- Compared with the silicon wafer, thin-film-transistor (TFT) circuitry is more cost-effective to make bigger fingerprint-on-display (FoD) sensors.
- Bigger sensors can support multi-finger detection, which has a higher security level, and the “blind unlock” feature, which is convenient for users to unlock with a larger recognition area and without having to look for the sensor.
- Unlike the TFT-based ultrasonic solution patented by Qualcomm, panel makers would prefer optical imaging TFT solutions, and they are more feasible on both rigid and flexible types.
- The technical approach of under-cell FoD has some advantages for panel makers:
- First, its approach does not notably influence the yield rate of manufacturing AMOLED displays as in-cell FoD.
- Second, panel makers have enough TFT capacity for FoD sensor productions.
- Third, the FoD integration or assembly can increase AMOLED display average selling prices (ASPs).
Solution makers improved silicon-based optical imaging CMOS image sensor (CIS) solutions significantly in 2019, for example, mature performance like signal-to-noise ratio (SNR), affordable solutions, such as optical lens type to leverage the compact camera module (CCM) supply chain, and module thickness, such as microlens type. Thicker optical lens type was cost-effective to increase adoptions, and thinner microlens type was preferred for 5G smartphones or high-end products.
Qualcomm’s 3D Sonic solution is TFT-based and only suitable for flexible AMOLED displays, but Century’s LTPS TFT circuitry, which is used to drive the ultrasonic sensor materials, have demonstrated TFT advantages, such as thin module and big sensor area. Qualcomm proposed an option of 20x30mm sensor area—named as 3D Sonic Max—at the end of 2019; it is much bigger than CIS-based solutions, which are 6x9mm or smaller. GIS and O-film are Qualcomm’s module makers.
OXI is another TFT-based solution maker, but its technology is optical imaging instead of ultrasound. Isorg’s sample demonstration in CES 2020 also adopted optical imaging TFT sensors. OXI’s mass production schedule was postponed to the end of 2020. For Qualcomm 3D Sonic, its optical imaging TFT-based FoD sensor is laminated beneath AMOLED displays but feasible for both rigid and flexible types. When TFT-based solutions mature, then bigger sensor area will be the critical focus in 2021. Bigger sensors will be in favor of TFT-based solutions and the supply chains. Some AMOLED panel makers particularly, such as Samsung Display (SDC), BOE, and Tianma, have started to allocate the development resources.
Bigger TFT-based FoD sensors are required
It is difficult to increase the sensor area with CIS-based solutions because of the wafer process and cost, which create an opportunity for TFT-based solutions. CIS chips are commonly produced at 8-inch wafer foundries. While the die size is bigger, the economic cutting rate and yield rate are lower. In addition, the silicon wafer is round, so the cutting efficiency for rectangular CIS chips is also lower. Goodix proposed a tilting-chip approach to make a bigger CIS-based FoD sensor in 2019, but it is not ready yet. The solution uses four CIS chips to detect the fingerprint image, and the imaging algorithm composes separate signals. Despite that, high wafer cost and tight 8-inch capacity will limit its feasibility.
The glass substrate for a TFT circuitry process is bigger. Most panel makers’ production lines have higher generation capacities than Gen 4.5. TFT circuitry process—on the glass or flexible plastic substrate—will be more affordable for bigger sensor sizes. The same optical imaging technology can be applied to a TFT-based sensor design, in which the photodiodes are patterned with TFT circuits, to replace CIS chips. Despite its lower resolution, a-Si TFT-based is qualified for 250–400 pixels per inch (PPI) resolution. Qualcomm uses the LTPS TFT circuitry to drive the piezoelectric materials and generate ultrasonic waves. TFT-based solutions can be bigger and achieve a half-display or full-display area. OXI even proposed an option of 25x45mm, which has yet to be produced. CIS chips are too expensive and not practical for such big area.
Replacing silicon wafer with TFT circuitry
Besides OXI, other solution makers and panel makers are interested in TFT-based optical imaging FoD solutions. Panel makers are already familiar in dealing with optical technology and light path design. In addition, the optical imaging technology does not have any patent issues or violations with the ultrasonic technology by Qualcomm. Samsung’s patent (US20180012069) illustrates the under-cell optical imaging structure, and the optical layer is a pinhole type. OXI proposed a module design of the optical imaging TFT solution.
OXI’s parts include a-Si TFT, with inorganic photodiodes, gate IC to sequence sensor pixels, read-out integrated circuit (IC), with analog-to-digital converter (ADC) and amplifier (AMP), to extract the signals because of the photoelectric effect, power IC (+/- 15 voltage) to drive TFT circuitry, and controller IC for pixel or fingerprint imaging. Increasing sensor area is feasible for more read-out ICs to increase detecting channels such as touch controller and display driver IC. The optical layer for light collimation is made of optical fibers provided by the supplier in Shandong, China. The bunch of optical fibers is sliced to generate thin optical layers, and the function is identical to the microlens of optical imaging CIS solution. The notable disadvantage of OXI’s module is its complicated module design; to simplify the module and assembly, the parts can be integrated into AMOLED displays. Therefore, panel makers are also interested as they have the capabilities to simplify the module.
Considering a potential moiré effect from AMOLED displays, OXI selected a lower sampling resolution of approximately 250–260ppi. The discrete fingerprint chip solution, which is a capacitive type, has 360–508ppi resolution. OXI’s FoD modules still need direct bonding with AMOLED displays because of weaker SNR performance. While SNR improves, air gap and assembly with the middle frame can lower its module cost and the possibility of damaging the display.
Compared with OXI, the major difference for Isorg’s FoD solution is its organic photodiode (OPD) materials. Common photodiode materials are inorganic such as indium gallium arsenide (InGaAs), gallium phosphide (GaP), and gallium arsenide (GaAs), which is a III–V compound semiconductor. These inorganic materials are widely found in camera and surveillance applications with CIS solutions. Inorganic and organic materials have the same photoelectric effect; the light absorption for energy-level jumping generates electrons and holes then creates currents through the electrodes. OPD materials have advantages such as sensor flexibility, low cost, simple coating process, and room temperature for manufacturing. Isorg uses the OPD materials supplied by Sumitomo Chemical. OPD materials are incompatible with oxygen, like AMOLED materials.
TFT-based FoD solutions and panel makers
Solution makers, such as OXI and Isorg, and panel makers, like BOE, Tianma, and SDC, are developing optical imaging TFT solutions to use TFT-based photo-detection circuitry with special optical layers, for light collimation, to replace CIS chips. This technical approach has some advantages for panel makers. First, its under-cell, or under-display, approach does not notably influence the yield rate of manufacturing AMOLED displays as in-cell FoD. Second, panel makers have enough TFT capacity for FoD sensor productions. Third, the FoD integration or assembly can increase AMOLED display ASPs. Compared with IC-based solution makers, like OXI, it seems more appropriate for panel makers to develop TFT-based optical imaging FoD solutions.
All available FoD solutions adopt the under-display approach, although FoD sensors have nothing to do with the displays and are just adhered beneath the displays. This situation is like add-on touch sensors, such as glass-film-film (GFF) and glass-to-glass (GG) structures, in which they are assembled to the displays but not integrated. Because the FoD sensor is far from the fingerprint sensor, which is also blocked by the display, its SNR is more challenging. However, under-display FoD development can be faster without considering the display extensively.
Some panel makers are interested in FoD development, but there are no feasible solutions along with mass production yet. FoD can be a useful feature to create add-on values on the display end, but it seems more technically difficult than embedded touch, such as in-cell or on-cell touch. Except for AUO working on the in-cell FoD (TFT LCD), other panel makers are more interested in the TFT-based solutions for AMOLED displays. Embedded FoD solutions for AMOLED displays will take longer time because these panel makers are suffering from raising the yield rates of displays and on-cell touch, which is directly patterned as YOUM On-Cell Touch AMOLED (Y-OCTA).
Under-cell FoD is like the under-display approach, but it is created for the panel makers’ business models instead of IC-based solution maker. In-cell or on-cell FoD sensor process is integrated with display manufacturing, but under-cell can be done separately. The under-cell sensor is further laminated beneath the finished AMOLED displays. Compared with the TFT-based ultrasonic solution patented by Qualcomm, panel makers would prefer optical imaging TFT solutions, and TFT solutions are more feasible on both rigid and flexible types.