Dear Francesco,
to follow up on Paolo's advice, I would do this: 1) freeze the molecule - i.e. do not relax it. 2) tighten the 2 thresholds a lot - conv_thr and tr2_ph (i.e. make them smaller by one order of magnitude, two, etc...). 3) use sssp pseudos: http://materialscloud.org/sssp/ The tables show exactly what is the convergence of the phonons with cutoff (for molecular solids in the case of N and O), using the scripts from http://materialscloud.org/sssp/downloads.html where you can find all the parameters we used 4) once you feel you are in control on the cutoffs and tolerance thresholds, you can start relaxing the molecule. In order to get very accurate results there, you need to have a very tight criterion on the relaxation, as well. nicola On 16/11/2017 10:09, Francesco Moriggi wrote: > Hello, i'm Francesco Moriggi, a master student at the Milan University, > i have a problem with ph.x. I am calculating NO frequency and i'm not > reaching a convergence in values compared to cutoff energy, using norm > conserving and ultrasoft pseudopotentials. Values i get are (i've put an > image too in the email): > > *Cutoff freq Norm freq US* > > ** > > *75 1879.0 1844.4* > > ** > > *80 1875.5 1858.2* > > ** > > *85 1890.6 1907.1* > > ** > > *90 1923.5 1895.2* > > ** > > *100 1889.6 1909.3* > > ** > > *110 1914.2 1893.1* > > ** > > *120 1890.5 1900.4* > > ** > > *130 1890.7 1901.0* > > ** > > *140 1914.7 1890.9* > > ** > > *150 1886.9 ////* > > With Gaussian (aug-cc-pv5Z) i've got *1889.21* cm-1. > > An example of the input of geometry optimization is: > > *&CONTROL** > ** calculation = "relax",** > ** prefix = "NO",** > ** pseudo_dir = > "/galileo/home/userexternal/fmoriggi/QE/NO/pseudoNO/ultrasoft",** > ** outdir = "./out",** > **/** > **&SYSTEM** > ** ibrav = 1,** > ** a = 20,** > ** nat = 2,** > ** ntyp = 2,** > ** nspin = 2,** > ** ecutwfc = 150.D0,** > ** ecutrho = 1200.D0,** > ** occupations = 'smearing', smearing = 'gauss', degauss=0.02,** > ** starting_magnetization(1)=-0.9,** > ** starting_magnetization(2)=1.5,** > **/** > **&ELECTRONS** > ** electron_maxstep = 500,** > ** conv_thr = 1.D-7,** > ** mixing_beta = 0.7D0,** > **/** > **&IONS** > **/** > **ATOMIC_SPECIES** > **O 16.00 O.pbe-van_ak.UPF** > **N 14.00 N.pbe-van_ak.UPF** > **ATOMIC_POSITIONS {angstrom}** > **N 1.16 0.0 0.0** > **O 0.000 0.0 0.0 0 0 0** > **K_POINTS** > **1** > **0.0 0.0 0.0 1.0* > > Later i've done an scf with a better conv_thr. This is the input. > > *&CONTROL** > ** calculation = "scf",** > ** prefix = "150",** > ** pseudo_dir = > "/galileo/home/userexternal/fmoriggi/QE/NO/pseudoNO/ultrasoft",** > ** outdir = "./out150",** > **/** > **&SYSTEM** > ** ibrav = 1,** > ** a = 20,** > ** nat = 2,** > ** ntyp = 2,** > ** nspin = 2,** > ** ecutwfc = 150.D0,** > ** ecutrho = 1200.D0,** > ** occupations = 'smearing', smearing = 'gauss', degauss=0.02,** > ** starting_magnetization(1)=-0.9,** > ** starting_magnetization(2)=1.5,** > **/** > **&ELECTRONS** > ** electron_maxstep = 500,** > ** conv_thr = 1.D-9,** > ** mixing_beta = 0.7D0,** > **/** > **&IONS** > **/** > **ATOMIC_SPECIES** > **O 16.00 O.pbe-van_ak.UPF** > **N 14.00 N.pbe-van_ak.UPF** > **ATOMIC_POSITIONS {angstrom}** > **N 1.166278 0.0 0.0** > **O 0.000 0.0 0.0 0 0 0** > **K_POINTS** > **1** > **0.0 0.0 0.0 1.0* > > And then i run the ph.x with this input: > > *phonon calculation - NO** > ** &inputph** > ** tr2_ph=4.0d-12,** > ** prefix='150'** > ** amass(1)=14.007,** > ** amass(2)=15.9994,** > ** outdir='/galileo/home/userexternal/fmoriggi/QE/NO/relax/ultra/out150'** > ** /** > ** 0.0 0.0 0.0* > > If you can help me it would be very useful to my future TiO2 based > calculations. > > > Best regards, > > Francesco Moriggi > > Webpage: http://users.unimi.it/ceotto/about_FrancescoMoriggi.html > > > > _______________________________________________ > Pw_forum mailing list > [email protected] > http://pwscf.org/mailman/listinfo/pw_forum > -- ---------------------------------------------------------------------- Prof Nicola Marzari, Chair of Theory and Simulation of Materials, EPFL Director, National Centre for Competence in Research NCCR MARVEL, EPFL http://theossrv1.epfl.ch/Main/Contact http://nccr-marvel.ch/en/project _______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
