Are the surfaces of CrO$_{2}$ half-metallic?

Previous photoelectron spectroscopy studies of CrO$_{2}$ have found either no density of states or a very low density of states at the Fermi level, suggesting that CrO$_{2}$ is a semiconductor or a semimetal. This is in contradiction to calculations that predict that CrO$_{2}$ should be a half-metallic ferromagnet. Recently, techniques have been developed to grow high-quality epitaxial films of CrO$_{2}$ on TiO$_{2}$ substrates by chemical vapor deposition. We present photoelectron spectroscopy measurements of epitaxial CrO$_{2}$(110)/TiO$_{2}$(110) and CrO$_{2}$(100)/TiO$_{2}$(100) grown using a CrO$_{3}$ precursor. In addition, measurements of epitaxial Cr$_{2}$O$_{3}$(0001)/Pt(111) films grown by thermal evaporation of Cr in an oxygen atmosphere are presented as a reference for reduced CrO$_{2}$ films. The measurements of the CrO$_{2}$ surfaces show no emission at the Fermi level after sputtering and annealing the surfaces in oxygen, even though our soft core photoemission data and low energy electron diffraction measurements provide evidence that stoichiometric CrO$_{2}$ is present. The consequence of this is that neither surface of CrO$_{2}$ is metallic. This behaviour could result from a metal to semiconductor transition at the (110) and (100) surfaces.