Dear I. Camps, Sorry for the late answer, just read your message now.
I guess the "SpinPolatized T" line is not sufficient in itself. Try several values for Spin.Total as well (the optimal value might be sensitive to your actual structures) and check in the outputs the (seemingly automatically set) value of Spin.Fix too. Set Spin.Fix in the FDF if necessary. NOTE-1: I may use another version of SIESTA (4.1.5), as I specify "Spin polarized" rather than SpinPolarized T". I hope you can find Spin.Total or its equivalent in your SW version, too. NOTE-2: I'm new to SIESTA, thus I do not know whether the minimum energy spin multiplicity is found by SIESTA automatically, or not. If yes, just make sure it does so (settings). If not, scan them . SCF is sensitive to the reliability of the applied multiplicity, while default mixer params work in most cases (or weight=0.1, history=8). On the other hand, DFT is a single configuration theory and you can introduce a little static correlation like improvement by using nonzero ElectronicTemperature (eg. 300 K or more). Too much of it, I believe, is not reliable. Try as low as enough for the SCF to converge in ALL of your simulations (affects Etot thus heat of reactions, as well). I think it is OK up to 1500 K or so. In addition, consider the appropriate spin multiplicities to be applied for the pure Ni structures in accordance with the other structures of your project, as eg. heats of reactions and other derived data should be reliable, too. I typically start by the pure metals and use their minimum energy multiplicities also for the derived structures -afterwards. Do not forget to recalculate spin for multiplied slabs (if any), as multiplicity is sort of extensive. I hope this helps, t On Sun, Aug 15, 2021 at 10:00 PM El-abed Haidar <ehai2...@uni.sydney.edu.au> wrote: > > No problem! Still though you would like to do a convergence test for Mesh cut > off because you might get surprised that you do not need that much energy > even when including the metals. > > Really eager to see the update of this work. > > Good luck! > > EL-abed > > > > El-abed Haidar | Doctor of Philosophy (Science) > Condensed Matter Theory (CMT) Group| School of Physics > THE UNIVERSITY OF SYDNEY | NSW | 2006 > > > > From: I. Camps > Sent: Sunday, 15 August 2021 6:04 AM > To: siesta-l@uam.es > Subject: Re: [SIESTA-L] << SCF convergence issues >> > > > > Thanks @Nick and @El-abed for your contribution. > > > > I tested with the following combinations: > > > > DM.MixingWeight: 0.25, 0.05, 0.02 and 0.01 > > SCF.Mixer.History: 1, 3, 6 and 9 > > SCF.Mixer.Variant: original and GR > > SCF.Mixer.Kick: 3, 10 and 50 > > > > The high value of the mesh cut off is due to this calculation being part of a > bigger study including two other metals (Cd and Pb) interacting with > boron-nitride nanotubes. This value was obtained after convergence study. > > > > The only thing that makes the system converge is changing DZP by SZP. > > > > []'s, > > Camps > > > > > > On Fri, Aug 13, 2021 at 5:00 PM Nick Papior <nickpap...@gmail.com> wrote: > > What did you play with for mixing weights and kick etc? > > > > A mixing weight of 0.25 is pretty high, if not excessively high in this case. > :) > > > > Did you try, say 0.02, or something like that? > > > > Also, kicks are *only* necessary when you have problems with stalls in > convergence. If you stall after 50 SCF, you should have a kick at that point, > but definitely not at every 3rd SCF, that may worsen your convergence. > > > > Den tor. 12. aug. 2021 kl. 22.02 skrev I. Camps <ica...@gmail.com>: > > Hello SIESTers, > > > > I am trying to optimize some 4 atom Ni clusters. There are 4 types: one > linear, one zigzag, one 2D and one 3D. > > > > Using the setup below, I don't get the calculations to converge even after > 1000 SCF steps (which is huge!). > > > > I already play with DM.MixingWeight, SCF.Mixer.History, SCF.Mixer.Variant, > SCF.Mixer.Kick, and nothing works. > > > > The tolerance is not even high, it is only 1E-3. > > > > The input is below. > > > > Any ideas, suggestions? > > > > Regards, > > > > Camps > > > > ########### > > SystemName NiM4-1Dlinear > SystemLabel NiM4-1Dlinear > > LatticeConstant 12.787740 Ang > %block LatticeVectors > 1.394587 0.000000 0.000000 > 0.000000 1.394587 0.000000 > 0.000000 0.000000 1.000000 > %endblock LatticeVectors > > NumberOfSpecies 1 > NumberOfAtoms 4 > > %block ChemicalSpeciesLabel > 1 28 Ni > %endblock ChemicalSpeciesLabel > > AtomicCoordinatesFormat Ang > > %block AtomicCoordinatesAndAtomicSpecies > 8.9168000 8.7888600 9.6866000 1 > 8.9168000 8.7888600 7.5067061 1 > 8.9168000 8.7888600 3.1011000 1 > 8.9168000 8.7888600 5.2809939 1 > %endblock AtomicCoordinatesAndAtomicSpecies > > > > -- SELF-CONSISTENT FIELD -- > PAO.BasisSize DZP > MD.TypeOfRun CG > MD.NumCGsteps 1000 > MinSCFIterations 3 > MaxSCFIterations 1000 > SpinPolarized T > MeshCutoff 500 Ry > DM.MixingWeight 0.25 > SCF.Mixer.History 6 # replace DM.NumberPulay > #SCF.Mixer.Variant GR > SCF.Mixer.Kick 3 > DM.Tolerance 0.001 > XC.functional GGA > XC.authors PBE > SolutionMethod diagon > > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the European > H2020 MaX Centre of Excellence (http://www.max-centre.eu/) > > > > > -- > > Kind regards Nick > > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the European > H2020 MaX Centre of Excellence (http://www.max-centre.eu/) > > > > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the European > H2020 MaX Centre of Excellence (http://www.max-centre.eu/)
-- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)
