reaxFF for graphene

LAMMPS LAMMPS Mailing List Mirror
1

Dear all;
I tried to simulate adsorption of a HCN molecule on a graphene sheet and I used reaxFF potential for whole system. In the relaxation stage, the graphene structure tear up and is unstable, even though I place a graphene sheet alone, the result was same.
here is my questions:

  1. is my ffield.reax file inappropriate or reaxFF can’t make the graphene stable?
  2. how can i get the proper file? (I used the ffield.reax file in bench folder of lammps)
  3. as long as i don’t want to use AIREBO potential because it doesn’t support N atoms, is there any suitable potential field for my system?
  4. if I want to add Cl atoms to my system, does reaxFF still support?

ffield.reax.budzienSi (21.6 KB)

2

data.graphene (3 KB)

data.hcn (447 Bytes)

3

I think you underestimate the difficulty of constructing a potential to represent particular chemical systems. You also failed to read the documentation for the pair_style reax/c in its entirety. The reason why we provide multiple ffield.reax parameter files is that each one has been developed for a particular chemical system. It was not developed to represent every possible chemical system that can occur with the elements listed in the file. If you grab the first parameter file that you see, your chances of it working for your particular system are not very high. A better approach is:

  1. Search the literature for papers that use ReaxFF to model chemical systems similar to the one you are interested in.
  2. Obtain the parameter files for promising parameterizations (possibly in the LAMMPS examples/reax directory)
  3. Figure out if you need to change the default settings of parameters specified in the ReaxFF control file
  4. Verify that you can accurately reproduce results published in the relevant paper

Only after you have successfully completed all these steps should you attempt to run your application.

Aidan