Greetings,
I was just wondering if anyone has suggestions for how to most accurately measure phonon dispersion while implementing the adiabatic core/shell model? I have read in the literature that the fix phonon algorithm is not necessarily built for the core/shell model, and I may see extra modes or other discrepancies in the phonon spectrum due to there being a pair of coupled particles for each ion.
Is there a safest route to take here? I.e., should I implement fix phonon on only one of the cores/shells groups, both the groups, or is there a modification to the cs model I might make that would improve the dispersion curve accuracy.
Thanks!