First of all, it is recommended to use the newer CHARMM36 potential instead.
Then, let me explain why your approach won't work. the polymerization features in CHARMM's RTF format are mostly convenient for 1D polymers; crosslinking and branching are usually implemented in the form of patches. For a 2D graphene sheet of relevant size, you'll need to apply so many patches that it will negate the advantage of the "polymer" approach altogether. In other words, I think you'll be better off generating the graphene sheet as one monolithic residue. The easiest way to do so starting from a mol2 file is the attached script. Please do heed the limitations listed at the beginning of the file.
Note that the script
- uses existing atom types (specifically CG2R61 from CGenFF, though it would work just as well with CA from the CHARMM protein force field). We recommend against introducing new atom types unless (1) you really know what you're doing and (2) have a really good reason. (1) is obviously violated in your case since you attempted to define different atom types for chemically equivalent atoms. (There are a few exceptional cases where that's worth doing, but graphene isn't one of them, and even if it were, 6 atom types would still be far too many.)
- will not generate IMPR statements. The reason for this is that 6-membered aromatic rings in CHARMM do not benefit from improper dihedrals; the out-of-plane behavior is adequately captured by the angle and (proper) dihedral potential alone and impropers would only over-rigidify them. See also this FAQ entry.
- defines all the bonds by BOND entries; as Lennart mentioned, "AROMatic" doesn't have any special meaning to CHARMM.
Finally, assuming you're using PBC, IF
you want an infinite graphene sheet that spans the periodic boundary, then things get a bit more complicated and you'll still need patches. I'm not experienced enough to be able to tell for sure how to best accomplish this; the other forum regulars may know more.