Spin resolved photoelectron spectroscopy of Fe(3)O(4): the case against half-metallicity

Many materials have been theoretically predicted to be half-metallic, and hence suitable for use as pure spin sources in spintronic devices. Yet to date, remarkably few of these predictions have been experimentally verified. We have used spin polarized photoelectron spectroscopy to study one candidate half-metallic system, Fe(3)O(4). Such experiments are normally hampered by difficulties in producing clean stoichiometric surfaces with a polarization that is truly representative of that of the bulk. However, by utilizing higher photon energies than have traditionally been used for such experiments, we can study polarization in 'as received' samples, essentially 'looking through' the disrupted surface. High quality, strain relieved, ex situ prepared Fe(3)O(4) films have been thoroughly characterized by diffraction, transport and magnetometry studies of their crystallographic, electronic and magnetic properties. The spectroscopic results are found to agree fairly closely with previously published experimental data on in situ grown thin films and cleaved single crystals. However, despite the higher photoelectron kinetic energies of the experiment, it has not been possible to observe 100% polarization at the Fermi level. Hence, our data do not support the claim of true half-metallicity for Fe(3)O(4).