Modified Pull Test for the Testing of Very Adherent Films

摘要

The effectiveness of a thin film on a substrate depends largely on the degree of adhesion of the film to its substrate. Much effort has been put into developing reproducible, quantitative adhesion tests. The pull test is a direct measure of the tensile force required to remove the test film normal to the substrate surface. A technique was developed to introduce a reproducible flaw at the film/substrate interface in an attempt to lower the apparent adhesion of the film to within the range of the pull tester. This technique also promotes film failure at the artificial flaw and subsequent crack propagation along the film/substrates interface. Zirconium films approximately 500 nm were electron beam evaporated onto the basal plane of single crystal sapphire substrates. The pull test was performed using a Sebastian I adherence tester. The pins were 3 mm in diameter and coated with epoxy. The adhesion of these zirconium films to their sapphire substrates was greater than the strength of the epoxy used to bond the test pin to the specimen. A thin layer of gold was sputter deposited onto a small portion of the sapphire substrate after the substrate had been cleaned. The test film was then deposited through a mask such that a portion of this new film overlapped the gold layer. Since the gold layer was non-adherent (experimentally) to the sapphire substrates used in this experiment, this served as a release layer for the test films. The mild shadowing effect of the mask used during gold deposition caused the edge of the gold layer under the test film to be tapered and thus this was treated as a reproducible crack and the analysis made using the concept of fracture toughness as indicated by the stress intensity factor, K. Results are discussed. 12 refs., 3 figs., 3 tabs. (ERA citation 13:054301)