Theory says it is not, at least up to now. Recent experiment showed it is. I'm not sure to be honest. This is what i actually was looking for: Where can i read more about this "switching path"(what is meant by this ?) ? I've briefly looked up the literature and this is that people do to study ferroelectrics. Do you know a good review on this subject ?
On Tue, Oct 1, 2013 at 10:01 AM, Pablo Aguado <[email protected]> wrote: > Is this perovskite ferroelectric? If so you could calculate the > polarization along the switching path (even if it does not go through a > high symmetry phase) to get a meaningful value of P. > > Pablo > > > On Tue, Oct 1, 2013 at 3:20 PM, Andrei Buin <[email protected]> wrote: > >> Thank you Pablo. The problem is this particular pervoskite does not have >> a centrosymmetric phase. I do realize that polarization is defined modulo x >> (since momentum operator defined modulo x), then i have to a have reference >> phase(which i dont). Thanks for the corrections for the PolarizationGrid >> block. Will try it. >> >> With Best Regards, Andrei Buin. >> >> >> >> On Tue, Oct 1, 2013 at 9:13 AM, Pablo Aguado <[email protected]> wrote: >> >>> >>> Hi Andrie, >>> >>> A couple of comments, first the absolute value of the polarization per >>> se is only defined modulo a quantum of polarization, so the value you get >>> is P = P_"real" + n P_q, where n is an integer number and P_q is the >>> quantum of polarization. P_q=e/A where A is the surface area of your unit >>> cell perpendicular to the direction with respect to which you are >>> calculating P. >>> >>> Since you don't know the value of n, to get the value of P_"real" you >>> usually calculate the polarization of a reference structure which P is >>> known by symmetry (in perovskites for instance, the centrosymetric phase >>> has either 0 or P_q/2). Then substracting the P from the centrosymmetric >>> you should be able to get the polarization of the polar phase (you might >>> still have some problems to identify the "true" polarization if the P of >>> the polar phase is comparable or larger than the quantum, see this >>> introductory paper for more details http://arxiv.org/abs/1202.1831) >>> >>> Another issue is the PolarizationGrid block you are using, which is the >>> example in the siesta manual and might not be the most suitable for you >>> simulation cell. Something like this would make more sense (re-read the >>> manual entry for the meaning of this block): >>> >>> %block PolarizationGrids >>> 20 4 4 yes >>> 4 20 4 yes >>> 4 4 15 yes >>> %endblock PolarizationGrids >>> >>> Best regards, >>> >>> Pablo >>> >>> >>> On Tue, Oct 1, 2013 at 2:39 PM, Andrei Buin <[email protected]> wrote: >>> >>>> Dear Siesta forum, >>>> >>>> I have 48(pervoskite CH3NH3PbI3) atoms in the 9 Angs x 9 Angs x 12 Angs >>>> unit cell. >>>> I'm trying to compute the Berry phase polarization usign >>>> PolarizationGrids, and i get >>>> insane dipole moment of: >>>> >>>> siesta: Macroscopic polarization per unit cell (Debye): >>>> siesta: Along the lattice vectors 2742.652644 2679.284584 4191.471845 >>>> siesta: Along cartesian directions 2742.652644 2679.284584 4191.471845 >>>> >>>> >>>> >>>> Input is attached. Forces are already converged. >>>> >>>> # 6.2 General System descriptors >>>> >>>> SystemName Tetra Super Unit Cell Siesta Calculations >>>> SystemLabel Tet # Short name for naming files >>>> NumberOfSpecies 5 >>>> NumberOfAtoms 48 >>>> >>>> %block Chemical_Species_Label >>>> 1 82 Pb >>>> 2 53 I >>>> 3 7 N >>>> 4 6 C >>>> 5 1 H >>>> %endblock Chemical_Species_Label >>>> >>>> >>>> >>>> >>>> # 6.3 Basis definitions >>>> >>>> PAO.BasisSize DZP >>>> PAO.EnergyShift 65 meV >>>> #PAO.SplitNorm 0.15 >>>> PAO.SplitTailNorm true >>>> PAO.SoftDefault true >>>> PAO.OldStylePolOrbs false >>>> >>>> %block PS.KBprojectors >>>> I 4 >>>> 0 2 >>>> 2000 -2000 >>>> 1 2 >>>> 2000 -2000 >>>> 2 2 >>>> 2000 -2000 >>>> 3 1 >>>> 2000 >>>> >>>> >>>> Pb 4 >>>> 0 2 >>>> 2000 -2000 >>>> 1 2 >>>> 2000 -2000 >>>> 2 2 >>>> 2000 -2000 >>>> 3 1 >>>> 2000 >>>> >>>> N 4 >>>> 0 2 >>>> 2000 -2000 >>>> 1 2 >>>> 2000 -2000 >>>> 2 2 >>>> 2000 -2000 >>>> 3 1 >>>> 2000 >>>> >>>> H 3 >>>> 0 2 >>>> 2000 -2000 >>>> 1 2 >>>> 2000 -2000 >>>> 2 1 >>>> 2000 >>>> >>>> C 4 >>>> 0 2 >>>> 2000 -2000 >>>> 1 2 >>>> 2000 -2000 >>>> 2 2 >>>> 2000 -2000 >>>> 3 1 >>>> 2000 >>>> %endblock PS.KBprojectors >>>> >>>> >>>> >>>> %block PAO.Basis # Define Basis set >>>> Pb 3 # Species label, number of l-shells >>>> n=6 0 2 # n, l, Nzeta, Polarization, NzetaPol >>>> 0 0 >>>> n=6 1 2 P 1 # n, l, Nzeta, Polarization, >>>> NzetaPol >>>> 0 0 >>>> n=5 2 1 # SZ for 5d >>>> 0 #0 >>>> >>>> >>>> # H in C6H6 DZP >>>> #Vova >>>> H 2 0.00000 >>>> n=1 0 2 E 11.36136 0.00928 >>>> 7.72405 2.19949 >>>> n=2 1 1 E 41.15301 0.00947 >>>> 2.89938 >>>> >>>> >>>> # C in C6H6 DZP >>>> #Vova >>>> C 3 0.00000 >>>> n=2 0 2 E 39.65304 6.21693 >>>> 7.40483 4.90026 >>>> n=2 1 2 E 27.05294 3.74121 >>>> 7.88345 3.11808 >>>> n=3 2 1 E 55.60264 0.01540 >>>> 3.93573 >>>> %endblock PAO.Basis >>>> >>>> >>>> >>>> >>>> >>>> # 6.4 Lattice, coordinates, k-sampling >>>> >>>> LatticeConstant 8.96 Ang # 6.05 350Ry, 5.93845 Ang Exp, 5.936 >>>> Exp, 5.84_LDA_Zhenya >>>> >>>> %block LatticeVectors >>>> 1.0 0 0 #18A crystal + 12Avacuum >>>> 0 1.0 0 >>>> 0 0 1.4375 >>>> %endblock LatticeVectors >>>> >>>> AtomicCoordinatesFormat Ang >>>> AtomicCoorFormatOut Ang >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> >>>> %block AtomicCoordinatesAndAtomicSpecies >>>> -0.03167 -0.11487 -0.35971 1 >>>> -0.12412 -0.42211 2.92565 2 >>>> 1.85899 2.43724 -0.06194 2 >>>> 4.40361 4.31212 -0.38969 1 >>>> 6.99200 6.23198 -0.88879 2 >>>> 2.46456 6.94775 -0.15215 2 >>>> 6.36927 1.84289 0.08180 2 >>>> -0.09247 -0.14207 6.12437 1 >>>> 2.57611 1.64461 5.69084 2 >>>> 4.35503 4.31005 6.09334 1 >>>> 6.18152 6.97008 6.46046 2 >>>> 4.59235 4.49865 2.91059 2 >>>> 4.43687 3.75124 9.32494 2 >>>> 0.44143 -0.12135 9.34642 2 >>>> 3.78065 0.42371 2.39084 3 >>>> 4.82458 -0.31863 3.15069 4 >>>> 3.28270 -0.20569 1.72441 5 >>>> 4.20292 1.18515 1.82761 5 >>>> 3.08493 0.84035 3.03696 5 >>>> 5.25142 0.34763 3.90531 5 >>>> 5.60253 -0.63714 2.45202 5 >>>> 4.36520 -1.18663 3.63088 5 >>>> 4.08008 0.15264 8.72657 3 >>>> 4.82907 -0.48545 9.84708 4 >>>> 4.39959 -0.21875 7.81266 5 >>>> 4.21776 1.18640 8.72722 5 >>>> 3.05391 -0.01415 8.80689 5 >>>> 4.48385 -0.06168 10.79380 5 >>>> 5.89626 -0.28577 9.71242 5 >>>> 4.64461 -1.56270 9.82703 5 >>>> -0.85527 5.05646 2.32772 3 >>>> 0.08950 4.13502 3.02005 4 >>>> -0.81071 4.97580 1.29648 5 >>>> -0.65212 6.05774 2.55366 5 >>>> -1.84958 4.87012 2.59300 5 >>>> -0.01489 4.27318 4.09978 5 >>>> 1.10925 4.37398 2.70535 5 >>>> -0.15654 3.10553 2.74484 5 >>>> 7.04591 2.46318 6.25281 2 >>>> 1.72806 6.16802 6.24656 2 >>>> -0.59017 4.78536 8.67712 3 >>>> 0.30844 3.96725 9.53987 4 >>>> -1.16287 4.17787 8.05170 5 >>>> -0.03405 5.40998 8.05433 5 >>>> -1.22474 5.36318 9.25668 5 >>>> 0.88403 4.63525 10.18598 5 >>>> 0.97946 3.39039 8.89784 5 >>>> -0.30006 3.29480 10.15037 5 >>>> %endblock AtomicCoordinatesAndAtomicSpecies >>>> >>>> >>>> #%block GeometryConstraints >>>> #position 1 2 >>>> #%endblock GeometryConstraints >>>> >>>> #kgrid_cutoff 20 Ang >>>> >>>> #BandLinesScale pi/a >>>> #BandLinesScale ReciprocalLatticeVectors >>>> >>>> #%block BandLines >>>> #1 1 1 1 L >>>> #20 0 0 0 G >>>> #20 1.5 0 1.5 K >>>> #10 2 0 0 X >>>> #20 0 0 0 G >>>> #%endblock Bandlines >>>> >>>> %block kgrid_Monkhorst_Pack >>>> 8 0 0 0 >>>> 0 8 0 0 >>>> 0 0 8 0 >>>> %endblock kgrid_Monkhorst_Pack >>>> >>>> >>>> >>>> >>>> # 6.5 DFT, Grid, SCF >>>> >>>> XC.functional GGA >>>> XC.authors PBE >>>> #SpinPolarized true >>>> MeshCutoff 375 Ry >>>> FilterCutoff 375 Ry >>>> MaxSCFIterations 199 # Maximum number of SCF iter >>>> DM.MixingWeight 0.05 # New DM amount for next SCF cycle >>>> DM.NumberPulay 10 >>>> DM.PulayOnFile false >>>> DM.NumberKick 50 >>>> #DM.KickMixingWeight 0.1 >>>> #DM.Tolerance 0.00001 # 0.0004 >>>> #DM.EnergyTolerance 0.00005 eV >>>> >>>> >>>> >>>> >>>> >>>> #%block PolarizationGrids >>>> # 7 3 3 yes >>>> # 3 7 3 yes >>>> # 3 3 7 yes >>>> #%endblock PolarizationGrids >>>> >>>> >>>> %block PolarizationGrids >>>> 10 3 4 yes >>>> 2 20 2 no >>>> 4 4 15 >>>> %endblock PolarizationGrids >>>> >>>> >>>> >>>> # 6.6 Eigenvalue problem: order-N or diagonalization >>>> >>>> SolutionMethod diagon >>>> #Diag.DivideAndConquer false >>>> #NumberOfEigenStates 10000 # total is ~27000 SZ >>>> #OccupationFunction MP >>>> #OccupationMPOrder 1 >>>> #ElectronicTemperature 300 K >>>> #ON.ChemicalPotentialUse true >>>> >>>> >>>> >>>> # 6.7 Molecular dynamics and relaxations >>>> >>>> MD.TypeOfRun CG >>>> #MD.NoseMass 600 Ry*fs**2 >>>> >>>> MD.NumCGsteps 9950 >>>> MD.MaxCGDispl 0.15 Ang >>>> MD.MaxForceTol 0.040 eV/Ang >>>> >>>> #MD.LengthTimeStep 1 fs >>>> #MD.FinalTimeStep 10000 >>>> #MD.InitialTemperature 550 K >>>> #MD.TargetTemperature 550 K >>>> >>>> >>>> >>>> >>>> >>>> >>>> # 6.8 Parallel options >>>> >>>> #BlockSize 426 # for 16 procs #852 for 8 # DZP 6813 orbitals >>>> #ProcessorY 8 >>>> #Diag.Memory 4 >>>> #Diag.ParallelOverK true >>>> >>>> >>>> >>>> # 6.9 Efficiency options >>>> >>>> # 6.10 Output options >>>> >>>> WriteMDXmol true >>>> WriteCoorXmol true >>>> WriteMDhistory true >>>> WriteXML false >>>> #WriteForces true >>>> #WriteMullikenPop 1 >>>> >>>> >>>> >>>> >>>> # 6.11 Options for saving/reading information >>>> UseSaveData true >>>> MD.UseSaveCG true >>>> >>>> #SaveRho true >>>> #SaveDeltaRho true >>>> #SaveElectrostaticPotential true >>>> #SaveTotalPotential true >>>> #SaveIonicCharge true >>>> #SaveTotalCharge true >>>> >>>> %block LocalDensityOfStates >>>> -7.15 12.00 eV >>>> %endblock LocalDensityOfStates >>>> >>>> %block ProjectedDensityOfStates >>>> -7.0 12.0 0.05 521 eV #min, max, broaden 100meV, steps every >>>> 25meV >>>> %endblock ProjectedDensityOfStates >>>> >>>> >>>> #WriteDenchar true >>>> >>>> #%block WaveFuncKPoints >>>> # 0.0 0.0 0.0 from 1040 to 1055 # traps # HOMO 1042 >>>> #%endblock WaveFuncKpoints >>>> >>>> >>>> #OpticalCalculation true >>>> #Optical.EnergyMinimum 0 eV >>>> #ptical.EnergyMaximum 4.5 eV >>>> #Optical.Broaden 0.01 eV >>>> #Optical.NumberOfBands 1100 # LUMO 1085 >>>> >>>> >>> >>> >>> -- >>> ----------------------------------------------------------- >>> Pablo Aguado Puente >>> [email protected] >>> ------------------------------------------------------------ >>> >> >> > > > -- > ----------------------------------------------------------- > Pablo Aguado Puente > [email protected] > ------------------------------------------------------------ >
