Is it possible to study/converge magnetic surfaces? repost..
marci
marc... at pci.uzh.ch
Wed Oct 26 13:30:06 UTC 2011
Dear Valerio,
I tried your system, 5x5 Ni(111) slab, and I could converge the
electronic structure by using more or less the same settings that were
in your input.
It needs many iterations and the energy keeps oscillating for a long
time before the algorithm can find a good minimum.
However, what is really annoying, is that at the end the electrons are
redistributed between the two spins in such a way that the final
magnetization is zero, in spite of the fact that the initial guess had
a high multiplicity.
It seems that with the present settings and system size, the algorithm
finds a minimum with no magnetization, and this should be also the
reason why starting from a magnetization different from zero it takes
such a long time to converge.
It is possible that one problem is the size of the system. One should
check larger boxes to verify that.
What I can tell for sure is that the optimization of the bulk (216
atoms) electronic structure, gives the expected magnetization (~0.6
magneton per unit cell) by using more or less the same SCF set up.
best
marcella
On Oct 24, 5:37 pm, Valerio Bellini <valerio... at unimore.it>
wrote:
> Dear all,
> I am trying to simulate with CP2K-Quickstep a magnetic surface, in
> particular
> a slab of 4 layers of Ni(111).
> I have not manage to converge the system. Here are my considerations:
>
> -) the two dimensional cell is not large (5x5), and I tried hexagonal
> and rectangular
> two-dimensiional cells. Since no k-points are considered in CP2K, I know
> that
> this cell is rather small, and in order to have converged properties one
> should
> use a larger one.
> -) I used the mixing/diagonalization options suggested for metallic
> surfaces
> (recent message from Marcella Iannuzzi on the 4th October).
> -) I tried to vary the ALPHA,BETA,NBROYDEN parameters. I also tried to use
> direct mixing+DIIS scheme. Things did not change much.
> -) I managed to converge the system only with OT scheme, but although I
> converge
> the system up to 10-7, the results does not seem to me converged to the
> ground state,
> since the magnetic moments of different atoms in the same layer are not
> the same,
> while they should be due to symmetric reasons.
>
> I made a step back and converged an isolated Ni(111) monolayer, and in that
> case managed to converge with OT but also with standard
> diagonalization+broyden
> mixing, and with the latter scheme I was able to reach always a
> converged system
> in terms of equal magnetic moments for all the atoms.
>
> I both case (ML and 4 layers slab) I set a multiplicity which is reasonable
> considering the total moment in the cell that the system should acquire
> (in case of the ML, I varied also the multiplicity and found the one
> giving the
> ground state, while for the 4 layers slab I set it to a reasonable value,
> could easily be it is not the ground state one).
>
> My questions are:
>
> 1) Are there chances that increasing the two-dimensional cell dimensions,
> convergence will be reached also for the slab?
> Before trying I would like to have an opinion on that; in other words,
> could the dimension
> of the two-dimensional cell be the responsible of the missing
> convergence with
> diagonalization+broyden techniques?
>
> 2) Have anybody ever managed to converged a magnetic surface with this
> code?
>
> Here is a typical input file I used:
>
> &GLOBAL
> PROJECT ./working
> RUN_TYPE ENERGY_FORCE
> PRINT_LEVEL MEDIUM
> &END GLOBAL
> &FORCE_EVAL
> METHOD Quickstep
> &DFT
> BASIS_SET_FILE_NAME ./BASIS_MOLOPT
> POTENTIAL_FILE_NAME ./GTH_POTENTIALS
> RESTART_FILE_NAME ./working-RESTART.wfn
> LSD T
> MULTIPLICITY 71
> &MGRID
> CUTOFF 500
> NGRIDS 5
> &END MGRID
> &QS
> EXTRAPOLATION PS
> EXTRAPOLATION_ORDER 3
> &END QS
> &SCF
> SCF_GUESS restart
> EPS_SCF 1.0E-7
> MAX_SCF 500
> &OUTER_SCF ON
> MAX_SCF 20
> EPS_SCF 1.0E-7
> &END OUTER_SCF
> ADDED_MOS 1000
> &SMEAR ON
> METHOD FERMI_DIRAC
> ELECTRONIC_TEMPERATURE [K] 300
> &END SMEAR
> &DIAGONALIZATION ON
> ALGORITHM STANDARD
> &END DIAGONALIZATION
> &MIXING ON
> METHOD BROYDEN_MIXING
> ALPHA 0.05
> BETA 1.5
> NBROYDEN 8
> &END MIXING
> &END SCF
> &XC
> &VDW_POTENTIAL
> POTENTIAL_TYPE PAIR_POTENTIAL
> &PAIR_POTENTIAL
> REFERENCE_FUNCTIONAL PBE
> TYPE DFTD3
> PARAMETER_FILE_NAME ./dftd3.dat
> &END PAIR_POTENTIAL
> &END VDW_POTENTIAL
> &XC_FUNCTIONAL
> &PBE
> &END PBE
> &END XC_FUNCTIONAL
> &XC_GRID
> &END XC_GRID
> &END XC
> &END DFT
> &SUBSYS
> &CELL
> PERIODIC XY
> ABC 12.45416482 12.45416482 40
> ANGLES 90 90 120
> &END CELL
> &COORD
> Ni 0.000000 0.000000 7.797460
> Ni -1.245416 2.157125 7.797460
> Ni -2.490833 4.314249 7.797460
> Ni 2.490833 -4.314249 7.797460
> Ni 1.245416 -2.157125 7.797460
> Ni 2.490833 0.000000 7.797460
> Ni 1.245416 2.157125 7.797460
> Ni 0.000000 4.314249 7.797460
> Ni 4.981666 -4.314249 7.797460
> Ni 3.736249 -2.157125 7.797460
> Ni 4.981666 0.000000 7.797460
> Ni 3.736249 2.157125 7.797460
> Ni 2.490833 4.314249 7.797460
> Ni 7.472499 -4.314249 7.797460
> Ni 6.227082 -2.157125 7.797460
> Ni -4.981666 0.000000 7.797460
> Ni -6.227082 2.157125 7.797460
> Ni -7.472499 4.314249 7.797460
> Ni -2.490833 -4.314249 7.797460
> Ni -3.736249 -2.157125 7.797460
> Ni -2.490833 0.000000 7.797460
> Ni -3.736249 2.157125 7.797460
> Ni -4.981666 4.314249 7.797460
> Ni 0.000000 -4.314249 7.797460
> Ni -1.245416 -2.157125 7.797460
> Ni 0.000000 1.438083 9.831216
> Ni -1.245416 3.595208 9.831216
> Ni 3.736250 -5.033291 9.831216
> Ni 2.490833 -2.876166 9.831216
> Ni 1.245417 -0.719042 9.831216
> Ni 2.490833 1.438083 9.831216
> Ni 1.245417 3.595208 9.831216
> Ni 6.227082 -5.033291 9.831216
> Ni 4.981666 -2.876166 9.831216
> Ni 3.736250 -0.719042 9.831216
> Ni 4.981666 1.438083 9.831216
> Ni 3.736250 3.595208 9.831216
> Ni 8.717915 -5.033291 9.831216
> Ni 7.472499 -2.876166 9.831216
> Ni 6.227082 -0.719042 9.831216
> Ni -4.981666 1.438083 9.831216
> Ni -6.227082 3.595208 9.831216
> Ni -1.245416 -5.033291 9.831216
> Ni -2.490833 -2.876166 9.831216
> Ni -3.736249 -0.719042 9.831216
> Ni -2.490833 1.438083 9.831216
> Ni -3.736249 3.595208 9.831216
> Ni 1.245417 -5.033291 9.831216
> Ni 0.000000 -2.876166 9.831216
> Ni -1.245416 -0.719042 9.831216
> Ni 1.245416 0.719042 11.864973
> Ni -0.000000 2.876166 11.864973
> Ni -1.245417 5.033291 11.864973
> Ni 3.736249 -3.595208 11.864973
> Ni 2.490833 -1.438083 11.864973
> Ni 3.736249 0.719042 11.864973
> Ni 2.490833 2.876166 11.864973
> Ni 1.245416 5.033291 11.864973
> Ni 6.227082 -3.595208 11.864973
> Ni 4.981666 -1.438083 11.864973
> Ni -6.227082 0.719042 11.864973
> Ni -7.472499 2.876166 11.864973
> Ni -8.717915 5.033291 11.864973
> Ni -3.736250 -3.595208 11.864973
> Ni -4.981666 -1.438083 11.864973
> Ni -3.736250 0.719042 11.864973
> Ni -4.981666 2.876166 11.864973
> Ni -6.227082 5.033291 11.864973
> Ni -1.245417 -3.595208 11.864973
> Ni -2.490833 -1.438083 11.864973
> Ni -1.245417 0.719042 11.864973
> Ni -2.490833 2.876166 11.864973
> Ni -3.736250 5.033291 11.864973
> Ni 1.245416 -3.595208 11.864973
> Ni -0.000000 -1.438083 11.864973
> Ni 0.000000 0.000000 13.898730
> Ni -1.245416 2.157125 13.898730
> Ni -2.490833 4.314249 13.898730
> Ni 2.490833 -4.314249 13.898730
> Ni 1.245416 -2.157125 13.898730
> Ni 2.490833 0.000000 13.898730
> Ni 1.245416 2.157125 13.898730
> Ni 0.000000 4.314249 13.898730
> Ni 4.981666 -4.314249 13.898730
> Ni 3.736249 -2.157125 13.898730
> Ni 4.981666 0.000000 13.898730
> Ni 3.736249 2.157125 13.898730
> Ni 2.490833 4.314249 13.898730
> Ni 7.472499 -4.314249 13.898730
> Ni 6.227082 -2.157125 13.898730
> Ni -4.981666 0.000000 13.898730
> Ni -6.227082 2.157125 13.898730
> Ni -7.472499 4.314249 13.898730
> Ni -2.490833 -4.314249 13.898730
> Ni -3.736249 -2.157125 13.898730
> Ni -2.490833 0.000000 13.898730
> Ni -3.736249 2.157125 13.898730
> Ni -4.981666 4.314249 13.898730
> Ni 0.000000 -4.314249 13.898730
> Ni -1.245416 -2.157125 13.898730
> &END COORD
> &KIND Ni
> POTENTIAL GTH-PBE-q18
> BASIS_SET DZVP-MOLOPT-SR-GTH
> &END KIND
> &END SUBSYS
> &END FORCE_EVAL
More information about the CP2K-user
mailing list