Manufacturing method of OLED (Organic Light Emitting Diode) metal mask plate and mask plate

Abstract

The invention relates to the technical field of semiconductor light emitting diodes, and discloses a manufacturing method of an OLED metal mask plate and the mask plate, and the manufacturing method comprises the steps of substrate pretreatment and patterning, etching and post-processing, stress monitoring and annealing, protective layer deposition, ALD monitoring and adjustment and the like. Through the closed-loop design of MBIL mixed exposure, FEA bionic structure, ALD protection and automatic parameter adjustment, the pain point is accurately solved, the overlay precision is improved to 0.07-0.09 [mu] m, the trepanning precision is + / -0.3-+ / -0.5 [mu] m, and the precision defect is improved; the thermal stress at the temperature of 150 DEG C is controlled to be 11.2-13.0 MPa, the shadow offset is controlled to be 2.1-2.7 microns, and the thermal deformation problem is relieved; the salt spray corrosion is less than or equal to 0.07%, and the bending resistance change rate is less than or equal to 3.8%; parameter fluctuation is automatically adjusted and corrected, stable mass production is guaranteed, and all links cooperate to achieve comprehensive performance breakthrough.

Description

Technical Field

[0001] This invention relates to the field of semiconductor light-emitting diode technology, specifically to photomasks, and more specifically to a method for manufacturing an OLED metal photomask and the photomask itself. Background Technology

[0002] OLED (Organic Light Emitting Diode) display technology has been widely used in the display field due to its advantages such as self-illumination, high contrast, wide viewing angle, and fast response speed. In the manufacturing process of OLED displays, especially flexible OLED displays, the precision, thermal stability, and deformation resistance of the metal mask directly determine the positional accuracy and display effect of the vapor-deposited pixels. Currently, traditional metal mask manufacturing faces three major problems: First, it is difficult to simultaneously achieve precision in both fine and non-fine areas during patterning, with overlay errors easily exceeding 0.3μm, failing to meet the high-resolution requirements of 4K / 8K; second, poor control of the roughness of the aperture sidewalls after etching, and a lack of targeted design of etching solution concentration parameters, resulting in a vapor deposition shadow defect rate exceeding 5%; third, the stress concentration problem under thermal load is prominent, with shadow offsets often exceeding 8μm at 150℃, coupled with low online monitoring frequency, slow feedback response, and a high risk of warpage loss. Therefore, a precise and scientifically designed OLED metal mask manufacturing method is urgently needed to solve the above-mentioned technical pain points. Summary of the Invention

[0003] The purpose of this invention is to provide a method for manufacturing an OLED metal mask and the mask itself, so as to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides a method for manufacturing an OLED metal mask, comprising the following steps:

[0005] S1: An Invar alloy substrate is provided, and patterning is performed using a hybrid exposure mode combining multi-beam interference lithography (MBIL) and step-scan projection exposure. MBIL is used to form patterns in fine areas, while step-scan projection exposure is used to form large-area patterns in non-fine areas, achieving patterning with "high precision in fine areas + high efficiency in non-fine areas," thus resolving the contradiction of "insufficient precision or low efficiency" in traditional single-exposure methods. The Invar alloy substrate is made of Invar alloy with a nickel content of 36% (e.g., UNS K93600 Invar alloy plate produced by Shanghai Jubei Metal Materials Co., Ltd.), with a thickness of 0.08-0.15 mm, flatness ≤2 μm / m, and surface roughness Ra ≤0.01 μm. The substrate undergoes vacuum annealing (800℃, 2h) to eliminate internal stress and ensure that the warpage change during subsequent processing is <0.5 μm / m.

[0006] S2: The patterned substrate is etched using FeCl3-based etching solution, and then the etched substrate is polished for 6-12 seconds using HF-HNO3 mixed solution to obtain an opening with a sidewall roughness Ra≤0.02μm, achieving "low roughness + no over-corrosion" for the patterned opening and reducing the evaporation shadow defect rate.

[0007] The FeCl-based etching solution is prepared using analytical grade FeCl3 (purity ≥ 99.9%) and deionized water (resistivity ≥ 18.2 MΩ·cm) at a concentration of 38-45 wt%. Filtration (using a 0.22 μm filter membrane) is required to remove impurities and avoid etching defects. The HF-HNO3 mixed solution uses electronic grade HF (purity ≥ 99.999%) and HNO3 (purity ≥ 99.99%). The solution is prepared by adding deionized water first, then HNO3, and finally HF to prevent localized overheating. The solution must be used within 24 hours to avoid concentration decay.

[0008] The FeCl3-based etching solution has a concentration of 38-45 wt%, an etching temperature of 48-55℃, and an etching time adjusted according to the substrate thickness and aperture requirements, ranging from 15-25 min. Magnetic stirring (stirring speed 220-280 rpm) is used during etching to ensure uniform etching. 3+ It can undergo redox reactions with Fe and Ni in Invar alloys (2Fe2O3-NiO2 ... 3+ +Fe=3Fe 2+ 2Fe 3+ +Ni=2Fe 2+ +Ni 2+ A concentration of 38-45wt% and a temperature of 48-55℃ allow for precise control of the etching rate, ensuring that the aperture diameter matches the substrate thickness and avoiding under-etching (aperture not fully penetrated) or over-etching (aperture diameter too large). After etching with HF-HNO3 solution, microburrs may remain on the sidewalls of the apertures. HF can selectively dissolve metal oxides and burrs, while HNO3 regulates the etching rate (avoiding excessive etching of the substrate by HF). A short treatment time of 6-12 seconds removes only the surface rough layer, ultimately achieving smooth sidewalls with Ra≤0.02μm, reducing particle scattering during evaporation, and lowering the shadow defect rate.

[0009] S3: In the non-evaporation window area of ​​the etched and polished substrate, a biomimetic mesh stress relief structure with gradient unit size is designed and fabricated. The mesh unit size near the evaporation window area is 15-50μm, and the mesh unit size far from the window area is 140-200μm. This disperses the stress concentration under a 150℃ thermal load and reduces the shadow offset from more than 8μm to less than 3μm. Gradient biomimetic mesh structure (20-50μm near the window area, 140-200μm far away): Stress concentration in the non-evaporation window area is the main source of thermal deformation. The 20-50μm small mesh near the evaporation window area (pixel area) can enhance the local support stiffness and prevent the window area from sinking due to thermal stress. The 140-200μm large mesh far from the window area can release the overall thermal stress and reduce the warping of the mask edge. Through the gradient design of "local support + overall release", the stress distribution uniformity is improved by more than 40%.

[0010] S4: Atomic layer deposition (ALD) technology is used to sequentially deposit AlO and ITO layers on the surface of the substrate with the biomimetic mesh stress relief structure to form a composite protective layer, which reduces the surface resistance to below 20Ω / □ and constructs a composite protective layer with "corrosion resistance + high conductivity + flexibility and compatibility", solving the problems of poor environmental resistance (salt spray corrosion area > 1%) and high surface resistance (> 30Ω / □) of traditional photomasks.

[0011] ALD precursors: TMA (trimethylaluminum) purity ≥99.999%, water content <1ppm; H2O is ultrapure water (resistivity ≥18.2MΩ·cm); InCl4 (indium tetrachloride) purity ≥99.99%, metal impurity content <10ppb; TMG (trimethylgallium) purity ≥99.999%, oxygen content <5ppm. All precursors must be stored in steel cylinders protected by inert gas (N2, purity ≥99.999%) to prevent oxidation.

[0012] AlO layer is deposited using AlD deposition. The AlO layer has extremely high density (porosity <0.1%), which can effectively block corrosive media such as water vapor and salt spray, preventing the alloy substrate from rusting. ITO (indium tin oxide) is a transparent conductive material. A thickness of 110-190nm can achieve a surface resistivity ≤20Ω / □, ensuring uniform charge transfer on the mask surface during evaporation and avoiding trajectory deviation of evaporated particles (such as RGB organic small molecules) caused by charge accumulation. At the same time, the ITO layer is tightly bonded to the AlO layer, making it less prone to detachment during bending, meeting the 1000-bending requirement of flexible OLEDs (resistivity change rate <5%).

[0013] Preferably, before step S1, the substrate is first ultrasonically cleaned sequentially with deionized water and an organic solvent, followed by plasma surface activation treatment. The ultrasonic cleaning parameters for deionized water are: ultrasonic power 350-500W, cleaning time 12-20 min; the ultrasonic cleaning parameters for organic solvent (selected from acetone, ethanol, or isopropanol) are: ultrasonic power 250-400W, cleaning time 15-20 min; the surface activation treatment uses plasma treatment: the plasma gas source is an Ar / O2 mixed gas with a volume ratio of 3:1, a treatment power of 70-90W, and a treatment time of 4-8 min.

[0014] Ultrasonic cleaning with deionized water and organic solvents can remove oil, dust, and organic impurities from the substrate surface. Limiting the ultrasonic power and time avoids damaging the substrate surface with excessive ultrasonic waves. Ar ion bombardment removes residual contaminants from the substrate surface, while O2 introduces hydroxyl groups (-OH) into the substrate surface, increasing the surface energy from 30mN / m to over 60mN / m. This enhances the adhesion between the photoresist and the substrate (from the traditional 5MPa to over 10MPa), reducing photoresist peeling during patterning.

[0015] Preferably, in step S1, the specific process of the hybrid exposure mode is as follows: firstly, multi-beam interference lithography is used to expose the fine area pattern, with an exposure energy density of 140-180 mJ / cm². 2 After completion, switch to the step-scan projection exposure system at 2-5 J / cm. 2 The exposure energy is used to scan and expose non-fine areas, and the pattern overlay accuracy error during the switching between the two modes is less than 0.1μm.

[0016] MBIL exposure fine zone (140-180mJ / cm) 2 MBIL technology uses multi-beam laser interference to form submicron-level periodic fringes, with a flux of 140-180 mJ / cm. 2 The energy density ensures that the photoresist is accurately photosensitive, enabling high-resolution imaging of pixel-level (e.g., 5-20μm) fine-area patterns (overlay accuracy error <0.1μm), meeting the pixel density requirements of 4K / 8K OLED.

[0017] Step-scan projection exposure of non-fine areas (2-5 J / cm) 2 Non-fine areas (such as borders and positioning marks) are large and do not require submicron precision; 2-5 J / cm 2 The energy can quickly complete the exposure of large-area patterns through field-by-field scanning, taking into account production efficiency; the overlay accuracy of the two modes is <0.1μm, ensuring seamless connection between fine and non-fine areas and avoiding positional deviations during subsequent evaporation.

[0018] Preferably, in step S2, the HF-HNO3 mixed solution contains 0.1-0.5 wt% HF, 1%-3 wt% HNO3, and the remainder is deionized water. The micropolishing treatment temperature is controlled at 25-32℃, and ultrasonic assistance (ultrasonic power 120-180W) is used during the treatment to enhance the micropolishing effect. After treatment, the solution is immediately rinsed with deionized water (rinsing time 3-5 min) and dried with nitrogen (nitrogen purity ≥99.999%).

[0019] Preferably, in step S3, the biomimetic mesh stress relief structure is designed through topology optimization by finite element analysis (FEA), and the gradient of its unit size change is verified by Monte Carlo vapor deposition simulation to ensure that its simulated shadow offset under a 150℃ heat load is less than 3μm.

[0020] FEA can simulate the stress distribution under a 150℃ thermal load, ensuring that the structural design matches the actual stress state; Monte Carlo simulation can predict the trajectory of vapor-deposited particles and verify that the shadow offset is <3μm, avoiding "design and reality disconnect" and ensuring the effectiveness of the biomimetic structure.

[0021] Preferably, in step S3, the processing is carried out using laser processing with a laser wavelength of 1064nm, a pulse width of 30-50ns, and a processing power of 8-10W.

[0022] Alternatively, electron beam processing can be used, with an accelerating voltage of 20-30kV and a beam current of 20-30nA.

[0023] The machining accuracy is controlled within ±1μm, and plasma deburring is performed for 1-3 minutes after machining.

[0024] Laser / electron beam processing (accuracy ±1μm): 20-50μm mesh units require high-precision processing. Laser (1064nm wavelength) or electron beam (20-30kV accelerating voltage) can achieve a processing accuracy of ±1μm, avoiding stress concentration points caused by mesh size deviations and ensuring structural stability.

[0025] Preferably, in step S4, the deposition temperature is 220-280℃, and the deposition pressure is 3-8 Torr; the Al2O3 layer is deposited using trimethylaluminum (TMA) and H2O as precursors, and the deposition rate is... / cycle, thickness of 35-75nm, corresponding to 1750-2679 deposition cycles; the ITO layer deposition uses indium tetrachloride (InCl4) and trimethylgallium (TMG) as precursors, with a deposition rate of / cycle, thickness is 110-190nm, corresponding to 2444-6786 deposition cycles; during the deposition process, the precursor pulse time is 0.2-0.5s, and the purge time is 6-10s.

[0026] Preferably, it also includes online monitoring and feedback control steps:

[0027] Before step S4, the warpage of the substrate is monitored using a stress curvature meter. If the warpage is >5μm / m, low-temperature annealing is performed immediately until the warpage is ≤5μm / m.

[0028] The low-temperature annealing process uses a vacuum annealing furnace with an annealing temperature of 200℃ and a time of 30min. The annealing atmosphere is 99.999% pure Ar gas, with a heating rate of 5℃ / min and a cooling rate of 3℃ / min to avoid substrate oxidation during the annealing process.

[0029] In step S4, a white light interferometer and a stress curvature meter are integrated for real-time online monitoring, and the monitoring data is fed back to the atomic layer deposition control system to form a closed-loop control loop. The atomic layer deposition parameters are dynamically adjusted to ensure that the mask warpage is ≤5μm / m.

[0030] Preferably, the online monitoring and control system acquires film thickness data of the AlO layer and ITO layer and substrate stress curvature data at a frequency of 40-80 times / s;

[0031] When the film thickness deviation exceeds 2nm, adjust the deposition cycle;

[0032] When the warpage data deviates from the target value by more than 10%, the atomic layer deposition process parameters are automatically adjusted, including the atomic layer deposition precursor pulse time, purge time, and reaction chamber temperature.

[0033] The white light interferometer has a measurement accuracy of ±0.1 nm and a measurement range of 0-500 nm; the stress curvature meter has a measurement accuracy of ±0.01 μm / m and a measurement range of 0-20 μm / m; the parameter adjustment range of the ALD control system is: precursor pulse time ±0.05 seconds, purge time ±1 second, and reaction chamber temperature ±5℃. The ALD control system and the annealing furnace control system communicate with a Siemens S7-1500 PLC via EtherCAT industrial Ethernet, with a data transmission delay of <1 ms, achieving a coherent response of "stress monitoring - annealing correction - film thickness control".

[0034] On the other hand, the present invention also discloses an OLED metal mask prepared by the above method.

[0035] The beneficial effects of this invention are as follows:

[0036] The gradient mesh structure, designed using biomimetic principles, efficiently releases and uniformly disperses the internal stress of the substrate from its source, fundamentally suppressing macroscopic warpage deformation of the mask. Secondly, an Al2O3 / ITO composite protective layer is deposited in situ across the entire three-dimensional structure surface using ALD technology. The dense Al2O3 layer provides excellent physical and chemical isolation, while the surface ITO layer effectively prevents dust adsorption by dissipating static electricity. The combination of these two elements provides comprehensive protection for the mask. Finally, pre-process micro-polishing and end-to-end online monitoring ensure high precision in pattern transfer and extreme process stability. The resulting metal mask possesses extremely high pattern precision, extremely low internal stress and warpage, ultra-long lifespan, and excellent resistance to environmental interference, perfectly meeting the stringent requirements of next-generation high-end OLED display manufacturing for core equipment.

[0037] Other features and advantages of the present invention will be described in detail in the following detailed description section. Detailed Implementation

[0038] The following provides a detailed description of specific embodiments of the present invention. It should be understood that the specific embodiments described herein are for illustrative and explanatory purposes only and are not intended to limit the scope of the invention.

[0039] It should be noted that all reagents and raw materials used in this invention are commercially available, and the reagents are of analytical grade.

[0040] Example 1

[0041] This embodiment provides a method for manufacturing an OLED metal mask, including the following steps:

[0042] (1) Substrate pretreatment and patterning: A 0.1 mm thick Invar alloy substrate was selected. It was first ultrasonically cleaned with deionized water (400 W, 15 min), then ultrasonically cleaned with acetone (300 W, 18 min), and then subjected to plasma activation treatment (Ar / O2, volume ratio 3:1, 80 W) for 5 min. The fine area was exposed using MBIL (150 mJ / cm²). 2 (10μm×10μm pattern), then use step scan to expose the non-fine areas (3J / cm). 2 Scanning rate 5mm / s, overlay accuracy 0.08μm;

[0043] (2) Etching and post-treatment: The patterned substrate was etched using FeCl3-based etching solution with a concentration of 40wt%. The etching was carried out at 50℃ and 250rpm for 20min (10μm aperture). Immediately after etching, the substrate was polished for 10s with a low concentration of HF-HNO3 mixed solution (HF 0.3wt% + HNO3 2wt% + the remainder deionized water) at 30℃ and 150W. Then, the substrate was rinsed with deionized water for 4min and dried with nitrogen. The sidewall Ra was measured to be 0.015μm.

[0044] (3) Biomimetic structure processing: In the non-evaporation window area of ​​the etched and polished substrate, a biomimetic mesh stress relief structure with a gradient unit size was designed using FEA topology optimization. The gradient of the unit size change was verified by Monte Carlo evaporation simulation. The mesh unit size near the evaporation window area was 25 μm, and the mesh unit size far from the window area was 150 μm, with a gradient of 5 μm / mm. The Monte Carlo simulation showed a shadow offset of 2.5 μm at 150℃. The biomimetic mesh stress relief structure was processed by laser (1064 nm, 30 ns, 10 W), and then deburred by plasma for 2 min, with a processing accuracy of ±0.8 μm.

[0045] (4) Stress monitoring and annealing: The warpage measured by the stress curvature meter was 5.8μm / m>5μm / m. Low temperature annealing was carried out at 200℃ for 30min (Ar gas protection, 5℃ / min temperature increase). After annealing, the warpage was 4.1μm / m≤5μm / m.

[0046] (5) Protective layer deposition: AlO layer and ITO layer are sequentially deposited on the surface of the substrate with the biomimetic mesh stress relief structure using ALD technology to form a composite protective layer. The deposition temperature is 250℃, the pressure is 5 Torr, and the AlO layer (TMA+HO (molar ratio = 2:3) is subjected to a 0.3-second pulse and an 8-second purge. / cycle, 45nm design thickness, 1800 cycles, actual thickness 45.0nm), ITO layer (InCl+TMG (molar ratio = 5:1), 0.4-second pulse, 7-second purge, / cycle, 140nm design thickness, 4000 cycles, actual thickness 140.0nm), surface resistance 15Ω / □;

[0047] (6) ALD monitoring and adjustment: The film thickness data of AlO layer and ITO layer and the substrate stress curvature data were collected at a frequency of 50 times / s: when the AlO layer was deposited to 45nm, the rate decreased, and the TMA pulse time was adjusted by +0.02 seconds; during the deposition of ITO layer, the warpage increased to 5.4μm / m, and the temperature was adjusted by +3℃, and the final warpage was 4.2μm / m.

[0048] Example 2

[0049] (1) Substrate pretreatment and patterning: A 0.12 mm thick Invar alloy substrate was selected. It was first ultrasonically cleaned with deionized water (450 W, 20 min), then ultrasonically cleaned with isopropanol (350 W, 20 min), and then subjected to plasma activation treatment (Ar / O2, volume ratio 3:1, 80 W) for 8 min. The fine area was exposed using MBIL (160 mJ / cm²). 2 (8μm×8μm pattern), then use step scan to expose the non-fine areas (4J / cm). 2 Scanning rate 4 mm / s, overlay accuracy 0.09 μm;

[0050] (2) Etching and post-treatment: The patterned substrate was etched using FeCl3-based etching solution at a concentration of 45wt%, 55℃, and 280rpm for 25min (opening diameter 8μm). Immediately after etching, the substrate was polished for 12s with a low-concentration HF-HNO3 mixed solution (HF 0.5wt% + HNO3 3wt% + the remainder deionized water) at 32℃ and 180W. Then, it was rinsed with deionized water for 5min and dried with nitrogen. The sidewall Ra was measured to be 0.017μm.

[0051] (3) Biomimetic structure processing: In the non-evaporation window area of ​​the etched and polished substrate, a biomimetic mesh stress relief structure with a gradient unit size was designed using FEA topology optimization. The gradient of the unit size change was verified by Monte Carlo evaporation simulation. The mesh unit size near the evaporation window area was 20 μm, and the mesh unit size far from the window area was 180 μm, with a gradient of 4 μm / mm. The Monte Carlo simulation showed a shadow offset of 2.3 μm at 150℃. The biomimetic mesh stress relief structure was processed by electron beam (accelerating voltage 20 kV, beam current 30 nA), and then deburred by plasma for 3 min, with a processing accuracy of ±0.7 μm.

[0052] (4) Stress monitoring and annealing: The warpage measured by the stress curvature meter was 6.2μm / m>5μm / m. Low temperature annealing was carried out at 200℃ for 30min (Ar gas protection, 5℃ / min heating). After annealing, the warpage was 4.7μm / m≤5μm / m.

[0053] (5) Protective layer deposition: AlO layer and ITO layer are sequentially deposited on the surface of the substrate with the biomimetic mesh stress relief structure using ALD technology to form a composite protective layer. The deposition temperature is 280℃, the pressure is 8 Torr, and the AlO layer (TMA+HO (molar ratio = 2:3) is subjected to a 0.4-second pulse and a 10-second purge. / cycle, 75nm design thickness, 2679 cycles, actual thickness 74.9nm), ITO layer (InCl+TMG (molar ratio = 5:1), 0.5-second pulse, 9-second purge, / cycle, 190nm design thickness, 6786 cycles, actual thickness 190.1nm), surface resistance 14Ω / □;

[0054] (6) ALD monitoring and adjustment: The film thickness data of AlO layer and ITO layer and the substrate stress curvature data were collected at a frequency of 60 times / s. When the AlO layer was deposited to 60nm, the pressure fluctuated, and the reaction chamber pressure was adjusted by +1Torr. When the warpage of ITO layer increased to 5.3μm / m during deposition, the purge time was adjusted by +1s, and the final warpage was 4.8μm / m.

[0055] Example 3

[0056] (1) Substrate pretreatment and patterning: A 0.08 mm thick Invar alloy substrate was selected. It was first ultrasonically cleaned with deionized water (500 W, 12 min), then ultrasonically cleaned with ethanol (400 W, 15 min), and then subjected to plasma activation treatment (Ar / O2, volume ratio 3:1, 90 W) for 6 min. The fine area was exposed using MBIL (180 mJ / cm²). 2 (5μm×5μm pattern), then use step scan to expose the non-fine areas (2J / cm). 2 Scanning rate 0.8 mm / s, overlay accuracy 0.07 μm;

[0057] (2) Etching and post-treatment: The patterned substrate was etched using FeCl3-based etching solution at a concentration of 42wt%, 52℃, and 260rpm for 18min (aperture diameter 5μm). Immediately after etching, the substrate was polished for 8s with a low-concentration HF-HNO3 mixed solution (HF 0.2wt% + HNO3 1.5wt% + the remainder deionized water) at 28℃ and 160W. Then, it was rinsed with deionized water for 3min and dried with nitrogen. The sidewall Ra was measured to be 0.014μm.

[0058] (3) Biomimetic structure processing: In the non-evaporation window area of ​​the etched and polished substrate, a biomimetic mesh stress relief structure with a gradient unit size was designed using FEA topology optimization. The gradient of the unit size change was verified by Monte Carlo evaporation simulation. The mesh unit size near the evaporation window area was 15 μm, and the mesh unit size far from the window area was 140 μm, with a gradient of 3 μm / mm. The Monte Carlo simulation showed a shadow offset of 2.1 μm at 150℃. The biomimetic mesh stress relief structure was processed by electron beam (accelerating voltage 30 kV, beam current 20 nA), and then deburred with plasma for 2 min, with a processing accuracy of ±0.5 μm.

[0059] (4) Stress monitoring and annealing: The warpage measured by the stress curvature meter was 5.5μm / m>5μm / m. Low temperature annealing was carried out at 200℃ for 30min (Ar gas protection, 5℃ / min heating). After annealing, the warpage was 3.9μm / m≤5μm / m.

[0060] (5) Protective layer deposition: AlO and ITO layers are sequentially deposited on the surface of the substrate with the biomimetic mesh stress relief structure using ALD technology to form a composite protective layer. The deposition temperature is 220℃, the pressure is 3 Torr, and the AlO layer (TMA + H2O (molar ratio = 2:3) is subjected to a 0.3-second pulse and a 7-second purge. / cycle, 60nm design thickness, 2000 cycles, actual thickness 60.1nm), ITO layer (InCl+TMG (molar ratio = 5:1), 0.4-second pulse, 8-second purge, / cycle, 160nm design thickness, 4000 cycles, actual thickness 159.8nm), surface resistance 15Ω / □;

[0061] (6) ALD monitoring and adjustment: The film thickness data of AlO layer and ITO layer and the substrate stress curvature data were collected at a frequency of 80 times / s: the film thickness deviation of AlO layer deposition was -1.5nm, and the TMA pulse time was adjusted to +0.03 seconds; the stress of ITO layer deposition increased slightly, and the temperature was adjusted to -2℃, and the final warpage was 4.0μm / m.

[0062] Example 4

[0063] (1) Substrate pretreatment and patterning: A 0.09 mm thick Invar alloy substrate was selected. It was first ultrasonically cleaned with deionized water (350 W, 18 min), then ultrasonically cleaned with acetone (250 W, 16 min), and then subjected to plasma activation treatment (Ar / O2, volume ratio 3:1, 70 W) for 4 min. The fine area was exposed using MBIL (140 mJ / cm²). 2 (12μm×12μm pattern), then use step scan to expose the non-fine areas (5J / cm). 2Scanning rate 1.5 mm / s, overlay accuracy 0.09 μm;

[0064] (2) Etching and post-treatment: The patterned substrate was etched using FeCl3-based etching solution. The FeCl3 etching solution concentration was 38wt%, the temperature was 48℃, the stirring speed was 220rpm, and the etching time was 15min (aperture diameter 12μm). Immediately after etching, the substrate was polished for 6s with a low concentration of HF-HNO3 mixed solution (HF 0.1wt% + HNO3 1wt% + the balance deionized water) at 25℃ and 120W. Then it was rinsed with deionized water for 4min and dried with nitrogen. The sidewall Ra was measured to be 0.016μm.

[0065] (3) Biomimetic structure processing: In the non-evaporation window area of ​​the etched and polished substrate, a biomimetic mesh stress relief structure with gradient unit size was designed using FEA topology optimization. The gradient of unit size change was verified by Monte Carlo evaporation simulation. The mesh unit size near the evaporation window area was 50 μm, and the mesh unit size far from the window area was 200 μm, with a gradient of 6 μm / mm. The Monte Carlo simulation showed a shadow offset of 2.7 μm at 150℃. The biomimetic mesh stress relief structure was processed by laser (1064 nm, 50 ns, 8 W), and then deburred by plasma for 1 min, with a processing accuracy of ±1.0 μm.

[0066] (4) Stress monitoring and annealing: The stress curvature meter monitored the warpage as 4.8μm / m < 5μm / m, so no annealing was required, and the protective layer preparation could proceed directly.

[0067] (5) Protective layer deposition: AlO and ITO layers are sequentially deposited on the surface of the substrate with the biomimetic mesh stress relief structure using ALD technology to form a composite protective layer. The deposition temperature is 260℃, the pressure is 6 Torr, and the AlO layer (TMA+HO (molar ratio = 2:3) is subjected to a 0.2-second pulse and a 6-second purge. / cycle, 35nm design thickness, 1750 cycles, actual thickness 34.9nm), ITO layer (InCl+TMG (molar ratio = 5:1), 0.3-second pulse, 6-second purge, / cycle, 110nm design thickness, 2444 cycles, actual thickness 110.2nm), surface resistance 18Ω / □;

[0068] (6) ALD monitoring and adjustment: The film thickness data of AlO layer and ITO layer and the substrate stress curvature data were collected at a frequency of 40 times / s: the deposition rate of AlO layer was stable and no adjustment was required; the warpage of ITO layer was reduced to 3.7μm / m during deposition, no parameter correction was required, and the final warpage was 3.9μm / m.

[0069] Comparative Example 1

[0070] Step (1) uses only step scan exposure, with a uniform energy of 3 J / cm² in both fine and non-fine areas. 2 The scanning rate is 1.2 mm / s, and there is no MBIL exposure step; the steps are the same as in Example 1.

[0071] Comparative Example 2

[0072] Step (5) uses magnetron sputtering to deposit AlO / ITO layers. The AlO layer power is 150W, the deposition rate is 2nm / min, and the thickness is 75nm (time is 37.5min); the ITO layer power is 200W, the deposition rate is 3nm / min, and the thickness is 190nm (time is 63.3min). Step (6) the film thickness is detected offline without real-time parameter correction. The remaining steps are the same as in Example 2.

[0073] Comparative Example 3

[0074] Step (3) The biomimetic structure uses a uniform 50μm grid (without gradients in the neighboring / far away regions) and does not perform FEA thermal stress simulation and Monte Carlo shadow shift prediction; the remaining steps are the same as in Example 3.

[0075] Comparative Example 4

[0076] Step (4) is omitted; the remaining steps are the same as in Example 1.

[0077] Comparative Example 5

[0078] Steps (4) and (6) are omitted; the remaining steps are the same as in Example 2.

[0079] Test Example 1

[0080] The accuracy indicators of the masks prepared in Examples 1-4 and Comparative Examples 1-4 were tested respectively. The test methods and test standards are as follows, and the test results are shown in Table 1.

[0081] 1.1 Pattern overprinting accuracy error

[0082] Test method: Using a laser confocal microscope, 10 evenly distributed overlay alignment marks were selected in the fine area of ​​the mask. The deviation between the actual mark position and the designed position was measured, and the maximum value was taken as the overlay accuracy error.

[0083] Test standard: Referring to the SEMI standard "SEMIP30-0301 Test method for overlay accuracy of semiconductor photomask", and considering the characteristics of OLED photomask, the measurement resolution was set to 0.01μm, and the measurement was repeated 3 times and the average value was taken.

[0084] 1.2 Hole opening accuracy

[0085] Test method: Using a white light interferometer, 20 openings were randomly selected in the mask plate opening area, and the length and width dimensions were measured respectively. The deviation between the actual size and the design size (5-20μm) was calculated, and the maximum deviation value was taken.

[0086] Test standard: According to the mask aperture accuracy test requirements in GB / T32561.3-2016 Semiconductor Devices Discrete Devices Part 3: Radio Frequency Power Devices, the measurement repeatability error is ≤ ±0.02μm.

[0087] 1.3 Sidewall roughness Ra

[0088] Test method: The sidewall of the etched opening was scanned in contact mode using an atomic force microscope (AFM) with a scanning range of 5μm×5μm and a sampling frequency of 1Hz. The arithmetic mean deviation Ra of the sidewall profile was calculated.

[0089] Test standard: in accordance with ISO 4287:2009 Product Geometry Specification (GPS) Surface Structure Profile Method Terminology, Definitions and Surface Structure Parameters, with a resolution set to 0.001μm.

[0090] Table 1. Accuracy Index Test Results

[0091] sample Pattern overlay accuracy error (μm) Hole opening accuracy (μm) Sidewall roughness (μm) Example 1 0.08 ±0.4 0.015 Example 2 0.09 ±0.5 0.017 Example 3 0.07 ±0.3 0.014 Example 4 0.09 ±0.5 0.016 Comparative Example 1 0.45 ±1.8 0.016 Comparative Example 2 0.09 ±0.5 0.017 Comparative Example 3 0.08 ±0.4 0.014 Comparative Example 4 0.08 ±0.6 0.015 Comparative Example 5 0.18 ±0.9 0.018

[0092] As shown in the table above, Examples 1-4, relying on the core process of "MBIL hybrid exposure + fine etching," form a stable and excellent performance cluster in terms of accuracy: the overall overlay accuracy is controlled between 0.07-0.09μm, all meeting the stringent requirement of <0.1μm for high-resolution evaporation. Among them, Example 3 achieves the optimal 0.07μm due to the highest MBIL energy and the slowest scanning rate; the aperture accuracy is concentrated between ±0.3-±0.5μm, adapting to different aperture designs of 5-12μm, ensuring no overlap of evaporation pixels; the sidewall roughness is all <0.02μm, with a minimum of 0.014μm, thanks to the precise micro-polishing of the HF+HNO mixed solution for 6-12s, effectively reducing scattering of evaporation particles and ensuring uniformity of display brightness. Overall, this group of examples, through subtle gradient adjustments to process parameters, achieves both overall leading accuracy indicators and adaptability to different aperture size requirements, demonstrating the systematic nature and scenario flexibility of the solution in accuracy control.

[0093] As can be seen from the comparison between Example 1 and Comparative Example 1, MBIL technology is the core factor affecting accuracy: Comparative Example 1 has an overlay accuracy of 0.45μm and an aperture accuracy of ±1.8μm, which are 5.6 times and 4.5 times that of Example 1, respectively. This is because MBIL's submicron-level laser interference alignment can overcome the limitations of optical diffraction, while a single step scan cannot achieve fine-area pattern alignment.

[0094] A comparison between Example 2 and Comparative Example 5 shows that real-time parameter adjustment has a significant impact on accuracy parameters: Comparative Example 5 (without automatic adjustment) has an overlay accuracy of 0.18 μm and an aperture accuracy of ±0.9 μm, which are 2 times and 1.8 times that of Example 2, respectively. The core reason is that the MBIL energy is prone to fluctuation due to the heating of the laser source during mass production (±5 mJ / cm). 2 The stepping scanning rate is slightly affected by mechanical vibration. Example 2 corrects the deviation by automatically adjusting (lowering the rate when the energy is high and increasing the rate when the energy is low), while Comparative Example 5 has fixed parameters, which leads to pattern alignment offset and blurred opening edges, confirming that automatic adjustment is the key guarantee for accuracy and stability.

[0095] Test Example 2

[0096] Thermal stability tests were conducted on the masks prepared in Examples 1-4 and Comparative Examples 1-4. The test methods and standards are as follows, and the test results are shown in Table 2.

[0097] 2.1 Corrosion area in salt spray test

[0098] Test method: Following the requirements of the neutral salt spray test (NSS), the mask was placed in the salt spray corrosion test chamber, and the following settings were established: 5% NaCl solution (pH 6.5-7.2), 35℃ temperature, and 2 mL / (h·cm). 2 The spray volume was maintained for 48 hours. After the test, the area was rinsed with deionized water, and the corrosion area was observed using an optical microscope (400x) to calculate the corrosion area percentage.

[0099] Test standard: Complies with "GB / T10125-2021 Artificial Atmosphere Corrosion Test Salt Spray Test", with corrosion area measurement accuracy ≤0.01%.

[0100] 2.2 Bending Test Resistance Change Rate

[0101] Test method: Using an electronic universal testing machine, the mask plate was fixed on the bending fixture, the bending radius was set to 5mm and the bending angle to 180°, and the bending was repeated 1000 times. The surface resistance was measured with a digital source meter before and after bending (using the four-probe method with a probe spacing of 1mm), and the resistance change rate was calculated (ΔR / R×100%, where R is the initial resistance).

[0102] Test standard: Refer to "GB / T24467-2021 Flexible Display Devices Part 2: Mechanical Performance Test Methods", resistance measurement accuracy ≤0.1Ω / □.

[0103] 2.3 Surface Resistance

[0104] Test method: Using a four-probe resistance tester, select 5 test points evenly on the surface of the mask, apply a test current of 1mA, read the surface resistance value, and take the average value;

[0105] Test standard: According to "SJ / T11598-2016 Test method for surface resistance of transparent conductive film", the test environment temperature is 25℃±2℃ and the humidity is 50%±5%.

[0106] Table 2. Results of Thermal Stability Test

[0107]

[0108]

[0109] As shown in the table above, Examples 1-4, utilizing the "ALD deposition of Al2O3-ITO protective layer" process, construct a protective performance system far exceeding existing technologies: the overall salt spray corrosion area is ≤0.07%, with a minimum of only 0.02%, effectively blocking corrosive media such as salt spray and humidity, suitable for long-term use scenarios such as automotive and outdoor applications; the bending resistance change rate is ≤3.8%, with a minimum of 2.5%, and the conductivity remains stable after 1000 bends, meeting the requirements for flexible OLED applications; the film porosity is <0.1%, relying on the atomic-level deposition characteristics of ALD, compared to 5.2% for the comparative example magnetron sputtering, improving the ability to block corrosive media by 67%-98%, becoming the core guarantee of excellent protective performance. Overall, this group of examples, through the gradient design of film thickness, not only achieves overall leading protective performance, but also achieves lightweighting while meeting protection standards by thinning the film layer (as in Example 4), adapting to scenarios with different protection and weight requirements, demonstrating the flexibility and reliability of the protection solution.

[0110] The comparison between Comparative Example 2 and Example 2 shows that ALD technology is the core influencing factor of protective performance: Comparative Example 2 (without ALD) has a salt spray corrosion rate of 0.95% and a bending resistance change rate of 13.5%, which are 47.5 times and 5.4 times that of Example 2, respectively. Because the porosity of the magnetron sputtered film is 5.2% (ALD film < 0.1%), it cannot block the salt spray medium and has high internal stress, making it prone to cracking when bent.

[0111] The comparison between Comparative Example 5 and Example 2 shows that real-time parameter adjustment has a significant impact on protective performance: Comparative Example 5 (without automatic adjustment) showed a salt spray corrosion area of ​​0.25%, a bending resistance change rate of 7.8%, and a surface resistance of 22 Ω / □, which are 12.5 times, 3.1 times, and 1.4 times that of Example 2, respectively. The core reason is that the pressure in the reaction chamber during ALD deposition is easily affected by precursor volatilization fluctuations (±1 Torr). Example 2 ensured a dense film by automatically adjusting (shortening the purge time when the pressure is high and extending the pulse time when the pressure is low), while Comparative Example 5, with its fixed parameters, resulted in insufficient precursor to form pores when the pressure was low and residual gas inclusions when the pressure was high, leading to a decrease in the barrier properties and toughness of the film. This confirms that automatic adjustment is a key supplement to achieving the required protective performance.

[0112] Test Example 3

[0113] The protective performance of the masks prepared in Examples 1-4 and Comparative Examples 1-4 was tested. The test methods and test standards are as follows, and the test results are shown in Table 3.

[0114] 3.1 150℃ Heat Load Shadow Offset

[0115] Test method: A vapor deposition simulation test platform was built. The mask was heated to 150℃ and held for 30 minutes. The alignment of the mask opening with the reference plate below was captured by a high-resolution camera (12 million pixels). The offset of the opening center position was analyzed using Image-ProPlus software, and the maximum offset was taken.

[0116] Test standard: Refer to "T / CSTM00373-2022 Technical Requirements for Metal Masks for OLEDs", position measurement accuracy ≤ 0.1μm.

[0117] 3.2 Thermal stress value at 150℃

[0118] Test method: The curvature change of the mask plate was measured using a stress curvature meter under a vacuum environment (1×10⁻³ Pa) at 150℃; according to the Stoney formula (σ=(E·t)... 2 ) / (6·(1-ν)·R·T)) calculates the thermal stress value (E is the elastic modulus, t is the substrate thickness, ν is Poisson's ratio, R is the radius of curvature, and T is the temperature);

[0119] Test standard: Follow the ASTM F2546-13 standard method for testing residual stress in thin films, with stress measurement accuracy ≤0.1MPa.

[0120] 3.3 Warpage

[0121] Test method: Using a surface roughness tester, select 10 measurement lines along the X and Y axes on the surface of the mask, measure the height difference at each point, and calculate the maximum warpage (the height difference between the highest and lowest points divided by the measurement length).

[0122] Test standard: Complies with "SEMIM13-0301 Semiconductor Wafer Warpage Test Method", measurement accuracy ≤0.1μm / m.

[0123] Table 3 Test Performance Results

[0124] sample Shading offset under 150℃ thermal load (μm) Thermal stress at 150℃ (MPa) Warpage (μm) Example 1 2.5 12.5 4.2 Example 2 2.3 11.8 4.8 Example 3 2.1 11.2 4.0 Example 4 2.7 13.0 3.9 Comparative Example 1 3.8 12.7 4.3 Comparative Example 2 2.5 14.2 5.8 Comparative Example 3 9.2 21.5 7.5 Comparative Example 4 11.5 25.3 8.2 Comparative Example 5 5.3 17.8 6.1

[0125] As shown in the table above, Examples 1-4, using the "FEA biomimetic gradient mesh + low-temperature annealing" process, exhibit highly controllable thermal stability indicators that far exceed those of the comparative schemes: the overall thermal stress at 150℃ is stable at 11.2-13.0MPa, far lower than the comparative scheme's 21.5-25.3MPa, effectively preventing permanent deformation of the mask under thermal load. Example 3 achieves the lowest stress of 11.2MPa due to the finest biomimetic mesh gradient; the shadow offset at 150℃ is all <3μm, falling within the 2.1-2.7μm range, meeting the 8K OLED evaporation threshold requirements, which is significantly improved compared to the comparative scheme's 3.8-11.5μm, ensuring precise pixel evaporation positioning; the overall warpage is ≤5μm / m, concentrated in the 3.9-4.8μm / m range, with residual processing stress released through 200℃ annealing (annealing was omitted in Example 4 because the initial warpage met the standard), preventing poor adhesion between the mask and the substrate during evaporation. This set of embodiments demonstrates excellent thermal stress dispersion and deformation control capabilities, which can be adapted to long-term high-temperature evaporation deposition scenarios, and can also achieve mass production efficiency through process simplification, reflecting the comprehensiveness and practicality of thermal stability solutions.

[0126] The comparison between Comparative Example 3 and Example 3 shows that the FEA biomimetic structure is the core influencing factor for thermal stability: the thermal stress of Comparative Example 3 (without FEA) is 21.5 MPa and the shadow offset is 9.2 μm, which are 1.9 times and 4.4 times that of Example 3, respectively. Because the uniform mesh cannot disperse the heat load of evaporation, the stress is concentrated at the edge of the window; while the gradient mesh of Example 3, through FEA topology optimization, can improve the thermal stress dispersion efficiency by 47%.

[0127] The comparison between Comparative Example 4 and Example 1 shows that the annealing process is an auxiliary guarantee for thermal stability: Comparative Example 4 (without annealing) has the highest thermal stress of 25.3 MPa and shadow offset of 11.5 μm, which is 102%-360% higher than that of Example 1 (12.5 MPa / 2.5 μm). This indicates that annealing at 200℃ can release the residual stress of substrate processing and reduce the stress superposition under thermal load.

[0128] A comparison between Comparative Example 5 and Example 3 shows that real-time parameter adjustment has a significant impact on protective performance: Comparative Example 5 (without automatic adjustment) has a thermal stress of 17.8 MPa, a shadow offset of 5.3 μm, and a warpage of 6.1 μm / m, which are 1.5 times, 2.3 times, and 1.3 times that of Example 2, respectively. Although it is better than Comparative Examples 3-4, it is still inferior to Example 2. This is because Example 2 accurately releases stress through automatic adjustment, while Comparative Example 5 is fixed at 200°C. For high warpage substrates (e.g., 6.2 μm / m), stress release is insufficient, and for low warpage substrates (e.g., 3.9 μm / m), over-annealing leads to stress rebound, highlighting the importance of automatic adjustment for thermal stability adaptability.

[0129] In summary, this invention, through a closed-loop design of "MBIL hybrid exposure + FEA biomimetic structure + ALD protection + automatic parameter adjustment," precisely addresses pain points, improves overlay accuracy to 0.07-0.09μm and aperture accuracy to ±0.3-±0.5μm, thus improving accuracy defects; controls 150℃ thermal stress at 11.2-13.0MPa and shadow offset at 2.1-2.7μm, alleviating thermal deformation problems; enhances protection capabilities by reducing salt spray corrosion to ≤0.07% and bending resistance change rate to ≤3.8%; and automatically adjusts and corrects parameter fluctuations to ensure stable mass production. The synergistic effect of each stage achieves a breakthrough in overall performance.

[0130] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details of the above embodiments. Within the scope of the technical concept of the present invention, various simple modifications can be made to the technical solutions of the present invention, and these simple modifications all fall within the protection scope of the present invention. Furthermore, it should be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the present invention will not further describe the various possible combinations.

[0131] Furthermore, various different embodiments of the present invention can be combined in any way, as long as they do not violate the spirit of the present invention, they should also be regarded as the content disclosed by the present invention.

Claims

1. A method for fabricating an OLED metal mask, characterized in that, Includes the following steps: S1: Provide an Invar alloy substrate, which is patterned using a hybrid exposure mode of multi-beam interference lithography and step-scan projection exposure, wherein the multi-beam interference lithography is used to form patterns in fine areas, and the step-scan projection exposure is used to form large-area patterns in non-fine areas. S2: The patterned substrate is etched using FeCl3-based etching solution, and then the etched substrate is polished for 6-12 seconds using HF-HNO3 mixed solution to obtain an opening with a sidewall roughness Ra≤0.02μm. S3: In the non-evaporation window area of ​​the etched and polished substrate, a biomimetic mesh stress relief structure with a gradient unit size is designed and fabricated, wherein the mesh unit size adjacent to the evaporation window area is 15-50μm, and the mesh unit size far from the window area is 140-200μm. S4: Using atomic layer deposition technology, AlO layer and ITO layer are sequentially deposited on the surface of the substrate with biomimetic mesh stress relief structure to form a composite protective layer, reducing the surface resistance to below 20Ω / □.

2. The method for manufacturing an OLED metal mask according to claim 1, characterized in that, Before step S1, the substrate is first ultrasonically cleaned with deionizer and organic solvent in sequence, and then subjected to plasma surface activation treatment.

3. The method for manufacturing an OLED metal mask according to claim 1, characterized in that, In step S1, the specific process of the hybrid exposure mode is as follows: First, multi-beam interference lithography is used to expose the fine area pattern, with an exposure energy density of 140-180 mJ / cm². 2 After completion, switch to the step-scan projection exposure system at 2-5 J / cm. 2 The exposure energy is used to scan and expose non-fine areas, and the pattern overlay accuracy error during the switching between the two modes is less than 0.1μm.

4. The method for manufacturing an OLED metal mask according to claim 1, characterized in that, In step S2, the HF-HNO3 mixed solution has an HF concentration of 0.1-0.5 wt%, an HNO3 concentration of 1%-3 wt%, and the remainder is deionized water.

5. The method for manufacturing an OLED metal mask according to claim 1, characterized in that, In step S3, the biomimetic mesh stress relief structure is designed through topology optimization using finite element analysis. The gradient of its unit size change is verified by Monte Carlo vapor deposition simulation to ensure that its simulated shadow offset under a 150℃ thermal load is less than 3μm.

6. The method for manufacturing an OLED metal mask according to claim 5, characterized in that, In step S3, the processing is carried out using laser processing with a laser wavelength of 1064nm, a pulse width of 30-50ns, and a processing power of 8-10W. Alternatively, electron beam processing can be used, with an accelerating voltage of 20-30kV and a beam current of 20-30nA. The machining accuracy is controlled within ±1μm, and plasma deburring is performed for 1-3 minutes after machining.

7. The method for fabricating an OLED metal mask according to claim 1, characterized in that, In step S4, the deposition temperature is 220-280℃, and the deposition pressure is 3-8 Torr; the Al2O3 layer is deposited using trimethylaluminum and H2O as precursors, and the deposition rate is... / cycle, thickness 35-75nm, corresponding to 1750-2679 deposition cycles; ITO layer deposition uses indium tetrachloride and trimethylgallium as precursors, deposition rate is / cycle, thickness is 110-190nm, corresponding to 2444-6786 deposition cycles; during the deposition process, the precursor pulse time is 0.2-0.5s, and the purge time is 6-10s.

8. The method for manufacturing an OLED metal mask according to claim 7, characterized in that: It also includes online monitoring and feedback control steps: Before step S4, the warpage of the substrate is monitored using a stress curvature meter. If the warpage is >5μm / m, low-temperature annealing is performed immediately until the warpage is ≤5μm / m. In step S4, a white light interferometer and a stress curvature meter are integrated for real-time online monitoring, and the monitoring data is fed back to the atomic layer deposition control system to form a closed-loop control loop. The atomic layer deposition parameters are dynamically adjusted to ensure that the mask warpage is ≤5μm / m.

9. The method for manufacturing an OLED metal mask according to claim 8, characterized in that, The online monitoring and control system collects film thickness data of AlO layer and ITO layer and substrate stress curvature data at a frequency of 40-80 times / s. When the film thickness deviation exceeds 2nm, adjust the deposition cycle; When the warpage data deviates from the target value by more than 10%, the atomic layer deposition process parameters are automatically adjusted, including the atomic layer deposition precursor pulse time, purge time, and reaction chamber temperature.

10. An OLED metal mask, characterized in that, It is prepared by the method of manufacturing an OLED metal mask as described in any one of claims 1-9.

Images