hard to remove nagative phonon bands when using classical potential

[quantumtoday]

I re-post my question/concern about phonon bands with classical potential here:

it seems very hard to remove negative bands of graphene if using the potential tersoff_C_2010 (even the structure was optimized). Did you succeed in using this potential?

My concern is that is there any rule when optimizing structures to be sure that negative bands will go away just by one or two times of optimization?

[Petr Khomyakov]

I would guess that it is rather difficult to say forehand if geometry optimization will resolve this issue without the actual optimization and then phonon frequency calculation for the optimized structure.

[Petr Khomyakov]

To add to my previous post, I have done phonon calculation for the graphene structure optimized using the tersoff_C_2010 classical potential, and I have not found any negative phonon frequencies for this optimized structure.

For hydrostatically-strained structure of graphene, the ATK-Classical calculation suggest that there exist a few pnonon frequencies having negative values. This might be due to the fact that strained graphene is more prone to buckling, i.e., structural large-scale instability; graphene rippling is known to take place in experiment on suspended graphene. One can not however fully rule out that there exists a bug. We will investigate this in more detail.

[quantumtoday]

Dear Petr Khomyakov,

Thank you for taking time to check this problem.

I did try calculations with 2D graphene sheet, but could not remove negative bands even the structure was relaxed (I tried several times).

I will test more, hope I can get luck.