Selective Area Growth for Vertical Power Electronics
Electricity generation currently accounts for ~40% of primary energy consumption in the U.S. and continues to be the fastest growing form of end-use energy. Power electronics are responsible for controlling and converting electrical power to provide optimal conditions for transmission, distribution, and load-side consumption. By 2030 as much as 80% of all electricity could pass through some form of power electronics. Applications for power electronics are widespread and include uses in power supplies, motor drives, grid applications, data centers, and distributed energy resources. Today, most power electronics are based on silicon semiconductor devices that have reached their efficiency limits at high power and frequency, due to the material limitations of silicon. Wide-bandgap (WBG) semiconductors such as gallium nitride (GaN) have superior electrical conductivity, breakdown properties, and switching speed. This allows for power converters with much improved efficiencies over silicon - while also dramatically reducing system size, weight, and form factor. Power semiconductor devices overwhelmingly use vertical architectures to realize high breakdown voltage (>1200V) and current levels, without having to enlarge chip size. The vertical architectures require the ability to add specific impurities to selected regions of a semiconductor to produce negative (n-type) and positive (p-type) electrical conduction, a process called doping. Currently, no doping process exists to form selective p-type regions in GaN. This is the major barrier to realization of GaN based vertical power electronic devices. The development of a selective p-type doping process will enable vertical GaN device architectures and unlock the potential of using the WBG semiconductor GaN in power electronics.
Project Innovation + Advantages:
Yale University will conduct a comprehensive investigation to overcome the barriers in selective area doping of gallium nitride (GaN) through an epitaxial regrowth process for high-performance, reliable GaN vertical transistors. Transistors based on GaN have emerged as promising candidates for future high efficiency, high power applications, but they have been plagued by poor electrical performance attributed to the existing selective doping processes. The team will demonstrate vertical GaN diodes through a selective area regrowth processes with performance similar to those made using current in situ techniques, which are non-selective and therefore less flexible. Key innovations in this project will be to use three-dimensional nanoscale characterizations to understand the regrowth interface formation at the nano scale, and to apply atomic-level manipulation to control impurities, and suppress extrinsic and intrinsic defects at the selective area regrowth interface. This will enable the electronic characteristics of the selective area growth p-n junction active region to be customized allowing for high performance GaN vertical transistors. The successful production of reliable and high-performance GaN vertical transistors on bulk substrates will be transformative to many mid-voltage applications including photovoltaic inverters, motor control, and hybrid automotive.
If successful, PNDIODES projects will enable further development of a new class of power converters suitable in a broad range of application areas including automotive, industrial, residential, transportation (rail & ship), aerospace, and utilities.
More energy efficient power electronics could improve the efficiency of the U.S. power sector. They could also significantly improve the reliability and security of the electrical grid.
More efficient power use may help reduce power-related emissions. Low-cost and highly efficient power electronics could also lead to increased adoption of electric vehicles and greater integration of renewable power sources.
Improved power electronics could yield a significant reduction in U.S. electricity consumption, saving American families and businesses money on their power bills.