Optical and Mechanical Properties of Cast-Dried Polyimide Films

• Main Authors: Chen, Po-Ju, 陳柏儒
• Other Authors: Liu, Ta-Jo
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/45249024775672119745

博士 === 國立清華大學 === 化學工程學系 === 97 === The objective of this research is to investigate some physical properties arising during the drying of polyimide film, namely birefringence and curl, both are caused by drying-induced stress. Curl is an out of plane displacement toward coated side or uncoated side at the edges of substrate. In general, coated film tends to shrink during drying and this shrinkage is inhibited by adherence to the substrate, which causes the build-up of a residual tensile stress and forces substrate to curl toward coated side. In microscopic viewpoint, this tensile stress also makes polymeric chains orientation parallel to the coating plane, which result in the difference between the refractive indices parallel and perpendicular to the coating plane, and that is the out-of-plane birefringence (OPBR). In this paper, the drying-stress induced birefringence and curl of soluble polyimide films was experimentally and theoretically investigated. The experimental results of OPBR and curl have been examined, and an operating window which is a region for stable and uniform film formation was also determined experimentally. In order to calculate the drying stress, the mass balance equation, the energy balance equation and the viscoelastic equation have to be solved, and then birefringence and curl can be evaluated by the stress-optical rule and the curling model. A one-dimensional (1D) model and a simple model have been developed to predict the drying stress. All governing equations of the mass balance, the energy balance and the viscoelastics are concluded in the 1D model. The computer aided solutions by the finite element method (FEM) were found. Three assumptions that would lead the simplification of the 1D model to reduce it to a simple model. The first assumption is that at the early stage of drying process, there is no concentration gradient inside the film. The second is that the variation of film thickness is negligible after stress built-up. The third is that after stress built-up, the coating material is a homogeneous viscoelastic material and deforms following a modified Maxwell model consisting of a spring connected to two parallel elements: a dashpot and a stick/slip. Because of the simplified governing equations, the simple model can be solved by the Excel spreadsheet. The predictions of 1D model are in reasonable agreement with experimental results especially for the substrates with high surface energy such as glass, but the model itself is rather complicated. On the other hand, although simple model does not predict the results as accurate as 1D model, it thus provides some physical insight on the formation mechanism of drying stress and is much easier to apply. One just has to be careful with two cases, i.e., rapid drying and large film thickness variation after built-up concentration. On-line experiments were also carried out and the results were compared with the theoretical models, it was found that the theoretical predictions are also reasonably accurate. An operating window which is a region for stable and uniform film formation was also determined experimentally, which also put the limit of drying-controlled OPBR and curl.