The main goal of this thesis is to better understand the photochromic mechanism in rare-earth (RE) oxyhydride thin films. When illuminated, these materials have shown to darken, and bleach to a transparent state when the illumination source is removed. The RE oxyhydride thin films used in this thesis are deposited using magnetron sputtering and are YOxHy and GdOxHy films with different storage (air and vacuum) and oxidation (air and dry air) conditions. The films were probed using Doppler broadening positron annihilation spectroscopy, spectrophotometric (optical) measurements, X-ray diffraction and spectroscopic ellipsometry, measured before, during and after in or ex situ ultraviolet illumination.

Series of illumination and bleaching cycles have shown to permanently induce vacancy defects, which are most likely divacancies. The photochromic effect is seen as a reversible process, which leads us to believe that the changes seen in the defects are not directly responsible for the photochromic effect. During in-situ photodarkening and bleaching partially reversible changes are seen, whereby systematic shifts in the positron annihilation Doppler parameters are exhibited. This shift correlates with the photochromic effect, where the irreversible formation of vacancies is seen as a result of UV illumination. This leads us to the idea of the formation of metallic domains that may be attributed to H- ions that become mobile under illumination and may cluster to form metallic domains. Furthermore, the formation of the vacancy defects correlates with an increase in bleaching times seen from optical measurements, suggesting that they could play a role in the memory effect. Another optical property seen is a change in the band gap. We have seen that during photodarkening, the band gap tends to widen and decrease during bleaching. This phenomenon can be associated with the Burstein-Moss effect and can be used to estimate the photo-induced electron density in the conduction band. Despite the changes observed from positron annihilation techniques and optical measurements, no significant differences were observed by X-ray diffraction. The lattice constants of the film before and after illumination were shown to be constant.

There are two hypotheses explaining the photochromic mechanism: the segregation of metallic domains as a result of H- ion mobility and an Anderson-Mott insulator-to-metal transition as a result of overlapping electron wave functions. Using spectroscopic ellipsometry, these models were tested for YOxHy. It is shown that the mechanism leading to the darkening of the films is best described by the formation of Y or YHx metallic domains and voids. The percentage of metallic domains and voids grows with illumination time and starts stabilising after 90 minutes of illumination. The fraction of metallic domains (1.9-2.6% after 150 minutes of illumination) is similar in magnitude as seen in previous studies. Furthermore, the voids probed by ellipsometry are much larger than the vacancy-related defects probed by positrons. Such voids could give rise to the formation of positronium in the films, as could be examined in future positron annihilation lifetime spectroscopy (PALS) experiments.

Qualitatively GdOxHy and YOxHy behave similarly when it comes to optical measurements, XRD and DB-PAS data. For both systems, an increase in bleaching time as a function of illumination and bleaching cycles, a widening and shrinking of the band gap and a decrease in contrast are observed. However, there are quantitative differences between these films. The bleaching constant is shorter for GdOxHy and the contrast is higher, which makes it seem more attractive for potential applications. However, GdOxHy could not yet be satisfactorily modelled, in contrast to the case of YOxHy using spectroscopic ellipsometry. The GdOxHy film could not be described by a single homogeneous layer and therefore requires more complex models to describe the behaviour of the film in future work.