DOS plot from various method NOT same
Posted: Mon Jul 13, 2015 9:34 pm
Hi,
I'm using the GW method (vasp.5.3.5 with potpaw_PBE.52 GW version of pseudopotentials) to look at the DOS of an isolated molecule. I plotted the DOS from three methods:
(1) Adding up the atom projected DOS, that is written in the DOSCAR, to get the molecules' DOS. LORBIT=11 was used. -> Blue curve in the attached figure.
(2) Total DOS that is written in the ver beginning of the DOSCAR. -> Red Curve in the attached figure.
(3) Took the eigenvalues directly from the OUTCAR and used same range and binning to get the histogram for the molecules' DOS. -> Green Curve in the attached figure.
The link for the plot is: https://app.box.com/s/7dnabt640gbtx3w1xqv2qsxwy0j2eg0y
Q1: Why am I getting differnt plots from these three methods?
Q2: Why TDOS, i.e. (2), so different than other methods especially in the LUMO region?
Thank you for the help.
I'm using the GW method (vasp.5.3.5 with potpaw_PBE.52 GW version of pseudopotentials) to look at the DOS of an isolated molecule. I plotted the DOS from three methods:
(1) Adding up the atom projected DOS, that is written in the DOSCAR, to get the molecules' DOS. LORBIT=11 was used. -> Blue curve in the attached figure.
(2) Total DOS that is written in the ver beginning of the DOSCAR. -> Red Curve in the attached figure.
(3) Took the eigenvalues directly from the OUTCAR and used same range and binning to get the histogram for the molecules' DOS. -> Green Curve in the attached figure.
The link for the plot is: https://app.box.com/s/7dnabt640gbtx3w1xqv2qsxwy0j2eg0y
Q1: Why am I getting differnt plots from these three methods?
Q2: Why TDOS, i.e. (2), so different than other methods especially in the LUMO region?
Thank you for the help.