No OH radical formed while using lammps

LAMMPS LAMMPS General Discussion
1

Hello List,

We are doing a combustion study of fuel (Cetane) using molecular dynamics (MD) simulations with LAMMPS using the ReaxFF module

Our model system consists of 1 Cetane and 39 Oxygen molecules in a 25Å Box and followed the NVT MD (at 1000 K) to simulate the combustion process and observed some dissociation of fuel molecules ie. cetane was broken and formed CO2, H2O, CO etc.

However, we did not observe the presence of any -OH radicals in our simulations, which we found in our equivalent NVT MD calculation with DFTB+.

Our Question is:
Is the ReaxFF in LAMMPS capable of predicting the formation of OH radicals ?
We guess that Formation of OH radical is expected from a first-principle calculation (though DFTB+ is not strictly an ab initio).

But it should be noted that formation of OH is actually occured in a fuel burning process (Please see: https://discovery.ucl.ac.uk/id/eprint/10084077/1/PhD%20Thesis_Muye%20Feng_201910.pdf ,ref. eg: Figure 4.5 Time evolution of OH species during ReaxFF NVT MD simulations of ethanol oxidation)

We still don’t know whether this discrepancy is due to an error in the input file or if there are some settings that we are not aware of. The input files have been linked below along with the dump files containing the initial and final iterations.

LINK OF INPUT: LAMMPS - Google Drive
Any help in this concern is much appreciated,
And Thanks in advance.

2

There are no objective restrictions for creation of OH radicals in ReaxFF simulations. If you look in the original paper describing the parametrization you used, you’ll see that they did detect some OH radicals (albeit for a different hydrocarbon).
You didn’t attach the log, so I’m not certain, but the timestep of 0.5 fs may be too large for atomic hydrogen in 1000K. If there are Dangerous builds in the end of the log, you should also change the frequency of neighbour list rebuilding (the most safe settings are neigh_modify every 1 delay 0 check yes).

3

Thank you so much for your response.
Link for Log file : mylog - Google Drive
I will do the recommended changes and change the timestep to 0.1 and redo the simulation. I have checked the log file and at the end it says Dangerous builds not checked.