A new procedure for measuring the decohesion energy for thin ductile films on substrates
- PDF / 1,200,267 Bytes
- 8 Pages / 576 x 792 pts Page_size
- 99 Downloads / 253 Views
A novel testing technique has been developed capable of measuring the interfacial fracture resistance, F,, of thin ductile films on substrates. In this technique, the thin film on the substrate is stressed by depositing onto the film a second superlayer of material, having a large intrinsic stress, such as Cr. Subsequent processing defines a precrack at the interface between the film and the substrate. The strain energy available for driving the debond crack is modulated by varying the thickness of the Cr superlayer. Spontaneous decohesion occurs for superlayers exceeding a critical thickness. The latter is used to obtain F, from elasticity solutions for residually stressed thin films. The technique has been demonstrated for Cu thin films on silica substrates.
I. INTRODUCTION The decohesion of interfaces between metals and nonmetals is a critically important technological issue.1'2 Many of the basic phenomena have been identified, particularly the role of the interface debond energy, F,, and its dependence on the loading mixity, ty (Fig. 1).3>4 The latter is a measure of the relative shear to tensile opening of the interface crack surfaces near the tip. Methods for measuring F, have also been devised and calibrated (Fig. 2).4-15 However, these methods typically require specimens made by using a high homologous temperature (T/Tm) processing step, such as the diffusion bonding of sandwich specimens: T and Tm denote the processing temperature and the metal melting point, respectively (Fig. 2). For interfaces that have experienced only low temperatures upon fabrication, there are many tests for the qualitative ranking of the interface fracture resistance. These include scratch,7 microscratch, 89 peel,10-11 and blister tests.12'13 However, the full quantification of most of these tests has not usually been possible because of the complex elastic/plastic stress fields involved. For example, models for the microscratch test9 are based on simplifying assumptions regarding the stress state and the decohered area. One exception is the peel test, which has been quantified.10-11 Even then, deconvolution of F, from the peel force is complex and subject to appreciable uncertainties. Moreover, this test has a mode mixity that differs in sign and magnitude from that associated with most practical decohesion problems.3 Blister tests are more readily interpreted but also provide mode mixities having inappropriate sign.12'13 Such tests have further limitations when applied to thin, opaque, ductile films. There are two practical problems. The pressure imposed to induce decohesion increases the stress in the film and may cause yielding. 1734
http://journals.cambridge.org
J. Mater. Res., Vol. 9, No. 7, Jul 1994
Downloaded: 02 Jun 2014
Moreover, the perimeter of the decohered zone is difficult to measure. A preferred test method for ductile films would be one in which the stresses in the film diminish as the decohesion extends and also have an energy release rate G independent of the decohered length. The present article addresses t
Data Loading...
