Comparison of Size-Dependent Structural and Electronic Properties of Anatase and Rutile Nanoparticles
journal contributionposted on 23.04.2009 by Vittorio Luca
Any type of content formally published in an academic journal, usually following a peer-review process.
Size-dependent variations in the electronic and structural properties of anatase and rutile nanoparticles have been compared. The anatase nanoparticles of the present study were prepared by hydrothermal ripening of an anatase sol and had diameters in the range 2−130 nm whereas the rutile nanoparticles were prepared by calcination of sol−gel derived rutile and had diameters in the range 3.6−60 nm. The hydrothermally ripened anatase nanoparticles had similar surface structures as deduced from the XANES as previously reported sol−gel anatase materials prepared through calcination (Luca et al., J. Phys. Chem. B 1998, 102, 10650). The optical band gap (Eg) of the anatase nanoparticles as deduced from their electronic absorption spectra showed some variation with size but Eg was not a smooth function of crystallite size, as would be dictated by the effective mass model for both types of anatase nanoparticles. In distinct contrast to the anatase nanoparticles, rutile nanoparticles showed a smooth size dependent variation in optical band gap in line with the dictates of the effective mass model. However, the XANES of the rutile nanoparticles was not dependent on size as it was for both the calcined and hydrothermally ripened anatase materials where the pre-edge XANES and EXAFS revealed a high concentration of distorted surface atoms with reduced coordination. The results suggest that sol−gel anatase nanoparticles consist of a core−shell structure in which the core is bulk-like and the shell interphase is less ordered with a high degree of Ti under-saturation. On the other hand, if such an interphase region was present at all in rutile nanoparticles, it was so thin as to avoid detection by XANES. The unique surface structure of anatase nanoparticles derived from sol−gel preparation methods is probably responsible for the lack of a clear quantum confinement effect.