[Pw_forum] Ru_atomic_calculation

[Kai Trepte]

Hello everyone,

I'm trying right now to compute some atomic energies of different
elements. For that, I used pseudopotentials from
http://theossrv1.epfl.ch/Main/Pseudopotentials. For Ru I get a very
strange behaviour, as it does not converge at all and the absolute
magnetization over the course of the scf cycle is somewhere around 300.
For ever other element, the computation works just fine. Is it possible
that the pseudopotential file is incorrect?

My input looks as follows:

 &control
    calculation ='scf',
    restart_mode='from_scratch',
    prefix='test_Ru',
    nstep = 500,
    pseudo_dir='/projects/nano-10/data/trepte/pseudo',
    outdir='/scratch/trepte/test_Ru'
 /
 &system
    ibrav = 0
    nat = 1,
    ntyp = 1,
    ecutwfc = 200,
    nspin = 2
    starting_magnetization(1) = 0.8
    occupations = 'smearing'
    degauss = 1.0d-5
 /
 &electrons
    mixing_beta = 0.1,
    conv_thr = 1.0d-7
    electron_maxstep = 400
 /
ATOMIC_SPECIES
 Ru 101.07 Ru.pbe-spn-kjpaw_psl.0.3.0.UPF
CELL_PARAMETERS bohr
28.34590317 0.0 0.0
0.0 28.34590317 0.0
0.0 0.0 28.34590317
ATOMIC_POSITIONS angstrom
Ru   7.5     7.5     7.5
K_POINTS gamma

Thanks in advance,
Kai


Kai Trepte
Technische Universität Dresden
Theoretische Chemie
Bergstraße 66b
01062 Dresden