Electronic structure of metal oxides is of great importance due to their huge use in heterogeneous catalysis, photochemistry, sensors, composite materials, etc. Among all metal oxides, TiO2 is considered a prototype for surface science studies and therefore it has been largely studied. The rutile (110) face is the most stable one and it presents two different surface reconstructions depending on the reduction level of the substrate, the annealing temperature reached during the sample preparation, and the annealing time. This reconstruction consists of Ti2O3 rows running along the [001] surface direction, as depicted by STM images and LEED I-V studies from our group[1,2].
We have measured the electronic structure of rutile TiO2 (110)-1x2 by means of angle resolved ultraviolet photoemission spectroscopy (ARUPS) for different photon energies. A previous theoretical work of our group pointed out the possibility of a 1D metallic character along the chains of the 1x2 reconstruction [1]. Using ARUPS and synchrotron radiation photoemission at different photon energies, we have looked for any electronic feature close to the Fermi edge related to the reconstruction. Although we have not found by photoemission any evidence of this state, we have been able to observe a characteristic dispersing peak of the reconstruction at around 7.5 eV below the Fermi level. In addition, evidence is given for a double component of the so-called defects peak at 1 eV of binding energy. This splitting is attributed to the contribution of two different kinds of Ti 3+ defects, related to different chemical environments: surface and bulk. This splitting is of about 0.4 eV. Besides, it is shown that the concentration of the oxygen vacancies (proportional to the Ti3+ peak) follows a decreasing distribution towards the bulk, which indicates a larger Ti/O ratio in the surface than in the bulk. We attribute this effect to the oxygen diffusion from the bulk to the surface during the reduction of the sample.
|