Hello lammps users,
How do you define O2 molecule in a reaxff simulation of O2 molecule deposition on silicon surface.
The simulation uses atom style = full and reads the O2 molecule with molecule command. some articles suggest it may be done in 2 ways. The first is defining bond_style and bond coeff in main script, like example in.deposit.molecule does. But I found O2 bonds do not break even under tens of thousands Kelvin temperature. This made the first way very suspicious. The second way is adding charges flag in molecule file. Some one used this way to define H2O molecules. I am not sure if this works and what charge should be used for O2.
Does anyone has some ideas about this?
Thanks,
S Wang
Hello lammps users,
How do you define O2 molecule in a reaxff simulation of O2 molecule deposition on silicon surface.
The simulation uses atom style = full and reads the O2 molecule with molecule command. some articles suggest it may be done in 2 ways. The first is defining bond_style and bond coeff in main script, like example in.deposit.molecule does.
for reaxff calculations you must not define bonds, angles, dihedrals, or impropers. those are all defined implicitly by the pair style. the typical atom_style for reaxff is thus “charge”, but “full” may also be used (it is a superset) for as long as no bond topology is present.
But I found O2 bonds do not break even under tens of thousands Kelvin temperature. This made the first way very suspicious.
this is just what you have told LAMMPS to do. (harmonic) bonds don’t break. they are thus only used in conventional non-reactive force fields, where this behavior is desired.
The second way is adding charges flag in molecule file. Some one used this way to define H2O molecules. I am not sure if this works and what charge should be used for O2.
you would have to set charges in both cases! you can use a charge of 0.0. the actual charge will be determined dynamically during the simulation due to charge equilibration.
Does anyone has some ideas about this?
i think before you continue running simulations, you should be reading up on how ReaxFF works as a force field and what is part of the calculation. then all your questions will become self-evident and you won’t make mistakes or have to ask here.
axel.
1 Like