Contamination of TiO2 thin films spin coated on rutile and fused silica substrates

The present work reports data for TiO2 thin films spin coated on unpolished fused silica (FS) and unpolished (001) rutile annealed at 200°, 350°, 450°, and 550 °C for 8 h in air. Characterization/testing consisted of glancing-angle X-ray diffraction, laser Raman microspectroscopy, atomic force microscopy, UV–Vis spectrophotometry, X-ray photoelectron spectroscopy, secondary ion mass spectrometry, transmission electron microscopy, and methylene blue (MB) photodegradation. The recrystallization depended on the type of substrate, where FS substrates yielded Si-contaminated anatase films of limited crystallinity (amorphous at 200 °C) and thickness ~160–200 nm while the rutile substrates resulted in well crystallized epitaxial films of ~200–300 nm thickness. The Si grain boundary diffusion coefficients were 4.59 × 10−19 m2s (200 °C) and 5.00 × 10−19 m2s (550 °C). This diffusion mechanism was supported by the XPS data before and after MB aqueous solution testing, which suggested that Si contamination was on the grain boundaries and was leached during testing. The thickness of all films increased slightly as the annealing temperature increased, which resulted from grain rearrangement and growth. The optical transmissions of all films were low and the reflectances higher, which resulted from the roughness and unevenness of the films. The optical indirect band gaps (Eg) from transmission and reflectance spectrophotometries decreased for the anatase films but increased for the rutile films with increasing annealing temperature. These trends reflect the dominance of competing mechanisms, where recrystallization and Eg decrease are dominant for anatase but void formation and Eg increase are dominant for rutile. The photocatalytic performances of the anatase films were superior to those of the rutile films. The performances corresponded to the annealing temperatures and so were linked directly to the crystallinity and, to a lesser extent, the surface area (reflected by the roughness and unevenness) since the other factors examined were inconsistent with the performance data.