https://doi.org/10.3390/nano12203692
“The impact of a titania (TiO2) support film surface on the catalytic activity of gold nanoparticles (Au NP) was investigated. Using the reactive dc-magnetron sputtering technique, TiO2 films with an amorphous, anatase, and nitrogen-doped anatase crystal structure were produced for a subsequent role as a support material for Au NP. Raman spectra of these TiO2 films revealed that both vacuum and NH3 annealing treatments promoted amorphous to anatase phase transformation through the presence of a peak in the 513–519 cm−1 spectral regime. Furthermore, annealing under NH3 flux had an associated blue shift and broadening of the Raman active mode at 1430 cm−1, characteristic of an increase in the oxygen vacancies (VO). For a 3 to 15 s sputter deposition time, the Au NP over TiO2 support films were in the 6.7–17.1 nm size range. From X-ray photoelectron spectroscope (XPS) analysis, the absence of any shift in the Au 4f core level peak implied that there was no change in the electronic properties of Au NP. On the other hand, spontaneous hydroxyl (–OH) group adsorption to anatase TiO2 support was instantly detected, the magnitude of which was found to be enhanced upon increasing the Au NP loading. Nitrogen-doped anatase TiO2 supporting Au NP with ~21.8 nm exhibited a greater extent of molecular oxygen adsorption. The adsorption of both –OH and O2 species is believed to take place at the perimeter sites of the Au NP interfacing with the TiO2 film. XPS analyses and discussions about the tentative roles of O2 and –OH adsorbent species toward Au/TiO2 systems corroborate very well with interpretations of density functional theory simulations.”
“For surface chemical studies, X-ray photoelectron spectroscopy (XPS) measurements were carried out on a K-Alpha instrument (Thermo Scientific, East Grinstead, UK) using a monochromatic X-ray beam (Al Kα) with a spot size of 300 × 300 μm2. The spectrometer is equipped with a flood gun for charge compensation, any charge-induced energy shift being corrected by fixing the C 1s line at 284.4 eV. For the peak-fitting procedure, a Shirley-type background was subtracted from the spectra while the peaks were fitted with symmetric Gaussian functions.”
”
“