band gap underestimated with MBJ???
Posted: Thu Oct 10, 2013 6:39 am
Hi everyone,
I have been trying MBJ calculation in VASP 5.3 using the new potential by including
in INCAR.
However I facing an interesting problem: the band gap is underestimated (opposite to what should be expected) for several systems, which only happened in VASP, Wien2k mbj give reasonable results. For example, a simple FCC latice of ZrNiSn has a band gap of ~0.5 eV with GGA, but a negative gap (~-.1eV) with MBJ. Another example is Mg2Si.
I've double checked with MBJ in WIEN2K which give a larger band gap.
Anyone has any idea on what happened?
Thanks
I have been trying MBJ calculation in VASP 5.3 using the new potential by including
Code: Select all
METAGGA = MBJ
LASPH = .TRUE.
However I facing an interesting problem: the band gap is underestimated (opposite to what should be expected) for several systems, which only happened in VASP, Wien2k mbj give reasonable results. For example, a simple FCC latice of ZrNiSn has a band gap of ~0.5 eV with GGA, but a negative gap (~-.1eV) with MBJ. Another example is Mg2Si.
I've double checked with MBJ in WIEN2K which give a larger band gap.
Anyone has any idea on what happened?
Thanks