Influence of water on the photocatalytic properties of synthetic nanocrystalline iron oxide films

Two sets of synthetic nanocrystalline iron oxide films were grown via Atmospheric Pressure Chemical Vapor Deposition (AP CVD) under identical conditions, barring the exclusion or inclusion of water vapor within the depositional environment. The morphology and electrochemical properties of the resulting films were examined and compared.;The presence of water during AP CVD deposition of nanocrystalline silicon doped hematite films impacts the photocatalytic properties of the resulting films. This is likely due to the inhibition of optimal doping of Fe2 O3 by the presence of water during growth. The presence of an amorphous iron phase in the wet preparation films, likely am-FeOOH, was suggested by Raman analysis, and is a possible source of lowered photocurrents in these films.;Wet preparation films were found to have a bandgap of 1.96 eV and an estimated film thickness of 214 nm; dry preparation films have a bandgap of 1.9 eV and an estimated film thickness of 412 nm thick. Irrespective of preparation method, the fundamental particle size had an average diameter of about 20 nm. Mott-Schottky analysis was used for donor density calculations; dry preparation films had a donor density of about 8.9*1018 cm-3 and wet films had a donor density of about 5.5*1016 cm -3. The differences in donor levels result in a significance difference in depletion layer width, which in turn may be responsible for lowered photocurrents in wet preparation films relative to dry preparation films.;Cyclic voltammetry indicates that at low applied potentials, loss of efficiency is dominated by intragap state recombination in both preparation methods, but at higher applied potentials, loss of efficiency in dry preparation films is dominated by reduction of radicals in solution by photogenerated electrons, and by intragap state recombination in wet preparation films. Evidence suggests that the intragap state is never fully emptied in the wet preparation films and thus optimum photocurrents are never achieved. The single slope seen in the Mott-Schottky analysis supports this assertion.