2 different ffield files with ReaxFF potential

LAMMPS LAMMPS Beginners
1

Dear LAMMPS users and developers,

I have a beginner question concerning the usage of ReaxFF potential.

I want to apply a ReaxFF potential to the molecule C6H11 BF4N2 (also known as: EMI-BF4).

Since I couldn’t find a single parameter file that can describe all the atoms of the molecule, I opted to use 2 separate files; one that describes the atoms C, H, N and F, and another one that describes the atom B.

I am using the following chunk of code:

pair_style hybrid/overlay reax/c NULL
pair_coeff * * ffield.reax.FC N C H NULL F
pair_coeff * * ffield.reax.AB NULL NULL NULL B NULL

But I am obtaining this error: “Pair coeff for hybrid has invalid style.”

My questions are:
1- is using 2 different files to describe one molecule correct ?
2- if yes, what is the correct way to do it ?

PS: I submitted a request to prof. Adri van Duin’s group in order to see if they have the force field file that I need, and in the meantime I am asking you these questions.

thank you so much,
Rafid.

2

No. This is a very bad idea.

3

Thank you for your prompt response.
Can I conclude from your answer that the only “good” way to define the force field is to find and use the adequate file that describes the whole molecule ?

4

Not just the whole molecule, but also for the conditions you are simulating in.
For empirical force fields (and ReaxFF qualifies as one) you cannot simply identify atom types with elements, but it also needs to consider the environment (e.g. temperature, density).
While ReaxFF is designed to represent all elements in a given simulation, the force field are much less transferable, i.e. the settings need to be compatible with the type of compound and the conditions.

The “rules” of what is adequate are different for different pair styles and simulation problems. Hence there are features like the hybrid pair style in LAMMPS, but they have to be used for the right reason and not the wrong reason. Outside of the incorrect syntax, you choice of hybrid is incorrect because it would simulate two separate systems (one for each force field file) without defining any interactions between them, which is fatally wrong for ReaxFF.

5

Thank you so much. I really appreciate your detailed answer.