Article: Coloring the Micro-LED revolution in AR, SmartPhone and TV Displays
By David Wyatt
Since the high-profile acquisition of Micro-LED innovators LuxView (Apple), and InfiniLED (Facebook / Oculus), and Sony's breathtaking "CrystalLED" wall-sized displays (at CES and NAB-show), Micro-LED’s (or uLED’s) have quickly become a hot-spot of attention.

And why not? Micro-LED's offer many of the advantages of emissive OLED displays, and yet are based on inorganic semiconductor chip-fabrication processes. Unlike sensitive organic-materials based Quantum Dots and OLED, Micro-LED displays and lighting products have zero heavy metals, and (inorganic) LED-like performance i.e. 30~100KHrs, with extremely high brightness, orders of magnitude beyond the industry best OLED – no surprise this new entrant is finding much excitement.

Quick recap: the tiny (3~10um) Micro-LED junction is grown on semiconductor wafers (like low-cost IC's) and then the bare junction die is cut, cherry-picked and placed directly onto the TFT backplane; which could be a Micro-Display chip (for AR), a large cheap A-Si glass backplane (for a SmartPhone or TV), or even a PCB for a large wall-sized outdoor display (e.g. Sony).

Since the backplane cost is decoupled from resolution & size, Micro-LED have highly scalable economics from small to large outdoor signage. Even application on amorphous-Si TFT backplanes, as used for lower-cost LCD (rather than dependence on LTPS needed by OLED) for cheaper/longer-lasting SmartPhone or SmartWatch displays. They support mounting on many types of flexible transparent backplanes without the need for thicker/rigid encapsulation layers, and due to their tiny feature size (3~10um), Micro-LED’s can support see-through flexible displays. Other enviable (inorganic) LED attributes include wider operating temperatures (up to 100~125C max, instead of 50~70C).

With extremely high luminosity up to 1,000,000 nits (rivaling direct sunlight), together with fine pixel micron pitch, they could be perfect for high-density micro-displays (6000ppi has been demonstrated, albeit monochromatic ... and we'll get to that), needed in AR head-mounted-display (HMD), and heads-up display (HUD) applications; hence the flurry of interest from Apple & Oculus. These high radiant flux applications, which could have junction temps of 150C, are all possible because of the robust inorganic semiconductor foundations. 

By comparison: when does anyone in the OLED industry realistically expect to find an organic-chemistry-based, high-brightness, high-flux-denisty, emitter of high-energy particles (i.e. Blue), that won’t degrade faster than a US-president's approval-rating?

The Market Potential

In a recent report, Yole Development made some bold predictions for future of Micro-LED disruption to the entire display industry.

Making Colorful, full of Color

One of the key challenges (and there are a few others) facing Micro-LED, is an old one, that dates back to the first use of LED's for LCD backlighting: How to get consistent color ? Is that with separate Red-Green-Blue LED junctions? Phosphor-converted Blue?

Firstly, let's step back and segregate into two distinct types of applications, of Micro-LED’s in Displays:
  • Type 1: White Micro-LED array, as a lighting product or as the thin (white) backlight-array sheet behind an LCD. Applications: (former) Lighting across any surface e.g. clothing. (latter) Enabling ultra-high-dynamic range (HDR), ultra-high contrast ratio’s (CR) and viewing angle e.g. Ultra-HD LCD TV’s and Portable LCD devices.
  • Type 2: An R-G-B Micro-LED array, as the pixelated surface of a display. Applications: Revolutionary new type of display disrupting both OLED and LCD. In AR, TV’s, Automotive, Smartphones and more. Again with realism, contrast and ultra-high dynamic range, far exceeding that of the best OLED.

A sneak peak into the future: Sony CrystalLED display in TV form

Quick Recap of LED's in Displays

There are good reasons modern LCD’s evolved to use White-LED's for the backlight, instead of R-G-B LED's or CCFL (Cold Cathode Fluorescent Light). Since CCFL's contained Mercury (a toxic heavy metal), they were quickly replaced with semiconductor solid-state backlights after the Blue LED was commercialized. But having 3 very different semiconductor junctions for each color component is problematic: each LED color requires a different semiconductor mix (e.g. Blue: InGaN vs Red: GaAsP/AlGaInP vs Green: AlGaP), and separate wafer construction process (i.e. >3x complexity). Moreover, each junction type has different electrical behaviours, differential efficiency, differential aging and differential performance even across one wafer. 
Furthermore, LED band gaps are extremely difficult to tune for custom color emission wavelengths, and Green has a very wide FWHM, limiting achievable color gamut of this path to less than 85% of Rec.2020 at best. Moreover, Display products based on this approach requires constant re-calibration throughout the lifetime (e.g. HP DreamColor). Outdoor RGB LED signage installations come with a maintenance contracts to manage the constant re-alignment needed. This is also why we will continue to see monochromatic Micro-LED prototypes first - realizing color on LED displays is hard. 

The 4K 6000ppi Micro-LED Micro-Display from Vue Real

By comparison, using one consistent type of high-efficiency junction (e.g. the Blue InGaN) + Yellow Phosphor to form a “dirty white” LED for LCD backlight, was the more robust, practical, cost-effective solution - although it severely limited the achievable color gamut, and impacted power efficiency, it’s helped enabled the proliferation of cheap LED LCD TV’s, 
which are now 71% of all displays made in 2016 by area (DSSC), and 40% by number (of all panels >9") at 264mil units/yr (IHS-Markit Display Panel Tracker 2017).

High-level path forward

Micro-LED’s arrays (of either type 1 or 2) would ideally emit all the colors, from the same Blue junction; from die cut from the same wafer, and binned for consistent performance. Such a Micro-LED array could employ 3 Blue-junction sub-pixels on one Micro-LED die, for each pixel; and use color conversion to turn into Blue, Green and Red (Note: if using a Blue junction of the correct center wavelength, then no conversion of the Blue is needed). This is a fairly attractive path, supporting a single consistent junction type at the wafer construction level, with consistent electrical performance, reducing the pick-place assembly task to picking up a single chip, of one type, for placement on the backplane.

Ye'old Phosphor Approach

Either: 1 Blue-junction with Yag (for Type-1), or (for Type-2) 3 Blue-junctions with Blue-converting phosphors for the Red (e.g. KSF) and Green (e.g. a-SiAlON). Key problem with this approach, it’s like "shoving a square peg, into a pin-hole”, the phosphor particles are: 8~20um, while the Micro-LED junction are 3um. And just like "High-Gamut" 2-phosphor LED’s, used in Wide-Gamut LCD backlighting: differential aging of the Red and Green phosphors needs to be carefully addressed, or else a visible color-shift can occur after as little as 1KHrs use. Furthermore, since Micro-LED’s rely on PWM to control brightness, the extremely long 10ms fluorescent decay of KSF Red Phosphor (~1000x that of a-SiAlON, >10,000x that of YAG) causes a Red-shift to occur. As many recent LCD applications are learning: this approach has it's challenges.

Ye'old CF Approach

Micro-LED’s could borrow the same trick as WOLED (e.g. as used by LG OLED products), by using a White-LED plus R/G/B Color Filter. Or Blue & Yellow LED’s plus G/R Color Filter (to make BGYR pixels). But adding a high transmission-loss absorptive filter can mean throwing away 2/3 of the energy in wasted heat, and reducing the achievable color gamut, while also adding layer-complexity; and if you’re advocating Micro-LED’s to displace OLED in power-sensitive SmartPhone applications, where's the fun in that?

Ye'old QD Approach

Depositing Quantum Dots in a Blank/Red/Green patterned color-conversion layer over Blue junction has been proposed. Assuming one mitigates the QD susceptibility to humidity & oxidization through quick encapsulating within an extra barrier film / glass layer; QD particles have the advantage of being small (<5 nm in diameter), but have always been challenged supporting high-flux environments. Since they radiate equally efficiently in all directions, and tend to self-absorb photons, thus making efficient light extraction extremely difficult when tightly packed (which is like only talking to yourself at a party... even less funny than cracking a joke about "absorptive filters"). All, except high-yield QD’s (i.e. heavy metal laden), have had relatively poor external quantum efficiency compared to phosphors. But given the choice of either packing QD’s in higher density patterning and taking the radiant efficiency loss, or else leaking Blue; worse than Amazon’s disastrous experiment with the QD approach (Google for: “Kindle Fire HDX Blue haze”).

Beyond the basic photonics & physics issues, the standard QD challenges remain: the dependence on heavy-metals for high color-gamut spectral-purity & higher-quantum-yield efficiency, limited operating temperature range, thickness and cost of the barrier film to protect from humidity and oxidization sensitivity, intolerance to high-flux light sources… etc. There may be some hope on the horizon with new Cd-Free CIGS materials, but getting it right on all fronts still seems even further away than cheaper+larger+longer-lifetime OLED's.

So, what would be the ideal solution?
To highlight the key points so far:
  1. Sticking with the core Micro-LED theme: long-lifetime robust inorganic materials. Compatible with Semiconductor wafer chip production. Zero heavy metals
  2. Smaller (sub-micron) color conversion particles, easily distributed evenly, within the small feature area with efficient light conversion and extraction (e.g. EQE >70%)
  3. Working over a wide operating temperature range. Support junction temps of 100~150C at max brightness
  4. Efficient conversion of 447~450nm Blue, in high flux (e.g. 100mW/cm2 ~ 2W/cm2) environment, with a wide luminosity range up to 1,000,000 nits.
  5. Support patterned area of 3 sub-pixel junctions, and converting into Blue, Red, Green within a thin layer
  6. Support tunable, narrow (FWHM) emission spectral peaks of 35nm, 20nm or smaller

At SID DisplayWeek 2017, PixelDisplay publicly demonstrated its first Vivid Color™ technology proof-of-concept, with ultra-narrow 10nm FWHM R/G/B primaries (twice the color purity of the best Quantum Dots), using inorganic, zero-heavy metal, LED-based design - see the SID WhitePaper: 67-4, for details on "Vivid Color LED for Ultra-Wide Gamut LCD" (

We're excited by the potential that Micro-LED’s hold for the future of Displays, and eager to make it happen sooner, more prolifically. With Vivid Color™ (Type 1) Micro-LED backlit LCD applications could achieve 97.8% Rec. 2020. And in the SID presentation, PixelDisplay revealed how >120% Rec.2020 Display is achievable using Vivid Color technology, making (Type 2) Micro-LED displays even more exciting. And what would be a more remarkable accomplishment for a new display technology, than realizing light, in all its colors ?

Rethink what's possible.

For more information contact:

(c) 2017, PixelDisplay, Inc. All Rights Reserved