Tersoff for carbon nitride with water

LAMMPS
1

I am trying to simulate adsorption of water on graphitic carbon nitride.

For the carbon nitride without water I draw the graphitic structure and I use it as initial geometry with tersoff potential and it is stable.

For water without carbon nitride I use SPC potential and it is stable.

To combine both, I wish to use Lennard-Jones potential between carbon nitride and water. I plan to optimize the LJ-parameters but to start I simply use the oxygen parameters for all cross interactions.

The issue is that the crystal is not stable when I add the hybrid interaction. I tried several ways to initializing the system, i.e first minimizing energy and then adding temperature or similar.

I appreciate hints with regards to conceptual or technical mistakes.

Below are my settings for the hybrid potential

Thanks

setting

atom_style full
dimension 3
boundary p p p
units real

pair_style hybrid lj/charmm/coul/long 9.0 11.0 tersoff
bond_style harmonic
angle_style harmonic
pair_modify mix geometric
special_bonds lj/coul 0.0 0.0 0.5
kspace_style pppm 0.0001

water

bond_coeff 1 450.00 1.000
angle_coeff 1 55.00 109.47
pair_coeff 2 2 lj/charmm/coul/long 0.0 0.0
pair_coeff 4 4 lj/charmm/coul/long 0.1553 3.166

carbon nitride

pair_coeff * * tersoff cn.tersoff C NULL N NULL

cross-terms

pair_coeff 1 2 lj/charmm/coul/long 0.1553 3.166
pair_coeff 1 4 lj/charmm/coul/long 0.1553 3.166
pair_coeff 3 2 lj/charmm/coul/long 0.1553 3.166
pair_coeff 3 4 lj/charmm/coul/long 0.1553 3.166

2

Unless you are an expert in force field parameterization, you should not do what you have been doing. I would say that the outcome speaks for itself … loudly.

Instead, you should search the published literature for simulation studies for the same or similar systems and find out what those studies have been using and choose the parameters that look most suitable for your needs.

3

Thanks for the hints. Yes previously I looked at literature on similar systems and I am considering their results. Its also part of my project to optimize these parameters.

The instability in my simulation was coming from the hydrogen atoms, as I assigned an LJ-potential between hydrogen and the carbon framework while SPC-water hydrogen has only bonded forces. I still have to clarify this.