My Community

Hello all,


I am writing concerning the format of the the DOSCAR file. In particular, I am interested in evaluating the Projected DOS (PDOS) to isolate band edges as a function of position. I calculated the PDOS using the approach illustrated in the VASP manual for bulk silicon. However, when I sum up the contributions from each atom and compare the result to the total, my answers aren't close at all. I expected some discrepancies due to numerical noise but at some energy levels I may be off by as much as a factor of 2.

Does anyone have any insight into this problem

-Tom

this issue has been discussed in the forum several times: the partial DOS is calculated by weighting the DOS with the projected integrated partial ccharges, integrated over the volumes of the 'atomic spheres' (with r=RWIGS) in all methods which are not based on localized orbitals basis sets. As space cannot be filled by spheres completely, there has to be some discrepancy. Please check whether the RWIGS (given in POTCAR) are a reasonable choice for the interatomic distances in your cell.