I don't know (yet) if he was the first to think of a pi-bond between a metal and ligand, but I have just read this:
"Atoms of transition groups are not restricted to the formation of single bonds, but can form multiple bonds with electron-accepting groups by making use of the electrons and orbitals within the valence shell." Pauling, 1939.
It is probably from The Nature of Chemical Bond. I will try to check my book to find the statement.