Uh, I see: you are using tetrahedra, which means that "degauss=0" in the pw.x run. When you run projwfc.x, unless you specify degauss explicitly, that is read from the output of the pw.x run. Hence, your PDOS calculation may be running with Gaussian smearing and degauss=0 which is nonsense. You must set degauss explicitly in projwfc.x, like to 0.007Ry (0.1eV). Conversely, the tetrahedron method is used in dos.x.
Have a look in INPUT_PROJWFC below "Important notices". Guido On 10/13/2012 09:12 AM, Cao TF wrote: > Dear Guido > Thank you very much for you kind reply. I have summed the PDOS of all > the atoms. The additional peaks still exist. I suspect it may be related > with the way how the PDOS are calculated in the PWSCF. In the Total DOS > and PDOS calculations the "smearing = tetrahedra" is used. And the > Kponits was set to 8 8 1 0 0 0 for the 6X6 supercell of graphene. So are > there any other parameters may be specially tested in the PDOS > calculation. Any suggestions will be greatly appreciated ! -- Guido Fratesi Dipartimento di Scienza dei Materiali Universita` degli Studi di Milano-Bicocca via Cozzi 53, 20125 Milano, Italy