You should choose an approach based on what you are trying to accomplish in the simulation. If you want to study changes in bonding, you will need to use ReaxFF in LAMMPS or else an electronic structure code. If not, you can use simpler potentials and builders like MolTemplate will generate all the necessary LAMMPS input.
there are plenty of examples in examples/reax and there is
documentation in the manual. the remaining information is in the
relevant publications, which are a must to study in order to do
meaningful simulations. as an advanced and complex force field, it
requires practice and care to use well. in other words, you'll first
have to practice reproducing already published data before moving on
to a new compound.
i can't find any reaxff forcefild file in which I can use lj parameters
directly.
what on earth is that supposed to mean? a suitable reaxff force field
file would provide all required parameters for all types of atoms
involved.