Removing GaN From GaN
Photoenhanced wet etching of InGaN leads to lift-off of high-quality heterostructures by Chris Youtsey and Robert McCarthy from Microlink Devices and Andy Xie from Qorvo
In the early days of the compound semiconductor industry, device production involved the growth of lattice-matched epilayers on GaAs and InP substrates. Now the range of materials has expanded to include GaN, which may be grown on mismatched substrates such as sapphire, SiC and silicon.
For all these material systems, the substrate plays an essential role in the growth process, serving as a crystalline template for building up a device structure, layer by layer. But once the growth is complete, the substrate is merely an expensive handle. It can even be suboptimal, with the device delivering a better performance – such as superior light extraction, reduced weight, enhanced thermal conductivity or increased flexibility – when its layers are transferred intact to a new, better handle.
One way to enjoy these benefits is to mechanically thin the substrate. However, that’s not ideal, as it’s a destructive process. Better is the use of an epitaxial lift-off (ELO) process, as this does not damage the substrate during the removal of the epilayers. This technique uses a sacrificial release layer that is selectively etched to separate the epitaxial layers from the substrate. By taking this approach, it is possible to reuse the substrate many times, promising considerable cost savings.
For the past ten years our team at Microlink of Niles, IL, has been pursuing this approach by developing an ELO technology for compound semiconductor materials. Our successes during that time have led to the use of ELO in the volume manufacture of inverted metamorphic multi-junction solar cells with a 6-inch diameter. These cells combine an exceptionally high efficiency of more than 30 percent in AM0 conditions with a specific power exceeding 3000 W/kg. The removal of the substrate holds the key to such a high specific power, and it also trims costs through its re-use. The very high efficiencies and specific powers of our cells make them strong candidates for serving the emerging market of high-altitude, long-endurance unmanned aerial vehicles. With this solar technology, flights can last for weeks and even months.
We are now breaking new ground by extending the capability of our technology so that it can be applied to GaN. In this material system we have pioneered the development of a relatively fast wafer-scale photochemical lift-off process that does not have to draw on an exotic material for the release layer – instead, it uses InGaN.