerns about the devastating
> > effects that the Russian military offensive has on their
> > country and on the free and peaceful scientific, cultural,
> > and economic cooperation amongst peoples
> > _______________________________________________
> > Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> > users mailing list users at lists.quantum-espresso.org
> > https://lists.quantum-espresso.org/mailman/listinfo/users
>
>
> _______________________________________________
> The Quantum ESPRESSO community stands by the Ukrainian
> people and expresses its concerns about the devastating
> effects that the Russian military offensive has on their
> country and on the free and peaceful scientific, cultural,
> and economic cooperation amongst peoples
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20230113/06b61fcb/attachment.html>
More information about the users
mailing list