Dear Charka
This topic comes up frequently. Questions like that are difficult
to answer without more information. I suspect you also want to know
how to set the force-field parameters as well.
Firstly, you need to decide: do you need explicit bonds (bonded
interactions) in your system? Generally speaking if there are
molecules in your system, then you do.
OR do you want to use many-body pair_styles? (such as tersoff, sw,
or AIREBO) which attempt to includes these effects
If you need explicit bonds in your system, you should give some
thought to the kind of force-field you intend to use. (3-body)
angle-types, and (4-body) dihedral types can usually be guessed by
atom type easily, but the number of (4-body) improper interactions,
and their atom order, can be highly force-field dependent.
Both moltemplate and topotools can guess all of the bonds, angles,
dihedrals and impropers in your system, according to atom-type.
Moltemplate currently requires you to supply a list of bonds, while
topotools can infer this by the distance between atom-pairs. I you
use the OPLSAA or AMBER/GAFF, force-fields, then moltemplate can
automatically guess the improper interactions. (Otherwise improper
interactions can be painful to figure out, no matter which tool you
are using.)
---- graphene ----
In your case, it looks like you are simulating a large sheet of
graphene (10816 atoms).
(If you don't need explicit bonds then you can build this easily in
moltemplate. You could create a system.lt file containing "import
graphene.lt", and "unitcells = new Graphene [5408]", and run
moltemplate using "moltemplate.sh -xyz grap2.xyz system.lt". If there
are other atom types in your system, you might need to add a few
pair_coeff commands to assign Lennard Jones parameters between the
carbon atoms and those other atoms. For example, read the end of this
file)
http://www.moltemplate.org/examples/nanotube+water/system.lt
However, there are a lot of questions: Do you intend to keep the
graphene sheet rigid, or do you intend to allow the carbon atoms to
move during the simulations? If you intend to freeze the positions of
the carbon atoms, then you don't need explicit bonds, and it does not
matter what force-field you use to describe interactions between them.
Incidentally, toptools has a nice tool to create sheets of graphene
with-or-without bonds. (If you want bonds, I think you must add them
manually at the periodic boundaries.)
---- graphene force-field ----
If it helps, someone used LAMMPS (with the AIREBO pair style) to
simulate the stretching of carbon nanotubes. I don't know if that
force-field would work for graphene. If so, then you don't need (or
want) to create explicit bonds between carbon atoms. (The images
below show bonds, but I think it is only to make it easier to
understand the images. I don't think they are explicitly present in
the simulation.) For more information on that, click these links:
http://lammps.sandia.gov/images/cnt1.jpg
http://lammps.sandia.gov/images/cnt2.jpg
http://scitation.aip.org/content/aip/journal/jcp/134/20/10.1063/1.3594197?showFTTab=true&containerItemId=content/aip/journal/jcp
http://lammps.sandia.gov/pictures.html
Cheers
Andrew