Paper TF-FrM5
the Curious Wetting Behavior of ALD Grown Al₂O₃ Thin Film Surfaces

Friday, November 3, 2017, 9:40 am, Room 20

The curious wetting behavior of ALD grown Al₂O₃ thin film surfaces

Yi Li, Brandon D. Piercy, Mark D. Losego

Conventional wisdom suggests that Al₂O₃ surfaces should be hydrophilic. Recently, it was demonstrated by the Parsons lab at NC State that under certain conditions a few cycles of trimethylaluminum-water atomic layer deposition (ALD) transforms cotton fiber surfaces from hydrophilic to hydrophobic. These researchers suggest that adventitious carbon from ambient atmosphere is the source of this hydrophobicity, but they make no note of the universality of this phenomenon. In our effort to better understand this phenomenon, we discovered that ALD grown Al₂O₃ thin films of “any” thickness (1 nm to >50 nm) on “any” substrate (including silicon) can be transformed into a relatively hydrophobic surface state (water contact angles > 80°) when heat treated under appropriate conditions. After heat treatment, ellipsometry detects several angstroms increase in thickness and XPS suggests this layer is some form of hydrocarbon. In this talk, we will discuss our evaluation of three hypotheses for the source of this hydrocarbon layer: (1) contamination from the heating environment (e.g., oven), (2) carbon contamination from the ALD deposition process (i.e., unpyrolized ligands), and (3) adventitious carbon from the atmosphere. To address hypotheses (1) and (3), Al₂O₃ thin films grown by ALD to 10 nm thickness on silicon are found to become hydrophobic if we remove them from the ALD reactor and then re-anneal them in either a “dirty” ambient air oven or the “clean” ALD reactor under low-pressure and ultra-pure flowing nitrogen conditions. To further verify the purity of the ALD reactor, we have done this annealing step without breaking vacuum in the ALD reactor, leading to the Al₂O₃ surface remaining hydrophilic. Films annealed to 700 °C for 4 hours to fully pyrolyzed any remaining organics (still hydrophilic) and then re-heated to 150 °C for 30 min are also found to turn hydrophobic, suggesting that residual carbon within the film is also not the source. Thus, we believe that the Al₂O₃ surface is attracting adventitious carbon from ambient atmosphere and then reacting with the surface to form a hydrophobic layer. This reaction, which is thermally accelerated, appears to be key to the transition. We will report on how both temperature and time affect this surface reaction and our best understanding of the surface chemistry based on various surface spectroscopy techniques.