​

Why are gallium nitride MOSFETs, rather than HEMTs, the key to truly monolithic microLED displays?

Traditional microLED development relies on separately fabricated silicon CMOS backplanes, which introduces high assembly complexity. By shifting to a monolithic, single-crystal gallium nitride (GaN) solution, manufacturers can integrate control circuits directly on the same wafer. However, choosing the right transistor architecture is critical to avoiding compatibility and growth issues during epitaxy.

Instead of utilizing High Electron Mobility Transistors (HEMTs), this approach leverages GaN MOSFETs. By integrating an NPN transistor stack directly beneath a conventional LED structure, both vertical and lateral transistors can coexist seamlessly. This co-design enables a highly compact active matrix configuration without requiring complex regrowth steps.

This architecture enables an NMOS active matrix drive circuit for each individual microLED pixel. Compared to the PMOS configurations typically forced by standard silicon CMOS backplane bonding, the GaN NMOS design delivers significantly higher driver performance, faster switching speeds, and superior thermal tolerance under high-brightness conditions.

In this short video, you can learn:
* Why GaN MOSFETs are more compatible with standard MOCVD LED growth than traditional HEMT devices.
* How integrating an NPN transistor stack beneath the LED enables vertical and lateral transistor layouts.
* The performance advantages of an NMOS active matrix configuration over standard PMOS silicon CMOS backplanes.
📋 **Clip Abstract** Matthew Hartensveld explains how Innovation Semiconductor integrates gallium nitride MOSFETs and microLEDs on a single-crystal GaN wafer. This monolithic approach avoids separately fabricated silicon CMOS backplanes to deliver compact, high-performance NMOS active matrix microdisplays.
🔗 Link in comments 👇

How can a single-wafer monolithic architecture eliminate the multi-step mass transfer bottleneck and costly pixel redundancy in microLED manufacturing?

The commercialization of larger-format microLED displays has long been bottlenecked by mass transfer step yields. Traditionally, red, green, and blue LEDs are grown on separate wafers, diced into micro-scale dies, and then systematically stamped onto a CMOS backplane. This multi-step process introduces high mechanical yield risks that scale exponentially with pixel count.

To combat sub-optimal transfer yields, manufacturers are forced to implement costly pixel redundancy strategies. This means doubling up on the number of red, green, and blue dies placed at every single subpixel site just to cover potential placement failures or dead dies. This approach significantly increases the raw material bill of materials and complicates repair cycles.

By fabricating the driver transistors and the microLEDs together on a single monolithic GaN wafer, the entire assembly flow is simplified. This unified stack enables direct transfer from the integrated wafer to the final display substrate, cutting down intermediate transfer steps. This step reduction inherently maximizes assembly yield while completely bypassing the need for redundant pixel architectures.

In this short video, you can learn:
* The key yield bottlenecks associated with traditional multi-step mass transfer stamping of discrete RGB microLEDs.
* Why manufacturers are forced to double up on subpixel dies to achieve acceptable defect densities on active backplanes.
* How a monolithic, co-fabricated transistor-and-LED wafer simplifies the display manufacturing supply chain and lowers production costs.
📋 **Clip Abstract** This clip highlights the inefficiencies of traditional microLED mass transfer methods and the costly redundancy measures required to overcome placement defects. Matthew Hartensveld proposes a monolithic GaN solution that merges drivers and emitters on a single wafer to dramatically improve assembly yields.
🔗 Link in comments 👇