Dear Mehmet and Prof. Andrei Postnikov, As I had mentioned earlier I had changed the vibra.h file to make maxx=10; maxy=10 and maxz=10. But now when I obtain the dispersion curve it is completely messed up. I believe that something else needs to be modified in the Vibra utility in order to accomodate more super-cells. Please could someone guide me on exactly what needs to be modified? According to Mehmets results the 5x5x1 supercell should be sufficient to give correct results.
Regards, Juzar Thingna Department of Physics, Center for Computation Science and Engineering, National University of Singapore. On Tue, Jul 27, 2010 at 2:39 PM, <apost...@uni-osnabrueck.de> wrote: > > Dear Mehmet, > > I used your input values but I still get a negative part in the curve. I > > tried re doing the calculation with LDA pseudo. For LDA there is no > > negative > > part whereas GGA (PBE) gives me negative part. > > Could someone comment on why this should happen for different pseudos. > > Dear Juzar, > I can suggest a small checklist to get to the bottom of your problem. > Forget the pseudos for the moment, they are not your principal worry. > > 1. Do you have, among Gamma-modes, three good acoustic ones, > with frequencies close enough to zero? > - if not, the problem is insufficient MeshCutoff. > > 2. Are other frequencies at Gamma positive? > - if not, you are calculating phonons not around the equilibrium > (this might be intentional, of course, in some cases). Go search > for a better equilibrium. Use eigenvectors of instable modes as a guide. > > 3. If Gamma modes are OK but negative modes appear someplace > along the dispersion curves. This can be, in principle, either > genuine (a true instability - but needs careful checking!), > but more often is due to insufficient supercell size in vibra calculations. > > I understand that this behaviour may vary with pseudopotential, but > in a collateral and random way; k-mesh and MeshCutoff convergencies > are much more important. You can go through checklist and enforce > stable phonons with any pseudopotential. (The values of stable modes > is a different issue). > > Good luck > > Andrei Postnikov > > > dont think there is a problem with the pseudo because it is well tested > > for > > other calculations.) > > > > Regards, > > > > Juzar Thingna > > Department of Physics, > > Center for Computation Science and Engineering, > > National University of Singapore. > > > > > > On Mon, Jul 26, 2010 at 9:42 PM, Mehmet Topsakal < > > topsa...@unam.bilkent.edu.tr> wrote: > > > >> Dear Jusar, > >> > >> Your structure in C2.fdf does not look like 2D-graphene. > >> > >> Try these > >> > >> > >> NumberOfSpecies 1 > >> NumberOfAtoms 2 > >> LatticeConstant 1.0007 Ang > >> %block LatticeVectors > >> 2.4644 0 0 > >> 1.2322 2.13423302 0 > >> 0 0 10 > >> %endblock LatticeVectors > >> > >> AtomicCoordinatesFormat fractional > >> %block AtomicCoordinatesAndAtomicSpecies > >> 0.33333333 0.33333333 0.5 1 > >> 0.66666666 0.66666666 0.5 1 > >> %endblock AtomicCoordinatesAndAtomicSpecies > >> > >> > >> Also the attached pdf is the phonon-spectrum that i calculated by using > >> phon > >> (http://www.homepages.ucl.ac.uk/~ucfbdxa/phon/<http://www.homepages.ucl.ac.uk/%7Eucfbdxa/phon/> > <http://www.homepages.ucl.ac.uk/%7Eucfbdxa/phon/> > >> ) > >> > >> On Mon, Jul 26, 2010 at 1:41 PM, Juzar Thingna <juza...@gmail.com> > >> wrote: > >> > >>> Hello Everyone, > >>> > >>> I'm trying to calculate the phonon dispersion relation of graphene > >>> sheet. > >>> In this calculation I get negative frequencies. Please could someone > >>> guide > >>> me where I'm going wrong? > >>> > >>> These are my input files: > >>> 1) C2.fdf > >>> > >>> SystemName graphene > >>> SystemLabel C2 > >>> > >>> NumberOfAtoms 2 > >>> LatticeConstant 2.46 Ang > >>> > >>> %block LatticeVectors > >>> 0.86 -0.5 0.0 > >>> 0.86 0.5 0.0 > >>> 0.0 0.0 20.0 > >>> %endblock LatticeVectors > >>> > >>> AtomicCoordinatesFormat NotScaledCartesianAng > >>> %block AtomicCoordinatesAndAtomicSpecies > >>> 0.0 0.0 0.0 1 12.0 > >>> 1.42 0.0 0.0 1 12.0 > >>> %endblock AtomicCoordinatesAndAtomicSpecies > >>> > >>> SuperCell_1 3 > >>> SuperCell_2 3 > >>> > >>> AtomicDispl 0.04 Bohr > >>> > >>> BandLinesScale ReciprocalLatticeVectors > >>> %block BandLines > >>> 1 0.000 0.000 0.000 \Gamma > >>> 20 0.500 0.000 0.000 M > >>> 20 0.333 -0.333 0.000 K > >>> 20 0.000 0.000 0.000 \Gamma > >>> %endblock BandLines > >>> > >>> 2) C2-siesta.fdf > >>> > >>> # graphene, 2 C in each unit cell > >>> # optimize structure(15/06/10/Tue) > >>> # cal. FC(16/06/10/Wed) > >>> # use ./vibrator to calculate phonon(16/06/10/Wed) > >>> > >>> SystemName graphene > >>> SystemLabel C2 > >>> > >>> NumberOfSpecies 1 > >>> %block ChemicalSpeciesLabel > >>> 1 6 C > >>> %endblock ChemicalSpeciesLabel > >>> > >>> PAO.BasisSize DZP > >>> XC.functional GGA > >>> XC.authors PBE > >>> > >>> MeshCutoff 200.0 Ry > >>> DM.MixingWeight 0.3 > >>> DM.NumberPulay 3 > >>> DM.Tolerance 1.0d-5 > >>> > >>> %include FC.fdf > >>> > >>> First I run the utility fcbuild followed by siesta and then the utility > >>> vibrator to get the spectrum. > >>> > >>> Regards, > >>> > >>> Juzar Thingna > >>> Department of Physics, > >>> Center for Computation Science and Engineering, > >>> National University of Singapore. > >>> > >> > >> > >> > >> -- > >> > >> Mehmet Topsakal (Ph.D. Student) > >> UNAM-Institute of Materials Science and Nanotechnology. > >> Bilkent University. 06800 Bilkent, Ankara/Türkiye > >> Tel: 0090 312 290 3527 ; Fax: 0090 312 266 4365 > >> UNAM-web : www.nano.org.tr > >> > >> > > > >
