[QE-users] Relaxation of Fe2O3

[Matteo Cococcioni]

Dear Anuja,

I don't have experience with this system and don't know how big the total
energy is. however, your convergence threshold for electrons (1.d-6) looks
a bit loose. that might result in not so accurate forces and, as a result,
ionic/cell steps are taken in the wrong direction. Try to tighten this
threshold. Also you may want to try to reduce the trust radius of the ionic
steps so the initial updates of ionic positions are not too big.
I assume you have studied the convergence wrt all the other parameters. Are
you sure, though, that a 888 k-point mesh is enough with such a small
Fermi-Dirac smearing of the distribution function? is the system metallic
in this phase?

Best regards,

Matteo



Il giorno dom 28 lug 2019 alle ore 18:08 Anuja Chanana <
anujachanana26 at gmail.com> ha scritto:

> Dear users,
> I am trying to relax the bulk structure of alpha-Fe2O3 in the rhombohedral
> crystal structure. The ground state is +--+. However, the crystal lattice
> doesn't relax and the total energy values keep fluctuating rather than
> showing any convergence. The following is the input file
>
> &CONTROL
>     calculation = 'vc-relax',
>     restart_mode = 'from_scratch',
>     title = 'Fe2O3',
>     prefix = 'Fe2O3',
>     outdir = './tmp_Fe2O3_rhom_scf_1',
>     pseudo_dir = '.',
>     forc_conv_thr = 1.0D-5,
>     tprnfor = .true.,
>     tstress = .true.
>     verbosity = 'high'
>     wf_collect = .true.,
>     nstep = 1000,
> /
>
> &SYSTEM
>   ibrav=5,
>   celldm(1)=10.2528224724d0, celldm(4)=0.5694631780d0,
>   nat=10,
>   ntyp=5,
>   ecutwfc = 80,
>   ecutrho = 640,
>   occupations='smearing'
>   smearing='fd'
>   degauss= 0.005,
>   nspin=2,
>   starting_magnetization(1)= 0.0,
>   starting_magnetization(2) =  1.0,
>  starting_magnetization(3) =  -1.0,
>  starting_magnetization(4) =  -1.0,
>  starting_magnetization(5) =   1.0,
>   lda_plus_u = .true.
>   Hubbard_U(2) = 3.8
>   Hubbard_U(3) = 3.8
>   Hubbard_U(4) = 3.8
>   Hubbard_U(5) = 3.8
> /
>
> &ELECTRONS
>     mixing_beta = 0.2,
>     conv_thr = 1.0D-6,
> electron_maxstep=1000,
> /
>  &ions
>  /
>
>  &cell
> /
>
> ATOMIC_SPECIES
>   O 15.999400d0 O.pw91-n-rrkjus_psl.1.0.0.UPF
>   Fe1 55.845000d0 Fe.pw91-spn-rrkjus_psl.1.0.0.UPF
>   Fe2 55.845000d0 Fe.pw91-spn-rrkjus_psl.1.0.0.UPF
>   Fe3 55.845000d0 Fe.pw91-spn-rrkjus_psl.1.0.0.UPF
>   Fe4 55.845000d0 Fe.pw91-spn-rrkjus_psl.1.0.0.UPF
>
> ATOMIC_POSITIONS {crystal}
>    O   0.9444650930d0   0.5556659533d0   0.2500659533d0
>    O   0.2500650930d0   0.9444659533d0   0.5556659533d0
>    O   0.0557952789d0   0.4445978598d0   0.7501978598d0
>    O   0.4445952789d0   0.7501978598d0   0.0557978598d0
>    O   0.5556650930d0   0.2500659533d0   0.9444659533d0
>    O   0.7501952789d0   0.0557978598d0   0.4445978598d0
>   Fe1   0.1446976654d0   0.1446981632d0   0.1446981632d0
>   Fe2   0.6448278513d0   0.6448300697d0   0.6448300697d0
>   Fe3   0.3554325205d0   0.3554337433d0   0.3554337433d0
>   Fe4   0.8555627065d0   0.8555656499d0   0.8555656499d0
>
> K_POINTS {automatic}
>   8 8 8 0 0 0
>
>
> Has anyone ever tried relaxing the structure before and faced a similar
> situation?
> What is the best possible solution if the system doesn't show a smooth
> convergence.
> Looking forward to a reply.
>
> Thanks and Regards
> Anuja Chanana
>
>
>
> [image: Mailtrack]
> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> Sender
> notified by
> Mailtrack
> <https://mailtrack.io?utm_source=gmail&utm_medium=signature&utm_campaign=signaturevirality5&> 07/28/19,
> 09:35:20 PM
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso)
> 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/20190729/d47d18e6/attachment.html>