Selective-Area Doping to Produce Vertical GaN Power Switches

February 22, 2018 by Paul Shepard

A team of researchers from the University of New Mexico (UNM) and Sandia National Laboratories has received funding from ARPA-E to develop fundamental knowledge needed to fabricate vertical gallium nitride (GaN)-based power electronic devices. The production of high-quality vertical GaN-based devices is proving to be difficult - largely due to design and materials challenges that are not well understood.

To date, many vertical GaN-based devices have shown poor quality (e.g., low break-down voltage and high leakage current), which renders them useless in high power applications.

"To realize the full potential of vertical GaN-based power electronic devices, fundamental research is needed to better understand what is happening at a microscopic level when making these devices," said Daniel Feezell, associate professor of electrical and computer engineering and principal investigator on the UNM team.

With recent funding from ARPA-E's new Power Nitride Doping Innovation Offers Devices Enabling SWITCHES (PNDIODES) program, the teams from UNM and Sandia are focusing on selective-area doping - the process of adding selected impurities to specific regions of the semiconducting material to achieve structures called p-n junctions.

Funding for the entire project is $1.895 million over two years, with $344,700 coming to UNM. The project will fund one graduate student and one postdoctoral researcher at UNM.

"Using the metal-organic chemical vapor deposition (MOCVD) system and the Nanofabrication Facility at the Center for High Technology Materials, we are studying selective-area regrowth and related doping and defects on nonpolar orientations of GaN, which constitute the regrowth sidewalls when forming p-n junctions in etched trenches," notes Feezell.

"The goal is to enable a fundamental understanding of the regrowth interface that will lead to high-voltage diodes with low leakage current," Feezell added.

The teams at UNM and Sandia are seeking to achieve selective-area doping using patterned regrowth of GaN p-n diodes with electronic performance equivalent to as-grown, state-of-the-art GaN p-n diodes.

They are working to obtain a deep understanding of the growth process, including the relationship among crystal growth conditions, etching methods and post-etch treatments, impurity control, and electronic performance.

Feezell concluded "Our research in this area will generate knowledge needed to address materials challenges for vertical GaN-based power electronic devices and help expedite their adoption in a variety of commercial applications."


The UNM / Sandia team is only one of several projects that comprise ARPA-E's PNDIODES program. Overall, these projects seek to develop transformational advances in the process of selective area doping in GaN and its alloys.

Summaries of current PNDIODES Projects

Adroit Materials, Inc. - Cary, NC

Selective Area Doping for Nitride Power Devices - $700,000

The Adroit Materials team seeks to establish selective area p-type doping of GaN by using ion implantation of magnesium and an innovative annealing, or heat treatment, process to remove implantation damage and control performance-reducing defects. By developing an in-depth understanding of the ion implantation doping process, the team will be able to demonstrate usable and reliable p-n junctions that meet or exceed PNDIODES program targets and enable a new generation of high-performance electronic semiconductor devices.

Arizona State University - Tempe, AZ

Effective Selective Area Doping for GaN Vertical Power Transistors Enabled by Innovative Materials Engineering - $1,500,000

The ASU team proposes a comprehensive research program to advance fundamental knowledge in the selective area growth of GaN materials in order to achieve selective area doping, leading to the development of high-performance GaN vertical power transistors. The team will develop a new fabrication process and determine the opportunities to solve the challenges of selective area growth for doping in GaN materials. The team will also conduct a materials study and investigate several issues related to GaN selective area epitaxial growth. If successful, the project will demonstrate generally usable p-n junctions for vertical GaN power devices that meet PNDIODES program targets.

JR2J, LLC - Ithaca, NY

Laser Spike Annealing for the Activation of Implanted Dopants in GaN - $647,750

The JR2J team seeks to use a fast, high-temperature technique called laser spike annealing (LSA) to activate implanted p-type dopants in GaN. This technique allows for the high temperatures necessary to activate the dopants, as well as to repair damage done during the implantation process. By keeping the laser spike duration very short (0.1-100 milliseconds), the technique also hopes to avoid damage to the GaN lattice itself. The team will experiment with various LSA annealing conditions, exploring temperatures and time scales of the technique.

Sandia National Laboratories - Albuquerque, NM

High Voltage Re-grown GaN P-N Diodes Enabled by Defect and Doping Control - $1,894,700

As described above, the Sandia team seeks to achieve selective area doping using patterned regrowth of GaN p-n diodes with electronic performance equivalent to as-grown state-of-the-art GaN p-n diodes. The team will work to obtain a deep understanding of the growth process, including the relationship among crystal growth conditions, etching methods and post-etch treatments, impurity control, and electronic performance. The team also seeks to address challenges presented by the regrowth technique using physics-based approaches.

State University of New York Polytechnic Institute - Albany, NY

Demonstration of PN-junctions by Ion implantation techniques for GaN (DOPING-GaN) - $720,000

The SUNY Poly team will focus on ion implantation as the centerpiece of its approach. Using new annealing techniques, the team will develop processes to activate implanted silicon or magnesium in GaN to build p-n junctions. P-type ion implantation and annealing will be performed using an innovative gyrotron beam (a high-power vacuum tube that generates millimeter-wave electromagnetic waves) technique and an aluminum nitride cap. Central to the SUNY Poly proposal is understanding the impact of implantation on the microstructural properties of the GaN material and effects on p-n diode performance.

University of Missouri - Columbia, MO

High Quality Doping of GaN through Transmutation processing - $250,000

The University of Missouri team will focus on the development of neutron transmutation doping—exposing GaN wafers to neutron radiation to create a stable network of dopants within—to fabricate an extremely uniform n-type GaN wafer. Specific innovations in this proposal concern an in-depth study of neutron transmission doping and a characterization of the resulting wafer, including analyzing electrical resistance, dopant concentration, unwanted impurities, and damage to the GaN lattice.

Yale University - New Haven, CT

Regrowth and Selective Area Growth of GaN for Vertical Power Electronics - $1,150,000

The Yale team seeks to conduct a comprehensive investigation to overcome the barriers in selective area doping of GaN through the regrowth process for high-performance, reliable GaN vertical transistors. The team will demonstrate vertical GaN diodes through regrowth and selective area growth 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 address the regrowth process at the nano scale, control of the crystal growth process to control impurities, electronic defects in the regrowth and selective area growth processes, and customizing the electronic characteristics of the selective area growth active region.