Can you read your post properly? I can’t. That indicates you have not read or are not following the guidelines and suggestions post for the LAMMPS categories on MatSci.org: Please Read This First: Guidelines and Suggestions for posting LAMMPS questions
It has “Please Read This First” in the subject for a good reason.
Besides, you model is bogus in multiple ways:
- ReaxFF uses a truncated, compensated Coulomb in Wolf summation. That is incompatible with long-range coulomb using Ewald or PPPM. It will lead to double counting of the long-range contributions from the ReaxFF atoms.
- When combining variable charge and constant charge models, you have a problem from the constant charge atoms not being polarized, but polarizing the variable charge atoms, which leads to unphysical leeching of energy
- You cannot compute the charge equilibration correctly when you have no parameters for the constant charge atoms. In fact, I don’t even see a charge equilibration fix.
Your error is a trivial error due to not properly using the syntax for pair_coeff commands when using a hybrid pair style. But this is not worth fixing because of the bad model overall.