I am calculating the band structure of S doped Strontium Titanate using vasp. After relaxing the system (using ISIF=3), the structure gets heavily changed, and Klabels and Kpath become different from the cubic pure supercell. but I have noticed in some papers that they are showing the same Klabels in both doped and undoped structures. What could be the possible reasons for different Klabels in my case?
Since you are relaxing the unit cell (ISIF=3), it changes and the related K-path also changes accordingly. Other papers might have not relaxed the unit cell. Doing only volume relaxation (ISIF=7) can also lead to the same K-path. However, maintaining the same K-path for doped samples may not necessarily be the best approach.
Ehsan Moradpur-Tari Thank you for your response. What would be your opinion if I initially give Kpath of the undoped structure in the KPOINTS file for bandstructure calculation? Would it be a correct approach? I am asking this because a friend of mine suggested me that step, but I was worrying if it would be valid.
Actually a lot of papers use this approach for the sake of better comparison of band structures. It is valid provided that the symmetry of the unit cell doesn't undergo any changes. For example, if its tetragonal, after doping it should remain tetragonal, however volumetric changes are acceptable.
I agree with Ehsan Moradpur-Tari but with doping, change of positions is also highly possible and ISIF = 7 doesn't include position change, so I believe Tarique Hasan Shushmoy relax the system with ISIF = 2, with finite NSW. That may help!
Payal Wadhwa thank you for your response and advice. I am going to try with isif=2.
- Badges
- Science topic
- Similar topics
- Journal Impact Factor