Mechanical Properties of Evaporated Gold Films. Hard Substrate Effect Correction

Nanoindentation tests using the Berkovich indenter tip were performed on 50 and 200 nm thick polycrystalline gold films deposited on hard substrates. Gold film hardness increased with the indentation depth due to the influence of the substrate. A procedure based on the Joslin-Oliver method was introduced to correct for the substrate effect. The method utilizes the fact that the measured elastic modulus did not change significantly with the indentation depth for the two films tested, which allowed reducing the contact area variation used in the hardness calculation. Introduction The ability to measure mechanical properties of thin films is important for various applications. Typically thin films have higher hardness compared with their bulk counterparts. There are many methods used for measuring thin films mechanical properties, which utilize different sample geometries, including freestanding films [1], microbridge [2], and microbeam cantilever beam [3] to name a few. However, nanoindentation is the most commonly used method for measuring mechanical properties. When indenting bulk polycrystalline samples their hardness and modulus do not change significantly with indentation depth, however, for thin films, the elastic modulus typically changes a lot at very shallow indentations (<10% of the film thickness) and tends to flatten at deeper indentations. Hardness exhibits large scatter at very shallow depths, and then changes steadily with the indentation depth. For the very shallow indentations, the hardness variation can be explained by the indentation size effect or strain gradient plasticity theory [4]. Hardness change at deeper indentations can be explained by the substrate effect [5], although hardness, a measure of the materials ductile properties, should not change significantly with indentation depth. Some researchers argue that the increase in hardness may be caused by the indentations performed on the grain boundaries rather than the grains [6]. However, the same may also happen in bulk materials, where experimental data normally does not show such changes. Generally speaking, the hardness and elastic modulus of thin films are affected by the substrate. In order to correct the hardness, lots of models have been introduced. One of the most commonly used is the Oliver and Pharr method [7]. Oliver and Pharr used the elastic unloading slope of the load-displacement curve along with the calibrated tip area function to assess the elastic and plastic mechanical properties: Mater. Res. Soc. Symp. Proc. Vol. 1086 © 2008 Materials Research Society 1086-U08-41