Estimated reading time: 5 minutes

Flexible Thinking: ‘Let There Be Light’–Emitting Diodes!

Display technology has become ubiquitous in today’s electronic products, driven largely by smartphones. This touchscreen technology increasingly dominates our means of communication, both for voice and text. Our connection to the world outside our homes, cars, and offices is now enjoyed 24/7 via the internet, to which they are invariably connected. In terms of time and rate of social change, such devices have gone from “nice to have” to “indispensable” in the seeming blink of the proverbial eye.

The key has been the mind-numbing advances in visual display technology, particularly the light-emitting diode (LED). The technology has roots back to the discovery of electroluminescence by H. J. Round in 1907, but did not see much commercialization until the red-light LED was demonstrated by Nick Holonyak at GE in 1962. Holonyak is commonly referred to as the “Father of the LED.”

The devices were not all that bright, but they were much less power hungry than the neon orange light-emitting “Nixie tubes,” which were common in the 1970s. They set the stage for the development of formats and colors: orange, yellow, green, and blue ushered in low power, high lumens per watt (lm/W). In home lighting, these long-lasting white-light LEDs have replaced incandescent bulbs , where most of the energy consumed is lost as heat. Efficiency has also increased rapidly. In 2000, a white light LED delivered roughly 20 lm/W; by 2010, it delivered 100 lm/W, and today the efficiency is sitting at an amazing 150–250+ lm/W.

Display technology using LEDs has relatively quickly gone through several stages in package formats. Initially, it was the dome-shaped or tubular LED, familiar to many electronics hobbyists. These are still relatively large and are most often used for larger-format outdoor displays because the individual pixels cannot be easily distinguished at a distance; however, they are unsuitable for personal electronics.

Developers created newer formats, including the first surface mount versions. What is truly amazing is that the internal construction of the original LED, a semiconductor die mounted on a lead frame with a bond wire and clear encapsulant, is nearly indistinguishable, conceptually, from modern LED packages.1 The major difference is that today's microLEDs can be 1,000 to 10,000 times smaller than those early through-hole devices.

Mini and microLEDs are often used in televisions, monitors, and smartphones. The technology progression for light-emitting electronic components is shown in Figure 1. This brings us to the present and the significant challenges of assembling devices as small as a mote of dust. The limitations of conventional pick-and-place assembly, the assembly of microLED displays, are summarized in the technology comparison chart shown in Table 1.

MicroLED technology employs microscopic light-emitting diodes, typically fabricated from gallium nitride (GaN), that individually function as red, green, or blue pixels. Depending on display resolution and application, driven by the distance of the viewer’s eye to the display, LED devices used in displays may range from approximately 3 to 50 µm in size, with many advanced display architectures targeting device dimensions in the 5–20 µm range. For perspective, a human hair is approximately 70–100 µm in diameter.

The manufacturing challenge is staggering. A 4K television contains more than 8 million pixels, and because each pixel generally requires separate red, green, and blue emitters, the display may require placement of over 24 million individual microLED devices. Future 8K displays may require nearly 100 million emitters.

Therein lies the rub. Conventional pick-and-place systems, which excel at placing packaged components measured in millimeters, become increasingly challenged when required to handle millions of fragile semiconductor devices measured in micrometers. At these dimensions, device handling, alignment accuracy, throughput, contamination control, and yield become major obstacles to commercialization. Even a small defect rate can translate into thousands of failed pixels, and even a single pixel failure in a field of millions can be picked up by the human eye.

The simple reality is that microLED devices are extremely thin and fragile, and any transfer process must avoid mechanical damage to the delicate devices while maintaining placement accuracy, often measured in a few micrometers. Furthermore, the economics of display production require that these transfers occur not one device at a time, but in massively parallel fashion.

For what it’s worth, I was invited to write about this topic for July’s issue, but it was quite propitious as I have been friends since 2008 with and an advisor to Dr. Jayna Sheats, founder of Terecircuits (it has changed names over the years). Dr. Sheats is a brilliant scientist with a PhD from Stanford and valuable experience as a researcher at HP Labs in Palo Alto.

The reason I say “propitious” is because Sheats and her team at Terecircuits have developed a unique and useful technology called Terefilm that seems almost custom-designed for microLED assembly, though its usefulness goes far beyond. The novel photopolymer platform is particularly interesting in that, rather than relying on mechanical pickup and placement of individual devices, Terefilm enables temporary bonding and highly selective release of semiconductor and/or other discrete devices (i.e., not just microLEDS) through photochemical activation and total gasification of the bonding material. Devices or groups of devices can be released from a donor substrate and transferred with high precision while minimizing mechanical handling and obviating cleaning.

In Table 1, the significance of the Terefilm approach to microLED assembly lies in its ability to replace serial assembly operations with highly parallel transfer processes as illustrated in the above chart. Note again that the technology is not limited to microLEDs alone but can be adapted to handling and assembling bleeding edge and prospective next generation discrete devices such as the 008004 (0.25mm × 0.125mm) and the experimental <0.2mm × 0.1mm.

For microLED manufacturing, the Terefilm approach thus offers several potential significant advantages:

• Simultaneous transfer of thousands or millions of devices
• Reduced mechanical stress on fragile emitters
• Improved placement throughput
• Lower contamination risk
• Enhanced yield potential
• Reduced manufacturing cost

While microLED displays represent one of the most visible applications for Terecircuits’ technology, the same challenges increasingly appear today in advanced semiconductor packaging, chiplet integration, heterogeneous assembly, silicon photonics, RF modules, and embedded electronics. In each case, manufacturers are confronted with the task of accurately positioning often vast numbers of extremely small and delicate devices at economically viable production rates—a challenge that will not get easier with current methods.

As electronic systems continue to shrink while component counts increase, the industry's future may depend less on how quickly machines can pick up components and more on how effectively materials and processes can move them in parallel. Technologies such as Terefilm may therefore represent not merely an improvement in assembly efficiency, but a fundamental shift in how next-generation electronic systems are manufactured.

A Closing Thought

Terefilm combines temporary bonding and debonding functionality with lithographic pattern ability. Through photochemical activation, individual devices or groups of devices can be selectively released from a donor substrate and transferred with high precision. Upon activation, the material decomposes cleanly into gaseous products without leaving residue, minimizing contamination and eliminating many cleaning requirements associated with conventional release methods. It seems to me, and to a growing number of others, that the technology has a definite “place at the electronics assembly table.” Check it out.

References

1. My former company, Tessera (thrice renamed and now called Aedia), applied for a patent for one of my disclosures, “Semiconductor Package Having Light Sensitive Chips in 1997. It concerned semiconductor packaging structures incorporating light-emitting semiconductor devices within a package architecture, and according to AI, the patent is cited in later optoelectronic and semiconductor packaging patents and appears repeatedly in patent citation records.

This column originally appeared in the July 2026 issue of I-Connect007 Magazine.