bug symmetrization charge density?
Posted: Mon Aug 29, 2016 9:00 pm
I am obtaining asymmetric charges on two Ru atoms in Ru2O4 that are symmetry equivalent. However, when I use ISYM=0 (instead of the default ISYM=2) I get the correct symmetric result. Could there be a bug in the charge density symmetrization in vasp?
Thanks, Tom
PS I uploaded the POSCAR, KPOINTS and INCAR files with/without ISYM=0. I used the default vasp version installed on cori.nersc.gov (version 5.4.1-gcc) and the PBE pseudopotentials from version 5.3.5.
From the run without ISYM=0 I find in OUTCAR:
total charge
# of ion s p d tot
----------------------------------------
1 0.575 0.979 6.618 8.171
2 0.582 1.001 6.472 8.054
From the run with ISYM=0 I find in OUTCAR:
total charge
# of ion s p d tot
----------------------------------------
1 0.578 0.990 6.545 8.113
2 0.578 0.990 6.544 8.112
Thanks, Tom
PS I uploaded the POSCAR, KPOINTS and INCAR files with/without ISYM=0. I used the default vasp version installed on cori.nersc.gov (version 5.4.1-gcc) and the PBE pseudopotentials from version 5.3.5.
From the run without ISYM=0 I find in OUTCAR:
total charge
# of ion s p d tot
----------------------------------------
1 0.575 0.979 6.618 8.171
2 0.582 1.001 6.472 8.054
From the run with ISYM=0 I find in OUTCAR:
total charge
# of ion s p d tot
----------------------------------------
1 0.578 0.990 6.545 8.113
2 0.578 0.990 6.544 8.112