Investigations of GaN-based vertical field effect transistors for applications in high-power electronics

点击翻译按钮获取中文摘要

In this project, normally-off current aperture vertical field effect transistor concepts (CAVET) will be investigated for the analysis, epitaxy, processing, and characterization of electronic power devices based on group III-N semiconductors. The aim of this project is to investigate vertical transistors in order to enhance the key device parameters on-state resistance and breakdown voltage on a higher level than in Si devices and lateral GaN-based devices. In contrast to previous work this shall be accomplished using epitaxial approaches for the definition of the aperture. For the required thick defect-poor GaN drift layer, hydride vapor phase epitaxial (HVPE) growth will be performed and analyzed based on our long-standing experience of this method. On these advancedlow defect density substrates, the actual device structures will be grown by metalorganic vapor phase epitaxy (MOVPE) using selective area epitaxy for the required current aperture layers in order to minimize any defect generation during lateral structurization. On the processing side, vertical structuring and complex process flows need to be analyzed and performed in order to explore the full potential of the vertical transistor in terms of current flow and breakdown stability. Novel normally-off concepts in combination with the CAVET aperture shall be investigated. Full device simulations atIAF will guide and support epitaxial growth and device processing. Particular focus will be put on the required doping profiles. Structural and electrical characterization will be performed to understand the prospects and limitations of this new class of devices for high-voltage operation. The work program is defined and executed jointly by groups at the University of Ulm, University of Freiburg and Fraunhofer IAF. The background of the groups combines the required in-depth expertise for this project in terms of semiconductor epitaxy, processing, modeling and characterization.