Large-Area Multi-Breakdown Characterization of Polymer Films: A New Approach for Establishing Structure–Processing–Breakdown Relationships in Capacitor Dielectrics

The ever-growing need for high-energy density and high operation temperature capacitive energy storage for next-generation applications has necessitated research and development on new dielectric materials for film capacitors. Consequently, various new approaches offering unique ways to tailor dielectric properties of polymers have recently emerged, and new materials such as dielectric polymer nanocomposites (PNC) are envisioned as potential next-generation dielectrics. Establishment of optimized formulation and processing conventions is however necessary in order to achieve improvement in dielectric breakdown properties. Importantly however, such material development puts dielectric breakdown strength assessment of polymer films in a central role in guiding material development process towards highly optimized functional materials. This is not a trivial task though, as the current state-of-the-art breakdown strength measurement techniques rarely provide statistically sufficient amounts of breakdown data from the application point-of-view, thus leading to impaired evaluation of the practical breakdown performance in film capacitors.

In this thesis, a new large-area multi-breakdown measurement method enabling detailed dielectric breakdown characterization of polymer films is developed and evaluated. Various aspects encompassing sample film preparation, measurement procedure, breakdown progression, discharge event characterization, breakdown field determination, data validation and statistical analysis are systematically and critically discussed. A data qualification process based on the self-healing discharge energy and breakdown voltage characteristics is developed and shown to be a sensible and convenient way to exclude non-breakdown events from the measurement data prior to Weibull statistical analysis. The measurement method is shown to enable high-statistical-accuracy breakdown characterization of both metallized and non-metallized polymer films of different nature, including laboratory-scale, pilot-scale and commercial-grade capacitor films. Statistical aspects on the area dependence are discussed and the problematic nature of Weibull area-extrapolation of small-area breakdown data to represent larger film areas is exemplified. The fundamental differences between the large-area multi-breakdown and the small-area single-breakdown measurement methods and the statistical aspects thereof are analyzed in more detail by the Monte Carlo simulation method.

The large-area multi-breakdown method is utilized to carry out a comprehensive analysis on structure--processing--breakdown relationships in conventional polymer and polymer nanocomposite films. Analysis on the effects of film processing, structure and morphology on the large-area multi-breakdown response of cast- and bi-axially oriented isotactic polypropylene (PP) films emphasizes the determining effect of processing-dependent film morphology in large-area dielectric breakdown response of PP films. Commercial capacitor-grade bi-axially oriented polypropylene (BOPP) films are shown to exhibit differences in breakdown distribution structure and weak point behavior in comparison to the laboratory-scale BOPP films, presumably due to differences in raw material, additives, thermal history and processing. Breakdown characterization of commercial metallized polymer films as a function of inter-layer pressure also emphasizes the importance of careful breakdown characterization under authentic operation stresses in order to ensure proper design and operation in practical applications.

BOPP-based polymer nanocomposite (PNC) films are studied with a particular emphasis on the effects of various compositional, structural and film processing factors on the breakdown behavior of laboratory- and pilot-scale melt-compounded BOPP nanocomposite films incorporating silica and/or calcium carbonate nanofillers. The optimum nano-filler content is found to reside at the low fill-fraction range (~1 wt-%), however, the fill-fraction itself is not the only determining factor, as compounds with equal nanoparticle content but with differences in e.g. compounder screw speed are found to exhibit large differences in breakdown response. Indications of possible silica-antioxidant interaction are also reported. Structural imaging of the films shows that nano-structural features cannot solely explain the observed large-area breakdown behavior – this aspect points towards large-area approach being necessary for the imaging techniques as well in order to reliably establish a link between structural properties and engineering breakdown strength. The results point out that up-scaling of PNC production is sensible with conventional melt-blending technology, although further development and optimization of nanocomposite formulations and processing are seen as necessary. Analysis on the ramp-rate-dependence of the breakdown response in dielectric polymer nanocomposite films also provides perspective on the importance of careful breakdown assessment when dielectric films of more complex internal structure are studied.