I would like to use "fix deposit" with ReaxFF to add atoms to a simulation in which I deposit molecules on a surface.

I have built working scripts for my system with other force fields. However, there is an issue with the ReaxFF implementation in LAMMPS where depositing new molecules into a system will cause an error (too many ghost atoms). As per the link, the error most likely arises because the USER-REAXC package pre-allocates memory for atoms based on the initial number of atoms in the simulation box.

I do know the maximum number of atoms needed before the simulation starts. I am looking for a workaround for this issue, which would allow me to pre-allocate memory for this number of atoms, which I can later add to the simulation box.