It won't be bonding to the H. Think about what the C6H5 ligand has to be.
EDIT:
And 400-1000 is usually called the "fingerprint" region. It's mostly due to carbon-carbon bonds and is really messy. IR is typically to prove you made the product/identifying unreacted reactant /solvent/side product. So you really want to see if there are new bonds formed in the product that aren't in the starting material. Forget the other stuff. But if you're doing something like FeCl2 + H2acac -> Fe(acac)2 + 2HCl an IR isn't going to prove s#*$, because the starting material and product IRs will be the same. That's pretty much the extent of my IR knowledge. Hope it helps.