VASP-CASTEP Optics Discrepancy
Posted: Wed Dec 19, 2012 4:05 am
Dear all,
I would like to calculate the optical properties of a narrow bandgap system (< 0.5 eV). I followed the following steps with VASP 5.3.2
-relaxed the system to a high energy tolerance < 1e-5 eV.
-Single energy point self consistent run with a minimal number of bands and a large number of KPOINTS to accurately calculate the charge density (EDIFF<5e-10).
-Did a non-self-consistant run with a large number of bands (factor of 3 times more) and a large number of KPOINTS using the ICHRAGE=11 and LOPTIC=TRUE.
I know that VASP will not give realistic results for wavelengths >1um (energies <1 eV) because it does not handle intra band transitions. But I would expect it to give realistic results for visible wavelengths (energies > 1eV).
When I compare the calculated refractive index results with experiment, the shape of the dispersion curve and the values of the refractive index is not at all accurate. However, if I take the same coordinates and run the calculation using CASTEP. The shape of the refractive index dispersion curve and values are much more realistic.
Can anybody offer some advice on how I can use VASP to calculate the dielectric function accurately for above band gap energies?
Thank you.
I would like to calculate the optical properties of a narrow bandgap system (< 0.5 eV). I followed the following steps with VASP 5.3.2
-relaxed the system to a high energy tolerance < 1e-5 eV.
-Single energy point self consistent run with a minimal number of bands and a large number of KPOINTS to accurately calculate the charge density (EDIFF<5e-10).
-Did a non-self-consistant run with a large number of bands (factor of 3 times more) and a large number of KPOINTS using the ICHRAGE=11 and LOPTIC=TRUE.
I know that VASP will not give realistic results for wavelengths >1um (energies <1 eV) because it does not handle intra band transitions. But I would expect it to give realistic results for visible wavelengths (energies > 1eV).
When I compare the calculated refractive index results with experiment, the shape of the dispersion curve and the values of the refractive index is not at all accurate. However, if I take the same coordinates and run the calculation using CASTEP. The shape of the refractive index dispersion curve and values are much more realistic.
Can anybody offer some advice on how I can use VASP to calculate the dielectric function accurately for above band gap energies?
Thank you.