Dec 3, 2024
9:00am - 9:30am
Hynes, Level 1, Room 110
Michael Snure1
Air Force Research Laboratory1
Integration of electronic materials and devices on to arbitrary substrates is of great interest to produce more compact and capable electronics systems. In particular, integration of GaN devices, like the high electron mobility transistor (HEMT), are of interest due to their high frequency and power handling capabilities, which far surpass Si and III-V based electronics. So far integration at the chiplet level has proven successful, improving power, and performance, while reducing area, design time and fabrication costs. By integrating just the thin active layers and bonding them to a common substrate, even closer integration can be achieved further improving the size, weight, and power (SWAP). Unlike Si and III-Vs, which have well established commercial epitaxial lift-off (ELO) methods, lift-off of GaN layers and devices from the growth substrate is particularly challenging. Of the ELO technologies for GaN, van der Waals (vdW) based lift-off offers great promise but is still in the early stages of development and has yet to demonstrate the scale and yield of other ELO technologies. vdW lift-off involves growth of GaN on a vdW buffer layer, BN on sapphire is used in this work, followed by mechanical separation of the GaN films from the substrate. We will highlight the potential of this process for integration of GaN films and HEMTs at scale. Both tape/stamp based mechanical lift-off and Ni spalling based lift-off will be compared. The effects of the BN vdW template and GaN epitaxy on transfer yield, area, and damage will be covered with the goal of achieving high quality films that can be easily lifted off and transferred. Finally, the impact of this transfer method on GaN HEMT performance will be discussed.