Is the microLED the next display revolution?

Displays based on microLEDs combine exceptional contrast with a vast colour gamut, tremendous brightness and a great deal of ruggedness, but will production be held back by their complexity?

BY ERIC VIREY FROM YOLE Dà‰VELOPPEMENT

Despite the rapid growth in solid-state lighting, the backlighting of displays remains a substantial market for the LED. For more than a decade, screens have been illuminated with these devices "“ initially housed in a traditional package, and more recently the chip-scale package "“ and they are now the omnipotent lighting source in LCDs.

A related success story for the packaged LEDs is as the illumination source in large video billboards, which are a common sight in stadiums, malls and video facades. Here, discrete-packaged LEDs containing red, green and blue chips form individual pixels, with pitches typically ranging from 1 mm to 40 mm, depending on the size of the display and its resolution.

As of today, LEDs have never been used as the direct emissive element "“ that is the pixel "“ in small-pitch consumer displays. Many issues prevent this from happening, including concerns related to cost and manufacturability. However, the idea of building a display with microLEDs and sub-millimetre pixel pitches can be traced back to the time when the LED was in its infancy.

During the last five years, interest in developing microLED-based displays has taken off. In 2014, excitement in the tech and display communities skyrocketed after Apple acquired Luxvue, a microLED display start-up. Last October, Oculus, the augmented reality/virtual reality (AR/VR) arm of social networking behemoth Facebook, bought microLED start-up InfiniLED and this May, Sharp, now part of the Hon Hai Foxconn group acquired eLux, another microLED outfit.

Given these acquisitions, the technology is not just a lab curiosity. What is driving this tremendous interest by consumer electronic OEMs and leading brands? It is that this technology, which features individual red, green and blue sub-pixels as independently controllable light sources, is capable of forming displays with high contrast, high speed, and wide viewing angles "“ attributes also found in the pricey OLED displays.

In fact, microLED displays should have the upperhand over OLED rivals, thanks to a wider colour gamut, a brightness that is orders of magnitude higher, a significantly reduced power consumption, a longer lifetime, greater ruggedness and superior environmental stability. What's more, as illustrated by Apple's recent patent filings, microLEDs could allow the integration of sensors and circuits, enabling thin displays with embedded sensing capabilities, such as fingerprint identification and gesture control.

Although microLEDs are still to reach the market, they are far more than just an idea on a drawing board. Back in 2012, at the Consumer Electronic Show, Sony showcased a full HD 55-inch television that featured microLEDs. This display, which received rave reviews from video enthusiasts walking by the booth, contained 6.2 million sub pixels "“ each an individually controllable microLED chip. However, Sony was noncommittal regarding a commercialisation timeline, and as of today, no microLED TV has ever made it into market.

An inherently complex technology

Today, there isn't a commonly accepted definition for microLEDs. However, in general, they are considered to be LED die with a total surface less than 2500 µm². This corresponds to a 50 µm x 50 µm square, or a circular die with a diameter of 55 µm. Based on this definition, microLEDs are on the market today "“ they were unveiled again by Sony, in 2016, in the form of a small-pitch, large LED video wall, with traditional packaged LEDs replaced by microLEDs.

The big question for both the LED and display industries is this: how far off is the small-pitch, consumer microLED display? It is this that can target cell phones, smartwatches, TVs, laptops and, more recently, virtual, augmented, and mixed reality, head-mounted devices.

Multiple challenges must be overcome before it is possible to realise the potential of microLEDs.

Getting there will not be easy. The art of making microLED displays involves processing of a bulk LED substrate into an array of microLEDs that are poised for pick-up and transfer to a receiving substrate, for integration into a heterogeneously integrated system: the display, which integrates LEDs, pixel-driving transistors, optics and so on. The epiwafers accommodate hundreds of millions of microLED chips, compared to just thousands for traditional LEDs.

There are two leading options for realising a display with microLEDs. One is to pick up and transfer the microLEDs individually, or in groups, onto a thin-film transistor driving matrix that is similar to the ones already used in OLED displays; and the other is to unite a full monolithic array of hundreds of thousands of microLEDs with a CMOS driving circuit.

If the first of these two approaches is adopted, assembling a 4K display requires picking up, positioning and individually connecting 25 million microLED chips (assuming no pixel redundancies) to the transistor backplane. Manipulating such small devices with traditional pick and place equipment produces a processing speed of around 25,000 units per hour. That's far too slow "“ assembling a single display would take over a month.

To address this concern, companies such as Apple, X-Celeprint and dozens of others have developed massively parallel pick and place technologies. They can process tens of thousands to millions of microLEDs simultaneously. However, when microLED sizes are just 10 µm, handling and positioning them with sufficient accuracy is very challenging.

There are also issues to overcome with the LED chip. When its dimensions are very small, its performance is held back by nefarious sidewall effects related to surface and subsurface defects, such as open bonds, contamination and structural damages. These imperfections lead to a hike in non-radiative carrier recombination. Sidewall effects can extend over distances similar to the carrier diffusion length, typically 1 µm to 10 µm: that's not a big deal in conventional LEDs, which have sides of hundreds of microns, but it's a killer in microLEDs. In these devices, it can limit the efficiency of the entire volume of the chip.

Due to these flaws, the peak efficiency of a microLED is often below 10 percent "“ and it can be less than 1 percent when device dimensions are below 5 mm. That's far lower than the best traditional blue-emitting "˜macro' LEDs, which can now produce peak external quantum efficiencies exceeding 70 percent.

Making matters even worse, microLEDs often have to be operated at very low current densities. They are typically driven way below the 1-10 A cm^-2 peak efficiency region, where traditional mid-power LEDs operate, because even at this low efficiency the LEDs are incredibly bright. If a cell phone had microLEDs operating at their peak efficiency, its display would deliver a brightness up to tens of thousands of nits, which is more than an order of magnitude higher than the brighter phones on the market today. The screen would be so bright that it would dazzle any user bold enough to look at it.

When LEDs operate at a very low current density, their efficiency is so low that the technology cannot fulfil its promise of trimming energy consumption. Consequently, addressing this issue is a key priority for companies involved in microLEDs. Options for increasing efficiency include the introduction of new chip designs and improved manufacturing technologies. Both approaches could reduce sidewall defects and keep electrical carriers away from the edges of the chip.

Developers of microLEDs also face challenges related to colour conversion, light extraction and beam shaping. All are subjects of intense research, licensing, and merger and acquisition activities.

Another requirement for modern displays is the elimination of dead or defective pixels. It is nothing short of utopia to realise a 100 percent combined yield in epitaxy, chip manufacturing and transfer. So microLED display manufacturers must develop effective defect management strategies. They could include pixel redundancies and individual pixel repair, with the approach governed by the characteristics and the economics of the display.

Low hanging fruits

MicroLED are capable of being deployed in any display application, from the smallest to the largest. In many cases, they would be even better than the ultimate combination of LCD and OLED displays. But feasibility and economic realities will initially limit the reach of displays to devices where they offer a significant or disruptive gain in performance and functionality, at a cost that is acceptable to the product.

Examining in detail the strengths, weaknesses, opportunities and threats for every single application that the microLED display could serve is well beyond the scope of this article. However, detailed analysis indicates that smartwatches and other wearables, such as microdisplays for AR/MR applications, are the most likely applications to initially showcase the capability of microLED displays.

Of these, it is the smartwatch that represents the lowest hanging fruit: it has a relatively small number of pixels and a mid-range pixel density. These characteristics lead to high cost efficiency at the chip and the assembly level. Due to this, displays are nearly within reach of the current technological development status of the microLED "“ and they have potentially differentiating features, including a reduced power consumption that allows for a longer battery life, and a higher brightness that enables outstanding outdoor readability.

If these displays start to take off, the technology may advance, with the introduction of various sensors within the display front-plane, such as those that can read fingerprints and provide gesture recognition.

Another major opportunity for microLEDs is in augmented- and mixed-reality headsets. Unlike virtually reality, where the user wears a fully enclosed head-mounted display that visually isolate them from the outside world, AR and MR applications overlay computer-generated images onto the real world.

MicroLED displays are made by dicing wafers into tiny devices, and tranferring them with a parallel pick-and place technology to a transitor backplane.

One of the requirements for these applications is that the overlaid image is bright enough to compete with ambient light, especially in outdoor applications. Another criterion is that the display must not be physically located in front of the eye "“ there it would obstruct the user's vision of the real world.

To satisfy these conditions, the display must be located in an unobtrusive location, with the image projected onto the eye, using either complex projection or waveguide optics with an optical efficiency of less than 10 percent. These requirements dictate that the display brightness ranges from 10,000 to 50,000 Nits, which is more than 10 times to 50 times that of the brightness of the best cell phone on the market.

Today, the microLED is the only candidate that has the potential to offer these levels of brightness while maintaining a reasonable power consumption and compactness. Encouragingly, the same reasoning can be applied to head-up displays in automotive and other environments "“ this class of display can be considered as a form of AR.

MicroLEDs can target a range of applications, which require different pixel densities.


A market where the microLED will struggle to make an impact is that of the smartphone. Here OLED displays are already delivering outstanding performance at a very competitive cost. If microLEDs are to be in the game, the size of the subpixel must be reduced to just a few microns. That will make it even harder to deliver acceptable efficiencies, and implement a parallel transfer approach. Success could come more easily in the TV. In this case, the drawback is that pixel densities are relatively low, with a spacing of around 100 µm in a 4K, 55-inch set. The low density hampers the efficiency of the transfer technology, because each cycle needs to move several thousand chips, rather than the hundreds of thousands needed for smartphones or smartwatches. Thriving in this market requires the development of alternative high-throughput assembly techniques.

But where's the supply chain?

There are many large companies and multiple start-ups working on microLEDs, from LED makers such as Epistar, Nichia and Osram; to display makers like Sony, AUO, BOE and CSOT; and original equipment manufacturers such as Apple and Facebook/Oculus. Undertake a thorough patent search and analyses

and you'll find more than a hundred companies that have filed for patents related to microLEDs.

Success in this market requires three major disparate technologies and supply chain elements to be brought together: LEDs, thin-film transistor backplanes and chip transfer. This creates a supply chain that is more complex and involved, compared to that of traditional displays. Each process is critical, and it will be challenging to manage every aspect effectively.

No single player is going to solve all the issues, and it's unlikely that a fully integrated manufacturer will emerge. In the smaller markets, such as augmented reality, small companies could bring together the different technologies needed for the product. But that's unlikely to happen in high-volume consumer applications, such as mobiles or TVs, where a strong push from a leading OEM is probably needed to establish a supply chain. Today, Apple has enough leverage and financial strength to be able to bring all partners together. Others that might also be capable of doing this include Oculus, which has invested in microLEDs, with a focus on AR/MR applications.

Within the supply chain, each player will try and capture as much added value as possible. For LED makers, two requirements particular to the microLED "“ very low levels of defects and high-resolution features "“ imply large investments in new clean room and lithography equipment. These are criteria that might be better suited to CMOS foundries. Traditional display makers will also come up against new challenges. They are used to manufacturing back and front planes in an integrated fashion, before delivering finished panels to OEMs.

A move to microLEDs could result in these players just providing a thin-film transistor backplane to producers of the final display assembly: either OEMs or outsourced semiconductor assembly and test players. There will be some winners from the introduction of microLED displays, regardless of the shape of the supply chain. Equipment manufacturers, including makers of MOCVD equipment, will increase their sales, as will suppliers of wafers.

How long will we have to wait until we see the first consumer applications? The science is established, but the microLED is an inherently complex display technology, with cost drivers different to those of incumbents, namely LCD and OLED displays. Based on the latest developments, and the level of maturity in the supply chain, it will be 2019 at the very earliest before a high-volume application hits the market.

There is a good chance that the microLED will succeed in various sectors. However, it is still too early to say whether it will take the industry by storm, or will crash and burn like many other "˜promising' technologies of the past. In any case, the vast accumulated and ongoing research and development on the topic should bear fruit and cross pollinate into other applications. It should lead to better, more efficient LEDs; high speed Li-Fi communication; and micro-device transfer technologies that aid other industries.