[QE-users] [SPAM] Re: Re: Re: Re: Assistance with Convergence Issues of Surface Calculations

Paolo Giannozzi paolo.giannozzi at uniud.it
Thu Sep 19 11:59:32 CEST 2024


Actually it works, sort of: the problem is that the structure is 
breaking down. With xcrsyden, look at the initial structure, then at the 
final one when the scf step does not converge.

Paolo


On 18/09/2024 11:40, Vinay Maithani wrote:
> Dear Giuseppe,
> 
> I can not thank you enough for your detailed suggestions and observations.
> 
> For reference, the surface I am working on is based on the material 
> presented in Energy Environ. Sci., 2013, 6, 148 (DOI: 
> https://doi.org/10.1039/C2EE23355J 
> <https://doi.org/10.1039/C2EE23355J>). I took the structural details 
> from the Supporting information (https://doi.org/10.1039/C2EE23355J 
> <https://doi.org/10.1039/C2EE23355J>) (Table 3, Page 4).
> I have constructed surface slabs from this structure, and comparable 
> surface studies on solid-state electrolytes have been reported in 
> publications such as Nanomaterials 2022, 12, 2795 (DOI: 
> https://doi.org/10.3390/nano12162795 
> <https://doi.org/10.3390/nano12162795>), Mater. Adv., 2022, 3, 3143−3150 
> (DOI: 10.1039/D1MA01246K), ACS Appl. Energy Mater. 2020, 3, 11061−11072 
> (DOI: https://dx.doi.org/10.1021/acsaem.0c02033 
> <https://dx.doi.org/10.1021/acsaem.0c02033>), ACS Appl. Mater. 
> Interfaces 2020, 12, 16350−16358 (DOI: 
> https://dx.doi.org/10.1021/acsami.9b23019 
> <https://dx.doi.org/10.1021/acsami.9b23019>), and Chem. Mater. 2018, 30, 
> 3019−3027 (DOI: https://doi.org/10.1021/acs.chemmater.8b00649 
> <https://doi.org/10.1021/acs.chemmater.8b00649>).
> 
> Despite implementing your recommendations,
> (a) including simplifying the setup by removing dipole corrections and 
> focusing on the gamma point, I continue to face convergence issues.
> (b) I am using pseudopotentials from PseudoDojo, generated with the 
> ONCVPSP code (version 3.3.0, 08/16/2017, D. R. Hamann), and I have 
> raised the ecutwfc to 90 Ry, as per your suggestion. However, I still 
> have not achieved convergence.
> (c) Additionally, I have tried lowering electron_maxstep and using 
> scf_must_converge=.false., but without success, and etot_conv_thr is not 
> reached, I am attaching a typical total energy convergence trend for 
> ionic minimization using the BFGS algorithm.
> (d) Additionally, I have tried using PAW and pseudopotentials from the 
> SSSP library without any luck! convergence is not achieved in 200 scf steps.
> I greatly appreciate your continued guidance and would be grateful for 
> any further insights you may have to help resolve this issue.
> 
> Thank you again for your time and valuable assistance.
> 
> Best regards,
> Vinay
> 
> 
> 
> On Thu, Sep 5, 2024 at 6:11 PM Giuseppe Mattioli 
> <giuseppe.mattioli at ism.cnr.it <mailto:giuseppe.mattioli at ism.cnr.it>> wrote:
> 
> 
>     Dear Vinay Maithani
>     Ok, let us say "unconventional amorphous structure" as you are sure
>     that it is meaningful, for the sake of discussion, even if I don't see
>     any pattern or repeated unit in the structure and I see strange P-S
>     "molecules" and direct Al-Li and Li-Li bonds which are unconventional.
>     Only for my curiosity, can you post a link to some publication where
>     this or a similar structure is investigated?
> 
>     However, the structure is so unconventional that it is possible that
>     you aren't able to obtain convergence, no matter what you try. To give
>     it a try I suggest that you:
> 
>     1) start it simpler (no dipole correction, gamma point as the
>     structure doesn't seem periodic).
> 
>     2) If you are using NC pseudopotentials (which ones?), then raise
>     ecutwfc as there are no Li pp in my experience that is well converged
>     at 50 Ry. You don't need to indicate ecutrho, instead, when you use NC
>     PPs.
> 
>     3) Don't specify nbnd when using occupation='smearing', the code
>     provised all the orbitals you need. Try Gaussian smearing, first, as
>     it is more stable.
> 
>     4) Try to reduce electron_maxstep to a much smaller value (if it
>     doesn't converge after 50-100 step it will probably not converge after
>     2000), and use this smaller value together with
>     scf_must_converge=.false. This will force the bfgs step on an
>     unconverged wavefunction, and can be very helpful to move your system
>     out of an unlucky initial structure with a few BFGS steps. However,
>     your structure is so unconventional that even such trick may fail.
> 
>     HTH
>     Giuseppe
> 
> 
>     Quoting Vinay Maithani <vinaymaithani21 at kgpian.iitkgp.ac.in
>     <mailto:vinaymaithani21 at kgpian.iitkgp.ac.in>>:
> 
>      > Dear Giuseppe,
>      >
>      > Thank you for your feedback. I would like to clarify that this is an
>      > *asymmetric
>      > polar slab* structure of a *solid-state electrolyte*. I have
>     thoroughly
>      > checked the structure, and it appears correct based on my
>     verification.
>      >
>      > Could you please explain why you believe the slab is wrong? Your
>     insights
>      > would be helpful as I continue to address the convergence issue.
>      >
>      > On Wed, Sep 4, 2024 at 3:26 PM Vinay Maithani <
>      > vinaymaithani21 at kgpian.iitkgp.ac.in
>     <mailto:vinaymaithani21 at kgpian.iitkgp.ac.in>> wrote:
>      >
>      >> Dear Quantum ESPRESSO community,
>      >>
>      >> I hope this email finds you well.
>      >>
>      >> I am currently working on large band gap semiconductors and have
>      >> encountered persistent difficulties in achieving convergence while
>      >> calculating the surface properties of these materials. Below are the
>      >> strategies I have attempted so far:
>      >>
>      >>    - *Mixing Beta:* Adjusted up to 0.1
>      >>    - *Mixdim:* Increased to 24
>      >>    - *Smearing:* Gaussian with degauss = 0.2
>      >>    - *Nbnd:* Added a minimum of 20% more than the number of
>     Kohn-Sham
>      >>    states
>      >>    - *Ecutwfc:* Increased from 50 Ry to 70 Ry
>      >>    - *Ecutrho:* Set to 8 times Ecutwfc
>      >>    - *Electron Max Step:* Extended to 2000 steps
>      >>
>      >> Despite these efforts, my calculations still fail to converge,
>     returning
>      >> the error:
>      >> *“Convergence NOT achieved after *** iterations: stopping.”*
>      >>
>      >> I am using norm-conserving pseudopotentials and have applied dipole
>      >> corrections, but the issue persists. I have attached the input
>     and output
>      >> files, along with the plot of potential data obtained via pp.x
>      >> (plot_num=11) and averaged in the xy-plane where a discontinuity
>     occurs in
>      >> the vacuum region, available here: Google Drive link
>      >>
>     <https://drive.google.com/drive/folders/1HoSFk11TMN4qAo8jZoSvyEymK92xGu2r?usp=drive_link <https://drive.google.com/drive/folders/1HoSFk11TMN4qAo8jZoSvyEymK92xGu2r?usp=drive_link>>
>      >> .
>      >>
>      >> I would greatly appreciate any insights or suggestions you may
>     have to
>      >> help resolve this problem.
>      >>
>      >> Thank you for your time and guidance.
>      >>
>      >> Best regards,
>      >> Vinay Maithani
>      >> PhD Candidate, IIT Kharagpur
>      >> vinaymaithani21 at kgpian.iitkgp.ac.in
>     <mailto:vinaymaithani21 at kgpian.iitkgp.ac.in>
>      >>
> 
> 
> 
>     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
>     <mailto: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
>     <http://www.max-centre.eu>)
>     users mailing list users at lists.quantum-espresso.org
>     <mailto:users at lists.quantum-espresso.org>
>     https://lists.quantum-espresso.org/mailman/listinfo/users
>     <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

-- 
Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 206, 33100 Udine Italy, +39-0432-558216
X SCUOLA ENERGIE RINNOVABILI 16-21/9/2024


More information about the users mailing list