Solvent-induced damage in polyimide thin films
- PDF / 1,652,874 Bytes
- 10 Pages / 576 x 792 pts Page_size
- 35 Downloads / 216 Views
Solvent induced crazes formed in strained polyimide thin films on different substrates have been studied. A fracture mechanics approach has been used to simulate craze evolution. The experiments and simulations have identified a critical prestrain below which craze formation does not occur. This strain decreases with increase in solvent exposure time, but also exhibits a threshold. Diffusion of the solvent into the film is considered to be responsible for the time-dependent nature of damage formation.
I. INTRODUCTION Polyimide films used for device packaging are susceptible to solvent induced damage, manifest as bands that extend across the film.1"3 The bands are motivated by tensile residual stress, caused by the thermal expansion misfit between the polyimide and the substrate. Prior studies of these bands have revealed that the tensile stresses are relaxed in the immediate vicinity of the bands,1 suggestive of craze or crack-like behavior. Research on the propagation of crazes and cracks within both bulk polymers and polymer films, subject to environmental degradation, has been widely addressed with the context of linear elastic fracture mechanics.4"8 The basic approach has been to presume that the material outside the crack is subject to essentially timeindependent small-scale yielding, while the crack faces within the craze zone experience tractions induced by the fibrils. The effects of the environment are then expressed through degradation of either the fibrils or the cracktip bonds. Based upon this background, the present article has the objective of providing a fracture mechanics approach for characterizing solvent induced damage in polyimide films. Given the inherent complexity of the problem, a combined experimental and modeling approach has been adopted.
laxation has important effects on the damage evolution and also complicates the interpretation of experimental information. B. Approach Among the variables expected to have an influence on damage within the films are the film thicknesses and the stress, as well as defects in the film. By using spin coating techniques, the film thickness can be varied over a wide range. However, reliance on thermal expansion misfit to govern the stress provides only a limited range. To extend this range, the films are produced on a ductile substrate. Then, by stretching the substrate, additional stresses can be introduced. An Al alloy substrate (6061T6) was used for this purpose. C. Procedures The films were spin-coated onto substrates, consisting of thin disks of either Si or glass and tensile specimens of 6061-T6 Al, then cured using temperatures up to 400 °C. The final thickness of the films was controlled between 2 and 20 /xm by adjusting the spin rate and the viscosity of the precursor. Subsequent to curing, strain gauges (manufactured by HBM) capable of Polyimide Film
II. EXPERIMENTAL
Strain Gauge
A. Material The material used in the present investigation (Thermid IP 6015) has the following characteristics. The material is essentially elastic at room temperatur
Data Loading...
