You can certainly make a slab of Fe and put a molecule over it and see what happens. This sort of calculation is done often; I haven’t seen it done with free radicals, but there’s no reason you can’t try. If you have a charged molecule, like you mentioned in another question that you asked, then you might have to think about what sized supercell might be needed to eliminate spurious electrostatic interactions between periodic images, but in principle there shouldn’t be much of a problem with setting up a system with a charged molecule near a slab.
One caveat for all of this, is that while you can set up the calculation and relax the structure, there’s no guarantee that the results will be accurate. Your choice of things like supercell size and exchange-correlation functional might end up being important, and it would be good to find papers on other slab calculations people have done to see what sorts of choices are typical, and how accurate you might hope your calculation can be.