Inkjet printing for microLED quantum dots

Inkjet printing for microLED quantum dots

Interviews |
By Nick Flaherty

3D printing can address the issues with microLED displays for AR headsets, Patrick Galliker of Scrona tells Nick Flaherty

The Swiss inkjet printing pioneer, a spin out of ETH Zurich, is scaling up its electrostatic MEMS-based 3D print head technology, with 48 and 128 nozzle versions planned for next year. These will be key for new ways to make microLED displays with quantum dots.  

“Last year we closed the Series A round to develop the inkjet print head. MEMS silicon has now taken back the market from laser printers and we want to take this to functional printing particularly in areas where additive manufacturing has huge value,” CEO Galliker tells eeNews Europe.

“What we were missing was a way to individually control the nozzles. Now we have the first print heads with digitality that can be individually controlled. Even if you can control the individual nozzles they still need extremely high voltages in the kV range with thousands of nozzles, so what you really need is lower voltages.”

“We are doing that with EHD using conventional inkjet drivers that operate at voltages below 100V,” he said. “This opens the whole design to thousands of nozzles to unlock production economics and one of the applications we are working on is microLEDs.

He cites the Apple Vision Pro, where there are issues with the microLED displays, as a key opportunity.

3D printing Quantum Dots

3D printing Quantum Dots

“The big benefit of microLED is you can create extremely small pixels (above) with a lot of light at a low energy cost. The problem is there are production issues and the main issue is doing colour conversion. There are a few methodologies for colour microLEDs, and the most tedious is mask transfer, with pixels built on different wafers and transferred to the display substrate. Many companies have been concentrating on that,” he said.

“Another strategy is to take a blue wafer and take a single block. What we come into play is to put a material on top of the blue LED as a filter. Rather than using a phosphor, instead we can print quantum dots.” 

This involves spin coating a super black resist to stop the microLEDs talking to each other. These are 10um deep holes with 3um diameter. “We print the quantum dots into the holes,” said Galliker.

“We need to scale the print heads further,” he said. “To make it compatible we need a lot of print heads each with 1000 nozzles. You can position the nozzles in a custom print head where the positioning of the nozzles aligns with the pixel array.”

“Our protypes have 10 nozzles and we would require print heads with 1000 nozzles, that’s our target. Our development cycles have become shorter and shorter. Going from 10 to 1000 nozzles with MEMS and conventional inkjet drivers is not so much of an issue. Getting to that step has taken 10 years and the next steps can be faster to be compatible with production,” he said. “Depending on a wafer size, we would have 10 heads in parallel, and each will deposit droplets at over 100kHz,” he said.

No thermal load

A key advantage is that there is no thermal load on material during patterning which allows materials such as perovskites. But there are other considerations for high speed, high volume production.

“You have to consider evaporation for high throughput – you can heat the substrate to promote the evaporation and we have an active solvent removal system so it does not condense on the print head as that can be a problem.  We can mix whatever we want into the solution, it doesn’t have to be just perovskites. 

Scrona has been working with Avontama in Zurich for green emitting perovskite materials as well as other materials. “We have also been collaborating with other quantum dots materials such as indium phosphide and cadmium sulfide,” said Galliker.

“We can also work with materials using resins that have more thermal stability which for microLEDs is a huge advantage,” he said. “The microLEDs and the quantum dot is relatively thick so you have strong aspect ratios so the light ends up in the black matrix. Instead we can print metals such as silver or aluminium as walls as an alternative to the black matrix to reflect the light back into the well.”  

The placement precision is better than 1um with an array, down to 100nm accuracy, and the AR displays are typically on a 3um pitch.

The current second generation print head has ten nozzles, while a third generation system in Q1 2024 will have 48 nozzle, 12 each with four different inks. This will scale up to 128 nozzles and from there to larger arrays.

“For the fourth generation we are still looking into the driver technologies, either our own or off the shelf, as the jump from 128 to 1000 raises questions. We expect this to be finalised in 2024,” he said.

3D printing interconnect 

Another issue is the attachment of the microLED to driver chip. “Creating these connections is not so simple and we can create bumps on each side of the wafer to create the electrical connection to the drivers,” he said.

“We also have compatibility with high viscosity materials, what is more important is to improve resolution with high precision and we are seeing interest in advanced packaging, back end, printing protection and encapsulation, underfill materials etc.”

“We have some big collaborations with major customers and we are setting those up at the moment. There is traction in packaging and microLED, some in displays and some in PCBs creating metal connections,” he said.








If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News


Linked Articles