Glancing angle deposition (GLAD) has found application in a wide range of fields requiring porous, high surface area thin films, including sensors, optics, and energy devices. [1] This diversity is due in large part to the wide range of compatible materials, including metals, semiconductors, and organic compounds. Metal oxides are of particular interest for energy storage and conversion applications since they can be tuned for a combination of transparency, electrical conductivity, and chemical and thermal stability. Achieving the desired stoichiometric phase is essential for controlling desirable metal oxide properties. Here we discuss the challenges associated with achieving phase control in porous GLAD films.

Metal oxide GLAD films typically deposit in an amorphous state, so post-deposition processing is one route used to access a particular crystallinity and stoichiometry. Thermal annealing conditions depend on the desired phase: anatase TiO₂ readily forms at a few hundred degrees Celsius in air; Ti₄O₇ generally requires longer anneals at high temperatures (1000 ºC) in a reducing atmosphere (eg. H₂ in carrier gas). [2] Annealing temperature, duration, and environment (e.g. oxidizing vs. reducing atmosphere) can all have an impact on film morphology, since coalescence or softening of structures is greatly enhanced at temperatures near the melting point of the metal oxide. Both the porosity of the film and the strength of the reducing atmosphere affect the extent of oxygen removal and morphology changes at relatively high temperatures, while still allowing access to a wide range of compositions (e.g Ti_nO_2n-1, n = 2 - 9), phases (e.g. monoclinic, tetragonal, or orthogonal Nb₂O₅) and the associated optical, electronic, and thermal properties.

We will present methods to retain the porosity and structure of GLAD thin films while achieving desired stoichiometry and phase via post-deposition annealing, with a specific focus on phase and crystallinity characterization using x-ray diffraction. We will attempt to correlate results from the Ti-O and Nb-O systems with results from other systems of interest (e.g. W-WO₃) [3] as part of a better understanding of phase formation in porous thin films.

[1] M.M. Hawkeye and M.J. Brett, J. Vac. Sci. & Tech. A, 25, 1317 (2007).

[2] J.R. Smith et al, J. Appl. Electrochem., 28, 1021 (1998).

[3] D. Deniz et al., Thin Solid Films, 518, 4095 (2010).

We thank NSERC, iCORE, Micralyne, and the National Research Council - Technology Development Program for financial support.