Energy Band Alignment between Anatase and Rutile TiO₂

Abstract

Using photoelectron spectroscopy, the interface formation of anatase and rutile TiO 2 with RuO 2 and tin-doped indium oxide (ITO) is studied. It is consistently found that the valence band maximum of rutile is 0.7 ± 0.1 eV above that of anatase. The alignment is confirmed by electronic structure calculations, which further show that the alignment is related to the splitting of the energy bands formed by the O 2p z lone-pair orbitals. The alignment can explain the different electron concentrations in doped anatase and rutile and the enhanced photocatalytic activity of mixed phase particles.SECTION: Surfaces, Interfaces, Porous Materials, and Catalysis A fter Fujishima and Honda 1 had reported on the photocatalytic activity of TiO 2 , the influence of crystal structure on this property was investigated intensively.2,3 Over the past 2 decades, it was commonly observed that mixed anatase/rutile systems show more favorable photocatalytic properties than pristine ones of either modification. 4−9 The synergistic effect of the mixed systems has been attributed to a built-in driving force for separation of photogenerated charge carriers. Such a driving force may result from either a built-in electric field or from energy barriers blocking charge transfer at the interface between anatase and rutile. The latter are described by the energy band alignment, which is well-studied for semiconductor interfaces. Connelly et al.11 recently reviewed several models that are trying to explain the synergistic effect of mixed anatase/rutile systems. Well-known are the rutile sink model of Bickley et al. 4 and the rutile antenna model of Hurum et al., 5 which place the band edges of rutile (energy band gap E g = 3.0 eV 12 ) in between the band edges of anatase (E g = 3.2 eV 13 ). Kavan et al.14 performed electrochemical measurements that located the conduction band edge of anatase 0.2 eV above that of rutile, which corresponds to aligned valence band maxima. These models, however, were not able to convincingly account for the observed synergistic phenomena. Only recently, Deaḱ et al. 15 as well as Scanlon et al. 16 found theoretical and experimental indications for an energy band alignment with valence and conduction band energies in rutile both located higher in energy than in anatase when brought into contact. With such a staggered energy band alignment at the anatase/rutile interface, photogenerated electrons will preferentially move to anatase due to its lower conduction band minimum energy E CB , and holes will move to rutile due to its higher valence band maximum energy E VB . Deaḱ et al. 15 used the alignment of branch point energies 10 for their calculations. For oxides, though, it has been shown that due to a low density of induced interface states, the alignment of branch point energies does not necessarily yield proper results for the energy band alignment. 17In this work, further evidence for a staggered energy band alignment at the anatase/rutile interface is provided by X-ray photoelectron spectroscopy (XPS)...

Water splitting and the band edge positions of TiO2

“…The spectro/photo/ electrochemical measurements by Jankulovska et al [8] confirmed the positive offset of 0.2 eV between the CBMs of anatase and rutile, and other measurements, like EPR [10], photoluminescence [11], photocatalytic Ag + reduction [12,13] and H 2 evolution [14] on mixed-phase particles, also reported the CBM of anatase being higher than that of rutile. In contrast to these works, all photoelectron spectroscopy (PES) studies -carried out on layers in electrical contact and in a vacuum environment -seem to agree that the CBM of rutile is lying higher [15][16][17]. The popular material in photocatalysis, Degussa (Evonik) P25 contains anatase/rutile mixed-phase particles, and there is a reasonable consensus that the enhanced photocatalytic activity is driven by exciton dissociation across the anatase/rutile interface, for which the CBM offset provides the necessary driving force.…”

Electric-dipole effect of defects on the energy band alignment of rutile and anatase TiO2

“…Such significant energy spaces between corresponding band edges for two TiO2 phases also was observed by experiments. The polycrystalline anatase thin films and rutile single crystals prepared by Pfeifer et al shows that the VBM (CBM) of rutile is 0.7 (0.5) eV above that of anatase according to the photoelectron spectroscopy analysis 47 .…”

Elongated titania nanostructures as efficient photocatalysts for degradation of selected herbicides

“…2. It is known that photocatalytic efficiency of TiO 2 is substantially increased in the sample containing both crystal phases (anatase and rutile) compared to single crystal sample [29]. Further increase of calcination temperature to 800 • C induced decrease of photocatalytic efficiency which is still higher compared to the as prepared titania nanotubes and samples calcined at temperatures lower than 700 • C. Such decrease in photocatalytic efficiency could be explained by the higher degree of particle agglomeration.…”

“…Considering degradation rates of selected herbicides, presented in Table 3, it could be concluded that there were no significant differences in photocatalytic activity of calcined sample. But, the overall increase in their photocatalytic activity in comparison to as prepared nanotubes is mainly the consequence of their generally higher crystallinity and existence of anatase and rutile phases (95%: 5%), which is a favorable condition (prerequisite) for the efficient photocatalytic degradation process on the surface of TiO 2 nanoparticles [29]. Taking into account negative electrostatic potential of studied molecules and the pH value of TNT 700 dispersions (7.2-7.8), it is clear that deprotonated surface of nanocrystals does not contribute significantly to increased efficiency of degradation process.…”