Today, GaN epitaxy is grown primarily on sapphire or silicon substrates, but neither is a perfect lattice match which causes dislocation defects which will impact the internal quantum efficiency of microLEDs. Further, as the microLEDs get smaller, sidewall defects have a greater impact, significantly lowering efficiency. Thirdly, the efficiency of red is much lower than green or blue. According to Professor Jeehwan Kim from MIT, the answer to these three challenges is remote epitaxy (paper 10.2 at the 2025 Vehicle Display and Interfaces conference) which produces ultrathin freestanding nitride membranes. Such an approach is one way to make stacked full-color microLED displays.

The key to remote epitaxy is the use of a graphene layer on the starting substrate (GaAs for red and SiC for G/B). The graphene layer inherits the underlying crystal structure which means less stress in the growth (reduced defects). The graphene also forms a weak van der Waals interaction with the substrate enabling an easy lift-off capability with reuse of the substrate as well. Conventional epitaxial growth with MOCVD equipment is then used to create the films and define the microLEDs. The final epitaxial film stack is less than one micron thick. This film stack can now be mechanically exfoliated and placed on CMOS, glass or flexible substrates for further processing.

Kim’s group has recently fabricated a stacked RGB display with 5000 ppi. This is a lab demo with no performance data, so clearly there is a long way to go to commercialization. Going forward, they have plans to work on a selective mass transfer method to glass for non-monolithic display applications. The group is also looking at using red quantum dots on the blue microLED as part of the vertical stack instead of AlInGaP red.