Heey Asiyeh Shokri

Well, VASP cannot directly simulate Raman spectra as far as I am aware of. You can, however, use VASP to calculate vibrational frequencies (phonon modes) and then use third-party tools like Phonopy or some available scripts to estimate Raman intensities based on those modes.

Best,

Gabriel Vinicius

Hello Gabriel Vinicius

Thanks for your respond. Actually I know that we can get vibrational frequencies and using phonopy or raman_sc we can get the raman spectroscoy. But I haven't seen such work on doped structures.

Regards,

Asiyeh

Huh, got you now.

Maybe it's because doping can introduce localized states, distort the lattice, or cause disorder and mixed phases, substantially increasing the complexity? I thought it was technically possible, but as you haven’t found anyone doing it, it seems more like a computational challenge in terms of complexity, requiring larger supercells, more precise convergence or something else.

Interesting..

Experimentally, there are overlapping effects which make the reconstruction of shifts by DFT pretty difficult.

On one side, you have changes to the lattice constant for the "quasi-ideal" doped crystal. This is an effect that you may be able to reconstruct by large supercells as suggested by Gabriel Vinicius . The required dopant concentrations for substantiate shifts may be quite large in this regard, see e.g.

Article Quantification of Yttria in Stabilized Zirconia by Raman Spectroscopy

so some people might start speaking about stoichiometry instead of doping.

On the other side, getting the dopant into a solid may require changes to the growth conditions, especially when we're talking about epitaxial layers. In our work (diamond growth) the introduction of a dopant may even swap the growth stress from compressive to tensile or vice versa, and that also swaps the shift of associated Raman lines.

Jürgen Weippert Thank you for your respond.