Thin Film Adhesion Study in Microelectronic Packaging
- PDF / 342,711 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 60 Downloads / 250 Views
THIN FILM ADHESION STUDY IN MICROELECTRONIC PACKAGING
H.S. JEONG, Y.Z. CHU*, M.B. FREILER, C. DURNING* Department of Electrical Engineering and Microelectronics Sciences Laboratories *Department of Chemical Engineering Columbia University, New York, NY 10027
AND R.C.
WHITE
ABSTRACT Fracture energy (G,) of BPDA-PDA polyimide (PI) on modified and unmodified Si surfaces was measured by the "blister" test as a function of final cure temperature. It is proven quantitatively that surface modification prior to thin film deposition enhances adhesion. Metal adhesion to PI was also measured by the same method. Reproducibility of the data was found to be exceptionally good for both cases. The linear elastic model is quite valid for the test of thin film adhesion. Therefore, it is believed that this test is best suited for G, measurements in the study of thin film adhesion for microelectronic packaging. 1. INTRODUCTION Adhesion measurements of thin films are an immensely complicated process regarding the bonding strength that can occur between any pair of materials. Testing interfacial adhesion typically involves applying external tractions by some means to the surface of the film. The application of this external traction generally results in a complex stress field at the interface because of the load application mechanism, the dissimilar materials involved and often large deformation. Experimentally measured adhesion strength - the force or work necessary to delaminate the two materials - then depends not only on the interfacial bonding strength, but also on many other factors such as the film residual stress, plasticity effects, thickness and, of course, testing mode. Adhesion testing methods such as the peel, pull, and scratch tests are generally qualitative or comparative in nature, although they can give valuable data if used intelligently [l]. A fracture mechanics approach - a crack propagation test without external traction - offers a unique way to quantify interfacial adhesion independent of the testing procedures used. This "blister" test requires the blister to propagate along the interface; a typical adhesion fracture mechanics test which has several advantages over any other testing method. Although described in the literature, the method has never been applied in a systematic way to the thin film structures for microelectronic packaging. Its main advantage is that most of the energy expended in the test contributes to the interfacial fracture, making possible determination of a value much closer to the true interfacial bonding strength between a thin film layer and a substrate. In addition, the concept of thermodynamic energy balance is used to relate the experimentally measured debonding energy to the adhesive fracture energy. Since this approach accounts for the total work expenditure in the debonding process, difficulties inherent in the stress analysis can be avoided. In these experiments, G, - the energy dissipated per unit Mat. Res. Soc. Symp. Proc. Vol. 239. 01992 Materials Research Society
548
inte
Data Loading...
