New strategies for application of defect engineering to TiO2 based photocatalytic materials

There is good reason to believe the performance of semiconductor metal oxides for catalytic applications may be improved through the application of principles related to defect engineering. Specifically, TiO2 has already demonstrated promise as a catalytic material for environmental remediation by photocatalysis in aqueous media as well as water splitting for hydrogen fuel production by photo/electrocatalysis. There remain a number of major hurdles preventing widespread implementation of such technologies related to the material’s optoelectronic properties. Previous applications of defect engineering to these hurdles have yielded a variety of point dopants which narrow the band gap to include portions of the visible spectrum, but carrier lifetime is reduced as the same dopants act as recombination centers. Dopant free defect engineering has demonstrated that the amorphous synthesis route has a direct result on the native carrier concentration of a polycrystalline finished production through the manipulation of medium range order; this control has enabled an improved drift current to drive minority carriers to the free surface, but has fallen short of its goal. Heterostructuring has produced some innovative material combinations that place band edges of complimentary materials close to redox potentials of desired reactions, but there are many limitations including difficulties with material selection, interfacial charge transfer, cost, and stability. This thesis endeavors to introduce and advance three distinct efforts within the spectrum of applicable defect engineering techniques. In dopant free material, the relationship between annealing temperature and final material properties of polycrystalline anatase TiO2 was explored. The soak temperature had a strong effect on crystallite size with a variation of about a factor of 2 with the larger ones being of a length scale comprising the majority of film thickness. However, there was no corresponding changes in density or carrier concentration. This demonstrated that measurements of crystallite size may not be used as a direct estimate of grain sizes. As grain size (viewed by film density) did not alter, there was no benefit with regard to carrier concentration through the removal of free surface area within the material. Attempts at reducing fully oxidized material by annealing in a hydrogen atmosphere produced thermally grown Magnéli phases without affecting the bulk crystallinity. These phases appear to have grown along grain boundaries through the removal of oxygen in the form of water. Because these phases form a separate contact with the substrate and top contacts, their effect of film carrier concentration is not perceivable until the filaments are fully formed, after which any further changes to the bulk material are obscured. Reversibility by thermal methods did not have good success possibly due to limitations to the diffusivity of molecular oxygen. Lastly, a new heterostructure design was explored to better understand the dominant mechanisms of minority carrier transport. In this system, SrRuO3, a strongly correlated metallic oxide is used as a visible light absorber. Both electrons and holes must transverse a rectifying interfacial barrier before crossing a TiO2 layer to a catalytic free surface. Minority holes avoid the need to tunnel or be excited beyond the TiO2 band edges via an interfacial surface defect. From the interface, their transport to the free surface is governed largely by diffusion as evidenced by a linear dependence of activity on TiO2 thickness and a weak band bending at the interface indicating little to no built in electric field. The performance of the material in an amorphous state was comparable to an epitaxially grown counterpart suggesting a robust system amenable to further development in lower cost systems. Performance was a bit under an order of magnitude below polycrystalline thin film TiO2 catalysts