Re: [Phonopy-users] negative frequencies

SourceForge Headquarters 1320 Columbia Street Suite 310 San Diego, CA 92101 +1 (858) 422-6466

SIGMA = 0.05 is quite a large smearing; you should check that for SIGMA 
-> 0, the forces converge. Also, EDIFFG = -0.01 is not tight at all, a 
good value would be 10^-4; however, depending on the system, you might 
have a corrugated PES and you can get the convergence with 10^-4 only by 
carefully adjusting the POTIM parameter and gradually increasing the k-mesh.

So, the answer to your question still remains open. If you don't check 
that your structure is really relaxed, then it is not possible to answer 
to your question. Keep in mind that the energy functional also plays a 
role, maybe PBE is not suitable for your system. Still, depending on 
what is the goal of your study and the physics of your system, you may 
conclude what you say from the results that you already have. I suggest 
to discuss it with your coworkers and your supervisor, if you have one.

A last observation. I can be wrong but from what you wrote, I guess you 
have a perovskite. Check if the negative frequencies correspond to 
octahedral rotations; if yes, then you have to really screen carefully 
possible different functionals other than PBE and use more stringent 
EDIFFG and SIGMA, as it is known that octahedral rotations are very 
sensitive to the simulation settings.

Best

Antonio

On 06. 04. 23 10:50, Ammar In wrote:
> Thanks for your reply.
>
> Yes I applied very tight parameters as follows:
> LREAL = .FALSE.
> PREC = Accurate
> ISMEAR = 0
> SIGMA = 0.05
> IBRION = 8
> EDIFF = 1E-8
> EDIFFG = -0.01
> ALGO = Normal
> GGA = PE
> ADDGRID = .TRUE.
>
> For my systems, these are bulk and lattice vectors for unit cell are 
> about 5.3, 5, 8.5. I have different systems but with the same range.
>
> So my question was, is it possible to ignored the small frequencies as 
> I see in previous discussions, and if the frequencies in band1 are 
> true and the structure is unstable?
>
>
> On Thu, Apr 6, 2023 at 5:39 PM Antonio Cammarata 
> <cam...@fe...> wrote:
>
>     Hello,
>
>     you didn't provide any information about the structure nor the
>     convergence. You should check that the SCF settings are OK (proper
>     energy functional, very small smearing, enabling extra grid for force
>     calculations etc.) and the tolerance on the SCF energy and forces are
>     stringent enough to obtain converged values. Still, the doubt
>     remains if
>     you have  bulk 3D structure but you have a sampling of 221, either on
>     the k-points or on the supercell. You should check the case of a
>     2x2x2
>     supercell with 2x2x2 k-mesh. You should also check the displacement
>     pattern relative to the small negative frequencies about Gamma to
>     check
>     if these are due to physical reasons or due to the lack of rotational
>     sum rules in phonopy, which can be solved by correcting the phonopy
>     results with other softwares like hiphive
>     (https://hiphive.materialsmodeling.org/advanced_topics/rotational_sum_rules.html).
>
>     Best
>
>     Antonio
>
>     On 06. 04. 23 10:19, Ammar In wrote:
>     > Hello,
>     >
>     > I have performed phonon dispersion for orthorhombic cells using
>     vasp
>     > DFPT and phonopy. I have used 221 supercells with 80
>     atoms, Kpoints 444.
>     > I got imaginary frequencies for example as attached. However, it is
>     > very small for systems 2-3 around gamma but for system1 it spreads
>     > over all points. I tested increasing the supercell to 222 with
>     > kpoint221 (due computational cost) for some of my systems, but
>     it did
>     > not help.
>     >
>     > Can these imaginary frequencies be ignored for systems2-3 and are
>     > these systems stable?
>     > I supposed system1 to be unstable.
>     >
>     > Thanks
>     >
>     >
>     > _______________________________________________
>     > Phonopy-users mailing list
>     > Pho...@li...
>     > https://lists.sourceforge.net/lists/listinfo/phonopy-users
>
>     -- 
>     _______________________________________________
>     Antonio Cammarata, PhD in Physics
>     Assistant Professor in Applied Physics
>     Advanced Materials Group
>     Department of Control Engineering - KN:G-204
>     Faculty of Electrical Engineering
>     Czech Technical University in Prague
>     Karlovo Náměstí, 13
>     121 35, Prague 2, Czech Republic
>     Phone: +420 224 35 5711
>     Fax:   +420 224 91 8646
>     ORCID: orcid.org/0000-0002-5691-0682
>     <http://orcid.org/0000-0002-5691-0682>
>     WoS ResearcherID: A-4883-2014
>
>
>
>     _______________________________________________
>     Phonopy-users mailing list
>     Pho...@li...
>     https://lists.sourceforge.net/lists/listinfo/phonopy-users
>
>
>
> _______________________________________________
> Phonopy-users mailing list
> Pho...@li...
> https://lists.sourceforge.net/lists/listinfo/phonopy-users

-- 
_______________________________________________
Antonio Cammarata, PhD in Physics
Assistant Professor in Applied Physics
Advanced Materials Group
Department of Control Engineering - KN:G-204
Faculty of Electrical Engineering
Czech Technical University in Prague
Karlovo Náměstí, 13
121 35, Prague 2, Czech Republic
Phone: +420 224 35 5711
Fax:   +420 224 91 8646
ORCID: orcid.org/0000-0002-5691-0682
WoS ResearcherID: A-4883-2014