Micro-LEDs – Hard to produce but Harder to Test
Over the last few years we have noted the difficulties in producing micro-LED displays. Based on structures that are only a few microns in diameter, Micro-LED displays are both complex and costly, yet represent one avenue for the display space that could become the de facto standard a few years down the road. We have spent much time trying to understand the techniques used to transfer Micro-LEDs from wafers to their final resting places in displays, but a portion of the difficulties involved in tha task of moving millions of such small devices from one location to another rest in the fact that not all of the Micro-LEDs on each wafer meet specifications, or in the extreme, might not work at all.
The only way such transfer systems can avoid moving lower quality or defective Micro-LEDs is to test each die while it is still on the wafer. Creating a ‘wafer map’ of a 6” die for Mini-LEDs (200um) would entail examining over 100,000 die, and those LEDs are an order of magnitude larger than the target size for high resolution Micro-LED displays, so the testing has to be both rapid and accurate as any defective die that is transferred to the final substrate must be repaired at a cost far exceeding the cost to produce. But testing does not stop there as each Micro-LED must also be classified as to a number of optical characteristics, particularly color and brightness, as mixing Micro-LEDs with different color characteristics or brightness will result in a display with uneven characteristics, which would be unacceptable to consumers.
In order to test such almost microscopic devices, very small ‘probes’ that contact each Micro-LED’s connection pads to activate it must be moved across the die, and the resulting light emission must be evaluated by a number of sensors that characterize the metrics of each die. Such a ‘wafer map’ is shown below, which shows how each die varies as to a number of characteristics, particularly wavelength, which would refer to ‘color’. LED test systems can use either PL (Photoluminescence) or EL (Electrical) stimulation to stimulate the Micro-LEDs to produce light , but an Israeli company, InZiv (pvt), who has just raised $10m from BlueRed Partners (pvt), a Singapore-based VC and OurCrowd (pvt) a VC platform in Israel that allows individuals to access venture investments. The attraction of InZiv is their tool, Omnipix 2.0, which combines both PL and EL testing processes and is able to test dies down to 1um, giving very accurate data on each die. Once that data is fed to a transfer device, only Micro-LEDs that meet specifications will be transferred, avoiding costly and time consuming repairs.
All in, while practical Micro-LED displays are still a few years away, considerable R&D and progress is being made toward solving some of the more complex production issues, and there is little hesitation toward capital investments in companies that look like they might have an interim solution to such problems. Last week South Korean based QMC (pvt) released a Mini-LED transfer tool that is said to have 30% better performance that Chinese competitors, while Kulicke & Soffa (KLIC) has dominated Apple’s (AAPL) Mini-LED business with its Katalyst and Luminex transfer/bonding tools, which are progressing from die transfer to multitask tools that would simplify the many steps required for Mini-LED and Micro-LED production.
https://youtu.be/zSY_qxsr5xI
The only way such transfer systems can avoid moving lower quality or defective Micro-LEDs is to test each die while it is still on the wafer. Creating a ‘wafer map’ of a 6” die for Mini-LEDs (200um) would entail examining over 100,000 die, and those LEDs are an order of magnitude larger than the target size for high resolution Micro-LED displays, so the testing has to be both rapid and accurate as any defective die that is transferred to the final substrate must be repaired at a cost far exceeding the cost to produce. But testing does not stop there as each Micro-LED must also be classified as to a number of optical characteristics, particularly color and brightness, as mixing Micro-LEDs with different color characteristics or brightness will result in a display with uneven characteristics, which would be unacceptable to consumers.
In order to test such almost microscopic devices, very small ‘probes’ that contact each Micro-LED’s connection pads to activate it must be moved across the die, and the resulting light emission must be evaluated by a number of sensors that characterize the metrics of each die. Such a ‘wafer map’ is shown below, which shows how each die varies as to a number of characteristics, particularly wavelength, which would refer to ‘color’. LED test systems can use either PL (Photoluminescence) or EL (Electrical) stimulation to stimulate the Micro-LEDs to produce light , but an Israeli company, InZiv (pvt), who has just raised $10m from BlueRed Partners (pvt), a Singapore-based VC and OurCrowd (pvt) a VC platform in Israel that allows individuals to access venture investments. The attraction of InZiv is their tool, Omnipix 2.0, which combines both PL and EL testing processes and is able to test dies down to 1um, giving very accurate data on each die. Once that data is fed to a transfer device, only Micro-LEDs that meet specifications will be transferred, avoiding costly and time consuming repairs.
All in, while practical Micro-LED displays are still a few years away, considerable R&D and progress is being made toward solving some of the more complex production issues, and there is little hesitation toward capital investments in companies that look like they might have an interim solution to such problems. Last week South Korean based QMC (pvt) released a Mini-LED transfer tool that is said to have 30% better performance that Chinese competitors, while Kulicke & Soffa (KLIC) has dominated Apple’s (AAPL) Mini-LED business with its Katalyst and Luminex transfer/bonding tools, which are progressing from die transfer to multitask tools that would simplify the many steps required for Mini-LED and Micro-LED production.
https://youtu.be/zSY_qxsr5xI
- Full wafer map of (350 µm) 2 LED die performance of a wafer - Source: Full wafer map of (350 µm) 2 LED die performance of a wafer C. Wetzel and T. Detchprohm Future Chips Constellation and Department of Physics Rensselaer Polytechnic Institute
RSS Feed