phonon band calculation: length of unit cell must lagre than 4 for Brenner

[berlin]

when I am doing phonon band calculation for Graphene Nanoribbon by Brenner potential (ATK classical) in ATK2015.1, errors comes out:

Traceback (most recent call last):
  File "(3,3) Graphene Nanoribbon.py", line 46, in <module>
    bulk_configuration.update()
  File ".\zipdir\NL\CommonConcepts\Configurations\AtomicConfiguration.py", line 1072, in update
  File ".\zipdir\NL\Calculators\BrennerCalculator\BrennerCalculator.py", line 126, in _update
  File ".\build\lib\site-packages\ase\atoms.py", line 643, in get_potential_energy
asapserial3.AsapError: The height of the cell (2.461) must be larger than 4

What is wrong?

[Anders Blom]

The Brenner potential (from ASAP) can't be used for very small cells, it's designed for large-scale MD simulations rather. Try a Tersoff potential instead.

[berlin]

But why the Materials Studio or Gulp woks fine with Brenner ?  And their results have a good agreement with GGA/LDA plane wave DFT calculations.

[Ulrik G. Vej-Hansen]

The implementation of the Brenner potential used in ATK requires a certain distance between periodic images of the atoms. It does not affect the results in any way, and as Anders writes, the Brenner potential is designed for larger systems.

[Anders Blom]

It's not really seen from the outside so clearly, but the Brenner potential comes from a different code (ASAP), and we don't have so much control over it. In the future we will reimplement the Brenner potential such that it can be used also for small systems. For now, I think the Tersoff or some other potential will give you just fine results for such a simple calculations.

[berlin]

Tersoff  does give results, but the Phonon Bandstructure from Tersoff  is not as good as Brenner, on the agreement to GGA/LDA plane wave DFT calculations or experiments for Graphene.

[Anders Blom]

The question is what you want to use it for. Since Brenner works fine - also in ATK - as soon as you make the cell larger, you can use that for any practical applications where you have a larger structure. If you are interested in 2 to 4 atom systems, you might as well use DFT

[berlin]

That is fine, thank you all very much!