ABINIT Discussion Forums

Dear users, Hi
I performed a phonon calculation on a full relaxed structure at the Gamma point.
After doing ANADDB, in the output file, the violation of charge neutrality for one displacement and direction is obtained about 0.19 and for most of the directions, the violations are zero or of order 0.002. Is this violation a critical error or I can ignore it?
My input and output files are attached.

Thanks
Masoud

Dear Masoud,
The charge neutrality is indeed violated by about 0.2e but your Born charges are around 3e, meaning that the violation is about 6%, which is fair.
It depends how much precision you want for the Born charges for example, or related properties, for example if you want to analyse variation of Born charges that are less or around 0.2e then you should improve the precision.
If you want to improve it, you can increase the number of k-point in the direction it is deviating the most.
Best wishes,
Eric

Dear Eric,
Thanks for your answer.
According to your suggestion, I increased the k-point (ngkpt from 6 8 6 to 6 14 6) and the violation of charge was decreased (from 0.2e to 0.07e).
Although this change has a small effect on the phonon frequency, the electronic contribution of the dielectric function has a considerable change.
Also, three acoustic phonons have now a small negative value. My previous acoustic phonons were exactly zero. Is this trend normal or I must try with more k-point?

Thanks,
Masoud