Narrow pixel pitch direct view LED displays for signage applications and high-performance LCD backlights relying on an increasingly high number of increasingly small LED chips (see our CES 2021 review and our initial thoughts of Apple’s miniLED iPad Pro backlight architecture) are driving rapid adoption of miniLEDs. The manufacture and assembly of high quantities of LED chips significantly smaller than what the industry has been used to over the last 20 years brings a new set of challenges in terms of handling, throughput, accuracy, and cost of ownership.
At the same time, microLED self-emissive displays that rely on even smaller chips, sometimes down to 2-3 µm in size, are progressing on all fronts. Investments are increasing, alliances and partnerships are multiplying, dozens of large corporations and startups have presented prototypes, and the first commercial products even hit the market in 2021. With prospects improving, equipment makers are increasingly interested in developing and offering microLED-dedicated tools. Although hindered by a lack of standard manufacturing flow, some are developing one-stop solutions, including mass transfer, inspection, repair and testing. As discussed extensively in our MicroLED Displays – Market, Industry and Technology Trends 2021 report, the availability of robust, manufacturing-ready, off-the-shelf tools from reputable equipment makers is a critical milestone in the young history of microLEDs: LCD or OLED didn’t take off until high volume manufacturing equipment became available.
ASMPT, a leading advanced packaging equipment provider, has been quick to react to the fast-growing miniLED market and the emerging microLED opportunity. The company is already offering a collection of tools to address the needs of both miniLED and microLED display makers.
Yole Développement analysts Eric Virey and Zine Bouhamri talked with Jonathan Ku, ASMPT’s Senior Director of business development, about the company’s strategy and prospects for miniLED and microLED.
Yole Développement (YD): First, could you introduce yourself and ASMPT to our readers?
Jonathan Ku (JK): My name is Jonathan Ku, ASMPT’s Senior Director of Business Development in Optoelectronics, with an emphasis on miniLED and microLED packaging solution development for Advanced Displays. I have worked in ASMPT for over 20 years, mainly focusing on the Opto/LED industry.
YD: There are no commonly accepted definitions of miniLED and microLED, which often creates some confusion. How does ASMPT define and differentiate the two technologies?
JK: Yes, we understand the confusion as we sometimes see promotions of microLED TV/ Displays that are actually assembled with miniLED products. ASMPT started developing miniLED & microLED packaging solutions in 2016, and at the time, we differentiated these two technologies by die size. (i.e., miniLED: 50um -200um; microLED: Less than 50um). Lately, we have altered our definition to be thatmicroLED has no carrier base, like sapphire.
YD: What do you see as the key challenges faced by miniLED device manufacturers?
JK: Cost! Whether related to yield, assembly productivity, the ability to bond smaller and hence less expensive die, everyone is driving for a lower cost. For all of these technical hurdles, ASMPT offers critical process solutions to support its customers. ASMPT’s products can sort, mix, assemble, and, if necessary, repair with the lowest cost of ownership.
YD: Are there differences in the requirements between RGB direct view LED and LCD backlight applications?
JK: There are big differences. LCD backlight is not a miniLED technology, actually. The die size is normally over 200um. And also, the unit pitch is large (Range 2mm- 10mm) so conventional die bonders can handle it with lower technical criteria except for the large panel size handling requirement. RGB direct view LED bonding pitch is much finer than backlight, and the die size is much smaller as well. One to two years ago, 5x9mil die size was still the mainstream, though this changed to 4x8mil last year. This year, most of our customers tested 3x5mil die bonding as they can enjoy the benefit of unit die cost-saving and prepare for fine pitch display (Less than P0.6). Several packaging houses even started 2x4mil die bonding evaluation to produce less than P0.4 ultra-fine pitch displays. However, substrate cost is the main hurdle for 3x5mil and 2x4mil high volume production. This requires substrate supplier breakthroughs to reduce the fine pitch substrate manufacturing cost.
YD: Could you introduce your miniLED offering?
JK: For RGB direct view LED die attach, we developed the Vortex platform. It provides a high speed & high precision pick & place COB solution for large LED display panels with small die handling capability, down to 50um x 100um LED chips. Thanks to its XYθ correction capability, VORTEX achieves REAL precision die placement with no discount on UPH for all usable die on the wafer. With its powerful inspection capabilities, VORTEX offers a high yield solution with post-bond inspection to guarantee the bonding quality. It is especially suited for ultra-fine pitch RGB miniLED displays down to P0.4.
For LCD backlight die attach applications, our AD420XL bonder is a high speed, high precision pick & place, miniLED COB solution for both large-sized LCD BLU for local dimming and ultra-fine pitch LED displays. It has small-die handling capability, down to 75um x 125 LEDs. Once again, with XYθ correction capability, it achieves REAL precision die placement with no discount on UPH for all usable die on the wafer. Targeted applications are large size LCD BLU for local dimming, PCB, glass substrate, etc.
We also offer a miniLED re-work Solution, the RW300. It is a one-pass re-work solution for RGB LED displays: it removes the defective die and rebonds a repair die. It is designed for fine and ultra-fine pitch RGB displays down to P0.5, and can handle multiple wafers and die size down to 75um x 125um. It is a clean process using a high-precision laser and a patented bonding technology with excellent tilt control. The tool can handle panel substrates up to 200×250 mm in size.
Finally, for customers requiring the ultimate in RGB display quality, where mass reflow cannot achieve the acceptable quality, we have a mass bonding platform, the AD300Mini, which delivers placement accuracy of ± 2 µm and angle accuracy of ± 0.005°. The Vortex rapidly reallocates die from sorted wafers onto intermediate carriers that the AD300Mini then bonds in-situ.
YD: MiniLED device makers seem to hesitate between different strategies. Some sacrifice capabilities and throughput and choose entry-level die bonders. Others seem keen on adopting more expensive, cutting-edge processes and equipment. Is this the result of the individual company’s strategies, or does it just reflect different application requirements?
JK: It really depends on both the different company’s strategies and applications. For example, in LCD backlighting (BLU) applications, some companies buy equipment for chip on board (COB) bonding, but some companies still use the traditional package on board (POB) technology. For RGB direct view displays, it typically depends on the pixel pitch. If the customer targets ultra-fine pitch displays(P0.4-P0.6) using smaller LED dies of 75um x 125um or even 50um x 100um, it becomes essential to buy high precision(for yield) and high speed(for productivity) machines. Some competitors have, to little avail,
tried to slow down their existing conventional die bonders to bond miniLEDs. In general, users need to sacrifice bonding yield and spend extra costs & resources for the re-work process.
Customers who demand high quality and good bonding co-planarity control displays have purchased ASMPT’s die re-allocation and mass bonding solution to achieve it. As mentioned above, for these applications, Vortex can be configured as a die re-allocation platform to facilitate mass transfer using the AD300Mini. This is a perfect fit for those P0.4 – P.6 ultra-fine pitch display applications. Finally, those customers who believe that miniLED is simply a stepping stone to microLED have purchased the Vortex with the AD300Pro, a solution that today can bond miniLED for the customer but is engineered and capable of in-situ bonding of the smallest microLED.
YD: Can you comment on your miniLED customer portfolio (applications, company types, location, etc?).
JK: For both fine pitch RGB direct view LED displays and LCD backlights, our customers comprise of panel makers as well as LED packaging houses and direct view key display makers in China, Korea, Taiwan, Japan, Europe, and the US.
YD: Are there additional unsolved challenges for miniLED that ASMPT is working to solve?
JK: Sure, it never stops. We are continuously working on improving productivity while increasing bonding accuracy. This will help our customers improve their yield for small die and ultra-fine pitch.
YD: MicroLED self-emissive displays are very promising but still in the early stages of development. When did ASMPT decide to jump into this market?
JK: ASMPT has worked with one microLED packaging pioneer in Taiwan since 2017 and offered the 1st mass pick and bond solution to the world.
YD: Could you introduce your offering in the microLED space? What are the major characteristics in terms of stamp size, cycle time, die size, bonding, etc.
JK: ASMPT was the early pioneer of microLED mass pick, mass transfer, and bond. As such, we released a production platform several years ago. The latest model is the AD300Pro, and it has helped us capture over a 90% market share with current pilot run players.
The most common transfer stamp size today is 37x37mm, but some companies are running 49x49mm. Bonding yield is greatly influenced by the co-planarity of the stamp to the substrate. Typical die size ranges from 5- 50um. Cycle time is highly material and process dependent, but our alignment time is only a few seconds. We have achieved hundreds of thousands of die bondings in one shot with 100% yield to affect hundreds of millions of UPH.
YD: Dozens of transfer process concepts have been developed by startups, display makers, research organizations, etc. What was the genesis of the AD300Pro? Did you develop the mass transfer process internally or collaborate with external partners? A few months ago, VueReal announced that the AD300Pro can be retrofitted with their mass transfer technology. Are you collaborating with other process developers?
JK: Yes, there are dozens of transfer process concepts in the market but most are in lab testing mode. The AD300Pro is the most mature mass transfer and bonding platform in current production mode. ASMPT develop mass transfer and bonding process internally. In addition, we worked closely with different material suppliers in the market to achieve our goals. We always discuss and are open to suggestions from other process developers about any collaboration opportunities.
YD: Stamp-based and laser-based are probably the most advanced and popular mass transfer concepts for microLED currently (with self-assembly a close third). What do you see as the pros and cons for each technology? Is ASMPT interested in laser processes for microLED?
JK: Stamp-based is still the most popular and mature mass transfer concept in the market. Laser-based is a feasible transfer process for microLED and the breakthrough needed is how to handle a few microns thick microLED epi-layer transfer. ASMPT is developing a unique laser process for microLED transfer, and we will release more to the market once more testing data is consolidated.
YD: Overall, remarkable progress has been made on microLED transfer. For most of the 30+ companies we interviewed when preparing the MicroLED 2021, we noticed a significant change compared to previous years: most no longer seem to consider mass transfer as a fundamental roadblock. In your opinion, what else is needed to further improve the processes and equipment and enable high-volume manufacturing of consumer microLED products?
JK: Yes, I agree that mass transfer is no longer a fundamental roadblock as most key players utilized the AD300Pro to achieve this. The key to enabling high-volume manufacturing of consumer microLED products should be the front and back-end processes, such as microLED wafer yield, sorting, final assembly, board re-work, etc.
YD: Do you ultimately expect to see a cross-over of microLED into miniLED technologies? Could mass transfer concepts such as those implemented in the AD300Pro also be useful for some miniLED applications?
JK: Definitely, Yes! That’s why ASM developed the OCEAN line (MiniLED feeder + AD300Pro Mass bonding) 2 years ago to meet this purpose and provide good bonding co-planarity control of the whole display panel.
YD: What is next for ASMPT in the miniLED and microLED fields? What should we be looking for in the next few months and years?
JK: Both miniLED and microLED use high-density matrix bonding processes that require a high speed and flexible packaging process to drive low cost and more efficient bonding. ASM is working in this direction and is developing innovative machines to support high-volume mini & microLED production. We are also making efforts to the front end and re-work process as a total solution to expedite mini and microLED products’ commercialization.
ASM also offers a factory automation solution, and some miniLED customers have utilized this as they have started high-volume production of RGB ultra-fine pitch displays (<P0.6) in the past year. MicroLED packaging line automation will be the focus in the next few months.
Mr. Ku Chi Ho, Jonathan is a Senior Director of business development of ASM Opto & Display Business Unit in the Semiconductor Solutions Segment.
He graduated from the City University of Hong Kong in 1997 with a Degree in Manufacturing Engineering. Mr. Ku joined ASM in 2000 as a Sales Engineer and was promoted to the Senior Director position. For the past 21 years that he has worked for Sales and Marketing role in Opto and material business unit. His wide exposure to the Opto electronics supply chain has given him ample opportunities to develop extensive customer contacts, a good understanding of market needs and an ability to provide solutions that satisfy customers’ current and future requirements. Mr. Ku has worked with the team on Mini and MicroLED business development since 2016. ASM offer critical process solution for Mini and MicroLED packaging in Advance Display Industry and expedite high volume manufacturing and commercialization.
Eric Virey, PhD, serves as a Principal Display Market and Technologies Analyst within the Photonics, Sensing & Display division at Yole Développement (Yole).
Eric has spoken in more than 50 industry conferences over the last 10 years and has been interviewed or quoted in multiple media including: The Wall Street Journal, CNN, Fox News, CNBC, Bloomberg, Financial Review, Forbes, Technology Review, etc.
Prior to joining Yole, Eric held R&D, engineering, manufacturing and marketing positions with Fortune 500 Company Saint-Gobain in France and the United States. Eric received a PhD in Optoelectronics from the National Polytechnic Institute of Grenoble. He is based in Portland, OR.
Zine Bouhamri, PhD. Team Lead Analyst, Imaging & Display Activities at Yole Développement (Yole).
Zine is managing the expansion of the technical expertise and the market know-how of the company.
In addition, he actively assists and supports the development of dedicated imaging collection of market & technology reports and monitor as well as custom consulting projects.
Prior to Yole, Zine oversaw numerous R&D programs at Aledia. During more than three years, he developed strong technical expertise as well as a detailed understanding of the display industry. He is author and co-author of several papers and patents.
Zine Bouhamri holds an Electronics Engineering Degree from the National Polytechnic Institute of Grenoble (FR), one from the Politecnico di Torino (IT), and a Ph.D. in RF & Optoelectronics from Grenoble University (FR).
MicroLED Displays – Market, Industry and Technology Trends 2020
Strong momentum for MicroLED with progress on all fronts. Cost is the biggest challenge, but Apple and Samsung are carving paths toward the consumer.