k points for bulk

[GJK]

Hi,

Is k points vs total energy is sufficient to find the global minimum of energy??

Some authors are reporting zero imaginary frequency to prove that our structure is at global minimum.How to calculate the frequencies(imaginary) using ATK???

Your reply is highly appreciated.

[Petr Khomyakov]

Total energy is not variational with respect to the number of k-points used to sample the Brillouin zone, i.e., it just converges to some value upon increasing the k-grid density. The convergence may also be rather non-monotonic, e.g., even and odd k-grids often provide a different convergence speed.

Your second question is not clear. What frequency are you talking about?

[Anders Blom]

@GJK
You seem to be mixing up concepts here. The point about negative (phonon) frequencies is a matter of structural stability, whereas the k-point convergence is a numerical issue regarding how to compute the correct total energy.
ATK can compute the phonon spectrum and you can check it for negative frequencies. Normally you perform a geometry optimization to obtain the lowest energy state, perhaps a local minimum, and then you can check that for stability. But in either case you should first find the required k-point sampling so that the total energy etc are computed accurately.