Highly Scaled N-polar GaN HEMTs for Ultra-High Efficiency at W-Band

Abstract

The nitrogen-polar (N-polar) gallium nitride (GaN) deep recess high electron mobility transistor structure has recently demonstrated exceptional large-signal performance at millimeter-wave frequencies. Within W-band, at 94 GHz, up to 8 W/mm of output power density has been obtained which represents a 4x improvement over traditional gallium-polar (Ga-polar) devices at similar frequencies while maintaining a high power-added efficiency in excess of 27%. These observed performance benefits of N-polar GaN arise from device design advantages associated with the inverted polarization-induced electric fields relative to a Ga-polar GaN-based transistor. For millimeter-wave devices in particular, the addition of a GaN cap layer in the device access regionsprovides the dual advantage of increased access region conductivity with control over dispersion. While excellent performance has been obtained from this device structure to date, room remains to further improve gain and efficiency of this device technology. The current N-polar GaN deep recess device is most limited by gain. Therefore, in this work we propose to extend this device to higher frequency operation by addressing multiple factors to increase the gain of transistor. To increase gain, both lateral and vertical device scaling will be implemented. Improvements to the epitaxial growth by metal-organic chemical vapor deposition will be used to vertically scale the channel while maintaining high conductivity. Lateral scaling will be addressed by themodifications to the fabrication process.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 17, 2020
Source ID
N000142012130

Entities

People

  • Umesh Mishra

Organizations

  • Office of Naval Research
  • United States Navy
  • University of California, Santa Barbara

Tags

Fields of Study

  • Materials science

Readers

  • Electronics Engineering
  • Semiconductor Device Technology

Technology Areas

  • 5G
  • 5G - Internet of Things
  • Microelectronics