Linda Schadler1 Aditya Prasad1 L Brinson2 Wei Chen3 Yanhui Huang4 Yichi Zhang3 Yixing Wang3

1, Rensselaer Polytechnic Institute, Troy, New York, United States
2, Duke University, Raleigh, North Carolina, United States
3, Northwestern University, Evanston, Illinois, United States
4, Lam Research Corporation, Livermore, California, United States

Nanodielectrics are extremely promising materials for use in high voltage cable transmission and as capacitor materials. It is now well accepted that the high volume of interfacial material creates opportunities for trapping of carriers, a reduction in conductivity, and a concomitant increase in dielectric breakdown strength and endurance strength. It has also become clear that the dispersion of the nanofiller is critical to optimization of both breakdown strength and permittivity. Thus, we have developed a ligand engineering approach that uses a bimodal population of molecules to control both dispersion and trapping behavior. A low graft density of long polymer chains ensures compatibility with the matrix and potentially entanglement or crosslinking with the matrix. A high graft density of charge trapping molecules modifies the composite trapping behavior. Using a variety of characterization methods including: PEA (Pulsed Electroacoustic Analysis), UV vis, microscopy, dielectric spectroscopy, breakdown strength, and endurance measurements, we have characterized the trapping and dielectric behavior of a variety of nanodielectrics. We have also developed a multiscale model that ranges from the nanoscale trapping behavior to the macroscopic breakdown behavior to attempt to create a methodology for nanodielectric design.

As part of a broader approach, we are developing dielectric data and tools for a new open source nanocomposite data resource, NanoMine. NanoMine has a growing set of data, microstructure quantification and reconstruction tools, and models that specifically incorporate interface behavior. We have run a comprehensive set of experiments that vary the mixing energy, shear stress, and strain rate in a polymer melt, and use a large set of TEM images, and NanoMine tools to quantify microstructure descriptors such as interfacial area, cluster size and volume fraction. We are using the data in NanoMine to correlate quantitative microstructure descriptors with dielectric properties with the goal of being able to develop design tools that identify the appropriate fillers, dispersion, and processing methods to meet specific combinations of properties. Such an approach to simulation and design could eventually accelerate the process of nanocomposite adoption for dielectric applications.