ABINIT Discussion Forums

Hi, I try to calculate phonon frequencies of the cubic phase of BaF2. all seems to be ok, but for the phonon wavevector .25 .25 0 the two lower bands give negative phononfrequencies of around -350 cm-1. I don't understand why. Can anybody help me? The input file is attached. Thanks in advance, christian.

This is probably not acoustic sum rule breaking, for once, so either the phase is genuinely unstable (ok, that won't really be the case) or you are not converged wrt some parameter...

Try doing just the soft phonon with even higher ecut/nkpt/nband. Btw, have you looked at the phonon scf convergence, to make sure it was ok? Which pseudopotentials are you using?

Matthieu

Thank you very much. I used for Ba the LDA Teter extended norm-conserving psp and for F the Troullier-Martins. Other pseudopotentials gave single negative phonon frequencies, too. What do you mean with "phonon scf convergence"? I checked the other convergences but the results did not change significantly. It is worthy to try a much higher value for nband than 12 in this case? greetings, christian

By phonon convergence I meant the D2ETOT convergence of ETOT - check the log.