[gmx-users] Graphene Sheet Rolling up into CNT

[Mark Abraham]

Justin A. Lemkul wrote:
> 
> 
> Darrell Koskinen wrote:
>> Hey GROMACS Gurus,
>> I was wondering if you could enlighten me. I performed a simulation of 
>> a graphene sheet in an ammonia gas at 240K using ffoplsaa parameters 
>> and see that the graphene sheet begins to curl up into what appears to 
>> be a nanotube. I did not run the simulation long enough to actually 
>> see it form a nanotube, but it appears to be headed in that direction. 
>> I am not sure if a graphene sheet would actually curl up on itself to 
>> form a CNT at this temperature. I believe CNTs are formed by ablation 
>> of pure graphite and therefore rolling up a graphene sheets must occur 
>> for the CNTs to form, but assumed that the local temperature increase 
>> resulting from the ablation was the cause of the rolling up. So I am 
>> wondering if you might be able to give me some idea of why this is 
>> happening? Could it be that the OPLS LJ parameters, since OPLS was 
>> created for liquids and not gases, are creating forces too great for a 
>> gaseous simulation thus causing too great an attraction between 
>> distant carbon atoms within the graphene lattice, which manifests 
>> itself in the rolling up of the graphene into a CNT? Or could the 
>> curling up be occurring because the graphene sheet is free floating in 
>> space as would happen during ablation. Maybe rolling up of graphene 
>> sheets does occur at such a low temperature. I will do some research 
>> on this possibility, but in the mean time please provide me your 
>> comments, if you can.
>>
> 
> My guess would be that the curling is a result of an inadequate model.  
> Several things come to mind, although probably a few of them have been 
> discussed at some point in the past.
> 
> 1. Appropriate use of improper dihedrals.  You shouldn't get crazy 
> bending or curling if the impropers are doing their job.
> 
> 2. Are you using a finite graphene sheet?  I think a better model would 
> be an infinite lattice (a periodic molecule).  You might get some 
> flexing, but you definitely won't get curling, if treated properly.
> 
> 3. The fact that you are simulating in the gas phase.  If the ammonia is 
> not close to the graphene sheet, then the graphene may only "see" itself 
> (depending on how your cut-off's are set up) and therefore be spuriously 
> attracted to itself.

Agreed, and perhaps the sheet is too thin to behave like bulk graphite. 
Certainly a single layer will not have (all) the properties of graphite.

Mark