A method, apparatus, and readable storage medium for photoresist spin coating

Abstract

This invention relates to the field of semiconductor process technology, specifically to a method, apparatus, and readable storage medium for photoresist spin coating. The problem addressed by this invention is how to achieve uniform film thickness of pigment photoresist on a square substrate. To solve this problem, this invention provides a spin coating method comprising: a first spreading step: while a predetermined amount of photoresist is dropped onto the center of a square substrate, the substrate is rotated at a first rotation speed for a first duration to drive the photoresist to spread and form a liquid colloid reserve layer that does not yet cover the edges of the square substrate; the first rotation speed is 100-200 rpm, and the radius of the liquid colloid reserve layer is controlled between 0.3 and 0.6 times the radius of the inscribed circle of the square substrate; a second spin coating step: after the first spreading step, the square substrate is continuously accelerated from the first rotation speed to a second rotation speed at an angular acceleration of not less than 8000 rpm / s and maintained for a second duration, allowing the liquid colloid reserve layer to expand and cover the surface of the square substrate to form a photoresist film.

Description

Technical Field

[0001] This invention relates to the field of semiconductor process technology, and more specifically, to a method, apparatus, and readable storage medium for photoresist spin coating. Background Technology

[0002] In the manufacturing of semiconductor display panels (such as LCD and OLED), a uniform photoresist film needs to be coated on the substrate. Pigment photoresists, because they contain solid pigment dispersions, have significantly different solvent composition, volatility characteristics, and leveling properties compared to ordinary photoresists.

[0003] When using conventional spin coating processes, due to the difference between pigment-based photoresists and ordinary photoresists, increasing the initial spin speed causes the photoresist to thin rapidly. The large surface area causes highly volatile solvents to evaporate instantly, and the edge areas form a skin prematurely, making it difficult to flow. In contrast, during conventional gradual acceleration, the effective shear force generated at the edges of the photoresist is significantly reduced, insufficient to drive the flow of the photoresist that has initially formed a skin at the edges. The simultaneous decrease in shear force and the increase in solid content at the edges of the photoresist make it difficult for the photoresist to spread further towards the corners after it has spread to the edges. Ultimately, this results in an uneven photoresist film on square substrates (such as color filters and photomasks), manifested as "rainbow-like stripes".

[0004] In related technologies, adding a cover plate during photoresist coating can control the airflow in the cavity to improve yield, but it cannot solve the problems of color streaks and uneven film thickness caused during pigment photoresist coating. Therefore, how to ensure the thickness and uniformity of the photoresist film coating on the surface of a square substrate is one of the problems that urgently need to be solved by those skilled in the art. Summary of the Invention

[0005] The problem solved by this invention is: how to solve the problem of uniform film thickness of pigment adhesive on a square substrate.

[0006] To address the above problems, embodiments of the present invention provide a method for spin-coating photoresist, applied to coating photoresist on a square substrate. The photoresist is a pigment-based photoresist containing a solid pigment dispersion. The spin-coating method includes:

[0007] First spreading step: When the preset amount of photoresist is dropped onto the center of the square substrate, the square substrate is rotated at a first speed for a first duration to drive the photoresist to spread and form a liquid colloidal reserve layer that does not yet cover the edge of the square substrate; wherein, the first speed is 100-200 rpm, and the radius of the liquid colloidal reserve layer is controlled between 0.3 and 0.6 times the radius of the inscribed circle of the square substrate.

[0008] The second spin coating step: After the first spreading step is completed, the square substrate is accelerated from the first rotation speed to the second rotation speed without interruption at an angular acceleration of not less than 8000 rpm / s, and maintained for a second duration, so that the liquid colloid reserve layer expands and covers the surface of the square substrate to form a film. The second rotation speed is not less than 1000 rpm.

[0009] Compared with existing technologies, the technical effects achieved by this solution are as follows: through the synergistic effect of the first spreading step and the second homogenizing step, the problem of filling the corners of the pigment adhesive on the square substrate is fundamentally solved. This design avoids premature skinning at the edges caused by the traditional "one-time full-coverage" process, and at the same time, it uses high acceleration to provide shear force that breaks through the yield stress of the pigment adhesive within milliseconds, allowing the adhesive to complete the corner rearrangement before the solvent evaporates.

[0010] In one embodiment of the present invention, the first duration is 1 to 3 seconds and the second duration is 3 to 5 seconds, so as to complete the planarization of the adhesive film before the solvent evaporates and causes the adhesive film to initially set.

[0011] Compared with existing technologies, the technical effects achieved by this solution are as follows: the precise matching of the first and second time durations matches the volatility characteristics of the pigment adhesive solvent, ensuring that the reserve layer can be fully formed without causing excessive solvent evaporation and skinning in the edge areas. This allows the adhesive film planarization process to be completed before the adhesive viscosity increases dramatically.

[0012] In one embodiment of the present invention, the second rotational speed is 1000-1600 rpm.

[0013] Compared with existing technologies, the technical effects achieved by this technical solution are as follows: the limitation of the second rotation speed range allows the second rotation speed and the amount of adhesive to work together to ensure complete coverage of the corners while controlling the final film thickness within the process window, resulting in film thickness uniformity that is significantly better than conventional rotation speed solutions.

[0014] In one embodiment of the present invention, after the second spin coating step is completed, an in-situ stabilization step is performed. The in-situ stabilization step is used to allow the adhesive film to be initially shaped by solvent evaporation before the square substrate is removed from the current process environment.

[0015] Compared with existing technologies, the technical effects achieved by this technical solution are as follows: by adding an in-situ stabilization step, the adhesive film can be initially shaped by solvent evaporation before the square substrate is removed from the process environment, effectively preventing secondary defects caused by handling disturbances, and completing the initial curing in the controlled environment of the spin coating chamber, thus solving the problem of qualified spin coating but handling failure.

[0016] In one embodiment of the present invention, the in-situ stabilization step is continued for a preset third duration, which is 5 to 15 seconds.

[0017] Compared with existing technologies, the technical effects achieved by adopting this technical solution are as follows: by limiting the duration of the third time, it is ensured that the surface of the adhesive film forms a shaped skin with sufficient mechanical strength, which enables the adhesive film to obtain the best resistance to handling disturbances while maintaining compatibility with subsequent processes.

[0018] In one embodiment of the present invention, during the in-situ stabilization step, the square substrate remains stationary or rotates at a third rotational speed lower than the first rotational speed.

[0019] Compared with existing technologies, the technical effects achieved by this solution are as follows: by limiting the rotation speed of the square substrate, secondary shearing disturbances to the formed film caused by high-speed rotation are avoided, ensuring that no new thickness fluctuations are introduced during the shaping process.

[0020] In one embodiment of the present invention, during the in-situ stabilization step, when the square substrate rotates at a third rotational speed, the third rotational speed is 40-70 rpm.

[0021] Compared with existing technologies, the technical effects achieved by this technical solution are as follows: the limitation of the third rotation speed optimizes the airflow distribution of the FFU in the homogenizing chamber, and the generated weak centrifugal force can effectively homogenize the airflow differences in the chamber and promote the uniform evaporation of solvent on the surface of the film. At the same time, the rotation speed is significantly lower than the setting of the first spreading speed, which also ensures that no hydrodynamic effects sufficient to change the film thickness distribution are generated.

[0022] In one embodiment of the present invention, in the first spreading step, the volume of photoresist dropped onto the center of the square substrate is 0.60 ml to 0.80 ml.

[0023] Compared with existing technologies, the technical effects achieved by this solution are as follows: the photoresist volume range, the 100-200rpm reservoir design, and the 1000-1600rpm spin coating process work together to ensure full filling of corners while avoiding edge defects, significantly improving the process window tolerance.

[0024] In one embodiment of the present invention, a spin coating apparatus is also provided. The above-mentioned spin coating method is applied to the spin coating apparatus. The spin coating apparatus includes: a rotating stage for supporting a square substrate; a servo motor connected to the rotating stage and capable of providing an angular acceleration of not less than 8000 rpm / s to drive the rotating stage to rotate; a dispensing device for dispensing photoresist onto the square substrate; and a controller for controlling the rotating stage and the dispensing device. The spin coating apparatus has all the technical features of the above-mentioned spin coating method, which will not be described in detail here.

[0025] In one embodiment of the present invention, a computer-readable storage medium is also provided, on which a program or instructions are stored, which, when executed by a processor, implement the steps of the control method described above.

[0026] Compared with existing technologies, the technical effect achieved by adopting this technical solution is that the above-mentioned control method can be realized through programs and instructions on a computer-readable storage medium. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the morphology of the colored patterns after the pigment adhesive film is formed on the square substrate of the present invention;

[0028] Figure 2 This is a schematic diagram showing the distribution of spectral measurement points in this invention;

[0029] Figure 3 This is a schematic diagram illustrating the relationship between spreading speed and spreading time.

[0030] Figure 4 for Figure 2 One of the schematic diagrams showing the distribution of spectral peaks measured at various measurement points;

[0031] Figure 5 for Figure 2 The second schematic diagram of the distribution of spectral peaks measured at various measurement points;

[0032] Figure 6 for Figure 2 The third schematic diagram of the distribution of spectral peaks measured at various measurement points;

[0033] Figure 7 This is a system diagram of the gelling equipment of the present invention;

[0034] Explanation of reference numerals in the attached figures:

[0035] 2. Measurement area; 100. Substrate; 110. Color pattern; 200. Spreading equipment; 210. Rotary stage; 220. Servo motor; 230. Dispensing device; 240. Controller. Detailed Implementation

[0036] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0037] [First Embodiment]

[0038] See Figures 1 to 6 In one specific embodiment, this application provides a method for spin-coating photoresist, characterized in that it is applied to coating a square substrate with photoresist, wherein the photoresist is a pigment adhesive containing a solid pigment dispersion, and the spin-coating method includes:

[0039] S100, First spreading step: When a preset amount of photoresist is dropped onto the center of a square substrate, the square substrate is rotated at a first speed for a first duration to drive the photoresist to spread and form a liquid colloidal reserve layer that does not yet cover the edge of the square substrate; wherein, the first speed is 100-200 rpm, and the radius of the liquid colloidal reserve layer is controlled between 0.3 and 0.6 times the radius of the inscribed circle of the square substrate;

[0040] S200, Second Spreading Step: After the first spreading step is completed, the square substrate is accelerated from the first rotation speed to the second rotation speed without interruption at an angular acceleration of not less than 8000 rpm / s, and maintained for a second duration, so that the liquid colloid reserve layer expands and covers the surface of the square substrate to form a film. The second rotation speed is not less than 1000 rpm.

[0041] The coating process for photoresists with rapid solvent evaporation is often more difficult than that for common photoresists. This is because the rapid solvent evaporation leads to faster curing and film formation, thus amplifying the impact of process variations or disturbances on film formation control, making uneven film thickness more likely. Furthermore, in the field of pigment photoresists, besides differences in solvent formulation, another significant difference lies in the presence of solid pigment dispersions, such as pigment particles. Unlike ordinary photoresists without solid pigment dispersions, these solid pigment dispersions are essentially suspended solid particles in solution. They also affect the viscosity, thixotropy (the shear-thinning characteristic of non-Newtonian fluids under sustained shear stress), and leveling characteristics of the photoresist, making it more difficult to obtain a uniform film after homogenization.

[0042] Taking a 77mm square substrate for a light filter as an example, made of glass, the goal is to coat the substrate surface with a pigment adhesive film of uniform thickness. When using EXCOLOR brand KT-B-01 pigment adhesive, the uniformity requirement is that within the area enclosed by the four sides (1.5mm) and corners (2mm) of the substrate as specified in the inspection standard, the spectral peak range should be less than 1.5nm. The process environment in the spin coater is typically maintained under standard conditions. Industry-standard process environment conditions generally include a temperature maintained between 22 and 23 degrees Celsius, a relative humidity controlled between 45% and 55%, and a positive pressure maintained inside the process chamber relative to the external environment to prevent contamination from external particles. In the original typical process, visible outward-expanding colored streaks would form on the surface of the square substrate after pigment adhesive film deposition.

[0043] The visible colored stripes are formed by the interference of reflected light after the incident light is reflected from the upper and lower surfaces of the adhesive film. The formation of these stripes indicates local differences in the film thickness, thus allowing for a quick check of the film thickness uniformity. Simultaneously, the film thickness can be quantitatively verified using a film thickness gauge or spectrophotometer. Generally, we consider thickness uniformity exceeding the preset tolerance to mean the formation of unacceptable visual colored stripes, or the range or standard deviation of the quantitatively measured film thickness or corresponding spectral peaks exceeding a given acceptable range. That is, within the area of ​​the substrate enclosed by four sides of 1.5mm and corners of 2mm as required by the inspection standard, the spectral data of the range of 100 squares within the substrate is better than 1.5nm. It should be noted that the inspection standard refers to the near disappearance of colored stripes on the surface of a square substrate, allowing for deviations outside the specified area. The range requirement in the standard is primarily because for filters used in displays, the edge areas are invalid areas and will be trimmed later.

[0044] Regarding the spin coating process for the aforementioned pigment resist, the uniformity was found to be excessive when using conventional spin coating techniques. Qualitatively, this manifested as visible color-streaked defects on the resist film surface after spin coating. Specifically, we recorded two spin coating methods. In the first method, namely Comparative Example 2 in the following table, the square substrate was rotated at 50 rpm, and a certain amount of pigment resist was simultaneously dropped onto the center of the square substrate and held for 1-3 seconds. At this low speed, the pigment resist spread slightly but remained mainly concentrated in the central area. Then, the square substrate was accelerated from 5000 rpm to a film-forming speed of 1600 rpm. This centrifugal force caused the photoresist to separate and form a resist film. After the square substrate with the resist film was removed, color-streaked defects extending to the corner areas of the substrate were visually visible. Figure 1 This is a schematic diagram of the morphology of the pigment-adhesive film after deposition on a square substrate. The morphology appears in a wavy pattern in the four corner areas of the square. To obtain detailed spectral absorbance distribution data behind this morphology, we designed multiple measurement points along the diagonal of the square. Figure 2 This is a schematic diagram of the distribution of spectral measurement points. The reason for arranging the measurement points along the diagonal of the square is mainly because the colored patterns are concentrated at the four corners of the square. Observing the difference in spectral absorbance between the corner region and the center region along the diagonal can reflect the trend of film thickness variation in the corner region. The area about 1.5mm inward from the upper edge of the substrate is the measurement region 2 defined in this invention where the range of spectral peaks needs to be controlled. In order to quantitatively measure the spectral peak variation on the substrate surface, especially in the corner region, we set 10 measurement points along each of the two diagonals of the substrate, the first diagonal D1 and the second diagonal D2. Each measurement point is the center position after the substrate is divided into 100 grid regions. In the subsequent experimental results, we will sequentially label the 10 measurement points on each diagonal as P1, P2, ..., P10.

[0045] In another spin coating method, namely Comparative Example 1 in the table below, the square substrate was kept stationary, and a certain amount of pigment photoresist was dropped onto the center of the square substrate. The square substrate was then immediately accelerated from 5000 rpm / s to a film-forming speed of 1600 rpm. This centrifugal spin-split the photoresist to form a film. After the square substrate with the film was removed, color defects were still visually visible. The wavy color defects were mainly concentrated in the corner areas isolated by the edges of the square substrate, or more simply, most of the color defects were concentrated in the corner areas outside the inscribed circle of the square substrate. Figure 1 The distribution shown is similar. Figures 4 to 6 As shown in Comparative Examples 1, 2, and others, we use... Figure 2 The diagram shows a broken line illustrating the distribution of spectral peaks at each point along the diagonal, as measured at the indicated measurement points.

[0046] Experiments showed that adjusting the spin speed and time, or using various combinations of different spin speeds, had little effect on improving the film thickness uniformity and eliminating color streaks in this pigment adhesive. Therefore, spin speed control is not the root cause of the color streaks in the pigment adhesive.

[0047] Based on this discovery, the technical solution of this application is not limited to adjusting the spin coating speed and time, because actual testing cannot solve the problem. Instead, it breaks down the original spin coating process into two stages.

[0048] In step S100, firstly, the square substrate is rotated at 150 rpm. Then, a certain amount of pigment adhesive is dropped onto the center of the square substrate and held for 1-3 seconds. At this low speed, the pigment adhesive spreads slightly but remains mainly concentrated in the central area. Next, the square substrate is accelerated to 1500 rpm at an acceleration of no less than 8000 rpm / s and held for about 3 seconds. This allows the photoresist to be centrifuged and formed into a film. After standing and initial shaping, it is removed and inspected; visually, the color patterns have largely disappeared. In this embodiment, to avoid the influence of the amount of photoresist used, a slightly redundant amount of photoresist, specifically 0.75 ml, is used to ensure that the photoresist can cover the entire substrate surface at the initial stage of the process design while achieving the predetermined thickness. The slightly excess photoresist can be removed during the homogenization process without any negative impact.

[0049] The 150 rpm rotation speed here is primarily to ensure that the photoresist added within a limited time of 1-3 seconds can spread to a suitable area on the square substrate, forming a disc-shaped liquid colloid reserve layer that does not yet cover the four corners of the square substrate. This prepares the substrate for the subsequent homogenization process. The target spreading area is controlled between 0.3 and 0.6 times the radius of the inscribed circle of the square substrate. The 1500 rpm film-forming speed provides sufficient centrifugal force for the photoresist to spread, expand, and thin out the substrate surface to form a film. This is usually determined by a combination of specific process film thickness requirements and the amount of photoresist added. In the first spreading step, rotating the square substrate at a first speed of 100 to 200 rpm is crucial. If the photoresist is added while the substrate is stationary, the velocity of the photoresist upon contact with the substrate is zero, and the photoresist tends to accumulate in the center, forming a raised clump. Even with subsequent high rotation speeds, it is difficult to completely eliminate the problem of excessive film thickness in the center. If the initial rotation speed is too low, for example, below 50 rpm, the centrifugal force generated by the rotation is insufficient to overcome the surface tension and contact angle resistance of the photoresist on the substrate surface, resulting in ineffective spreading of the photoresist and only the formation of irregular liquid coverage areas. Conversely, if the initial rotation speed is too high, for example, above 300 rpm, the centrifugal force is too large during the dispensing process, and splashing will occur before the photoresist is fully dispensed, not only wasting the photoresist but also causing surface defects. Therefore, in this embodiment, the initial rotation speed is strictly controlled between 100-200 rpm, coupled with an initial dispensing time of 1-3 seconds. This parameter window ensures that the photoresist obtains a suitable tangential velocity at the moment it falls into the center of the substrate, and under the balance of centrifugal force and surface tension, it gently spreads to form an approximately disk-shaped liquid colloidal reserve layer with a radius 0.3 to 0.6 times the radius of the inscribed circle of the square substrate. This liquid colloidal reserve layer is in a dynamic flow state, and because the rotation time is short and the speed is low, there is no significant solvent evaporation that would lead to surface crusting or accumulation of solid particles, thus ensuring the subsequent homogenization effect. After the first spreading step, the area where the pigment adhesive forms on the substrate can be referenced. Figure 1 It diffuses outward in a basically circular pattern, but does not contact the edge of the square substrate. We also experimented with the radius of the spread adhesive region at different time lengths and spreading speeds. Figure 3 Data on spreading radius under different time and rotation speed conditions.

[0050] Choosing a first rotation speed of 100 to 200 rpm, coupled with a relatively short initial duration (1 to 3 seconds), is crucial for controlling solvent evaporation. If the rotation speed is too low or the dispensing is static, the adhesive will remain in the center of the substrate for too long, causing solvent evaporation to occur first at the edges of the central liquid mass, forming a high-viscosity surface layer. During subsequent high-speed spin coating, the adhesive must break through this high-viscosity ring to diffuse outwards, resulting in a final film thickness distribution that is thicker in the center and thinner at the edges. By setting the aforementioned first rotation speed, the adhesive is gently spread out upon drop, increasing the specific surface area of ​​the liquid layer while maintaining its fluidity, thus preventing the formation of a thick local liquid mass in the center. The centrifugal force generated by this rotation speed ensures continuous flow of the adhesive for a short period, minimizing solvent evaporation from the surface and the accumulation of solid pigment particles, preventing the formation of a high-viscosity surface layer, and ensuring a relatively uniform solvent content distribution throughout the liquid colloid reservoir layer, which is beneficial for the stable distribution of the adhesive during subsequent spin coating.

[0051] In step S200, after the first spreading step, the second spin coating step proceeds directly without interruption. "Without interruption" means that the servo motor immediately executes an acceleration command at the same time as or within a very short time after the first spreading step, without any manually set pauses or deceleration. This is because the solvent in the pigment adhesive evaporates quickly, and the surface area increases after spreading. Any pause could lead to excessive solvent evaporation on the surface and edges of the liquid colloid reserve layer, forming a high-viscosity surface layer, making spin coating difficult and ultimately resulting in streaks at the edges of the film.

[0052] Providing an angular acceleration of at least 8000 rpm is one of the key thresholds in the experiment. Experiments show that when the angular acceleration is below this value, such as at 7500 rpm, the shear stress applied to the adhesive does not increase rapidly enough to completely overcome the thixotropy of the pigment adhesive instantaneously, thus insufficiently reducing its viscosity. As the adhesive flows towards the corners, although it is flowing, it still maintains a relatively high viscosity, resulting in significant flow resistance. This leads to insufficient redistribution of the adhesive in the corner areas of the square substrate, ultimately producing slight streaks or occasional adhesive defects.

[0053] When the angular acceleration increases to 8000 rpm or higher, the substrate rotation speed increases dramatically in a very short time. The shear stress applied to the liquid colloidal reservoir instantly exceeds the yield limit of the internal pigment particle agglomeration structure. This causes the viscosity of the adhesive to rapidly decrease to a level close to that of its solvent system itself in the initial stage of spreading, giving the adhesive excellent flowability and self-leveling ability. This high acceleration also greatly shortens the time required for the adhesive to spread from the center to cover the entire substrate, typically reaching the target uniform spreading speed within 0.2 seconds. The speed of the entire physical spreading process is much faster than the speed at which the viscosity recovers due to solvent evaporation. This ensures that the adhesive can quickly cover the substrate surface in a uniform and stable fluid state before its rheological properties deteriorate, effectively filling the four corner areas, thereby completely eliminating color defects caused by uneven flow. Conversely, if the angular acceleration is low, such as 5000 revolutions per second, the viscosity cannot be reduced rapidly. The spread of the adhesive during the homogenization process is unstable, and as the solvent evaporates, the flow resistance will increase. Eventually, the flow may stop before reaching the four corners, resulting in missing adhesive or severe streaking defects.

[0054] The following are the experimental results of the process debugging experiment. It should be noted that Comparative Example 2 is the experimental data for uniform coating using 0.65 mL of KT-B-01 pigment adhesive. The spreading time was 3 seconds, and the uniform coating time was 3 seconds, meaning the total process was kept to around 6 seconds, which is relatively short and ensures that the impact of solvent evaporation is relatively controllable. The spreading speed was 50 rpm, and the adhesive was dripped onto the center of the substrate at the beginning of spreading. The uniform coating speed was 1600 rpm, and the transition acceleration was 5000 rpm / s.

[0055] Comparative Example 1 is based on Comparative Example 2, but the spreading speed is reduced to zero, which is static dispensing.

[0056] Comparative Example 3 increases the spreading speed to 150 rpm based on Comparative Example 2.

[0057] Comparative Example 4 increases the acceleration to 7500 rpm / s based on Comparative Example 3.

[0058] Comparative Example 5 increases the spreading speed to 250 rpm and the acceleration to 9000 rpm / s, based on Comparative Example 4.

[0059] Comparative Example 6 increases the acceleration to 8000 rpm / s based on Comparative Example 3.

[0060] Example 1 presents experimental data on the homogenization of 0.75 mL of KT-B-01 pigment adhesive. The spreading time was 3 seconds, and the homogenization time was 3 seconds, meaning the total process lasted approximately 6 seconds. This short time ensured that the impact of solvent evaporation was relatively controllable. The spreading speed was 150 rpm. At the start of spreading, adhesive was dropped onto the center of the substrate. After 3 seconds, the diameter of the spread adhesive was approximately 50% of the side length of the square substrate. Then, the homogenization speed was increased to 1600 rpm, with a transition acceleration of 8000 rpm / s.

[0061] Example 2 is based on Example 1, but the spreading speed is reduced to 100 rpm, while the transition acceleration is increased to 9000 rpm / s.

[0062] Example 3 increases the spreading speed to 200 rpm, based on Example 2.

[0063] Example 4 is based on Example 1, but the amount of adhesive is adjusted to 0.65 mL and the spinning speed is increased to 12000 rpm.

[0064] Example 5 is based on Example 4, but the amount of adhesive is adjusted to 0.65 mL, the spinning speed is reduced to 1500 rpm, and the transition acceleration is reduced to 9000 rpm / s.

[0065] Example 6 is based on Example 5, but the spin coating speed is reduced to 1000 rpm and the transition acceleration is reduced to 8000 rpm / s.

[0066] Examples 7 and 8 are based on Example 2, with the amount of adhesive adjusted to 0.6 mL and 0.8 mL respectively.

[0067] Example 9 increases the acceleration to 12000 rpm / s, based on Example 6.

[0068] Example 10 is based on Example 9, but the spin coating speed is reduced to 1000 rpm / s.

[0069] Example 11 is based on Example 10, but the transition acceleration is reduced to 8000 rpm / s.

[0070] Example 12 is based on Example 5, but the spin coating speed is increased to 1600 rpm / s.

[0071] Table 1: Results under different spreading speeds, homogenizing speeds, and accelerations

[0072]

[0073] Table 2: Spectral peak data corresponding to each measurement point on the diagonal after homogenization.

[0074]

[0075] Table 3: Spectral peak data for each measurement point on the diagonal after homogenization (Part 2)

[0076]

[0077] Refer to Comparative Examples 1 and 2 in Table 1. In Comparative Example 1, due to static dispensing, the adhesive accumulates into a thick clump at the center of the substrate. Even with subsequent acceleration of 5000 rpm / s, it is impossible to effectively propel the thick adhesive in the center to the edges, especially the corners, resulting in severe adhesive deficiency at the corners, with a range exceeding 4 nm. In Comparative Example 2, the centrifugal force provided by a rotation speed of 50 rpm is insufficient to overcome the high initial yield stress of the pigment adhesive. The adhesive spread is extremely limited, and the final film exhibits obvious wavy streaks with a range exceeding 3 nm, far exceeding the acceptable standard of 1.5 nm. This indicates that imparting a certain initial tangential kinetic energy to the adhesive is essential. When using the scheme in Comparative Example 5, although the adhesive can spread rapidly when the spreading speed is too high, the excessive centrifugal force stretches the adhesive too thinly in a very short time, leading to a sharp increase in specific surface area. For pigment adhesives containing highly volatile solvents, this will cause a large amount of instantaneous solvent evaporation, resulting in a dramatic increase in viscosity at the edges of the adhesive. When entering the second spin coating step, the high-viscosity adhesive at the edges lost its fluidity. Even with greater acceleration, the increased viscosity of the adhesive itself was insufficient to propel the liquid adhesive evenly and stably to the four corners of the square substrate, ultimately leading to the fatal defect of missing adhesive at the corners, with a range exceeding 2 nm. Both Comparative Examples 1 and 6 exhibited the problem of unmeasurable spectral peaks at the corners due to missing adhesive, but the causes differed. Comparative Example 1 was due to the failure to employ a rotational spreading method in the first spreading step, while Comparative Example 6 was caused by insufficient rotation speed in the second spin coating step.

[0078] For Examples 1-3, it can be observed that when the first rotational speed is strictly controlled between 100-200 rpm, the adhesive solution, under the perfect balance of centrifugal force and surface tension, gently spreads into a reservoir layer of the preset size. At this point, the adhesive solution is neither difficult to move due to accumulation, nor does it cause solvent evaporation and skinning due to excessive spreading, thus laying an ideal fluid foundation for subsequent high-speed homogenization. The range of all examples is consistently between 1.2 nm and 1.5 nm, perfectly meeting the target.

[0079] Refer to Comparative Examples 3 and 4 in Table 1. In Comparative Example 3, although the spreading speed (150 rpm) was suitable, the shear stress increase provided by the acceleration of 5000 rpm / s was too gradual to instantly destroy the agglomeration structure of the pigment particles inside the pigment adhesive. The adhesive maintained a high viscosity as it advanced towards the corners, and when it encountered the geometric island effect at the corners of the square substrate, flow lag and physical accumulation occurred, forming obvious wavy streaks with a range of approximately 3 nm. In Comparative Example 4, increasing the acceleration to 7500 rpm / s reduced the streaks (becoming slight streaks), and the range decreased to 1.5-2 nm, but this still failed to cross the acceptable threshold. This indicates that 7500 rpm / s still does not reach the critical shear force threshold for completely activating the low-viscosity state of the pigment adhesive.

[0080] Referring to Examples 1, 2, and 4, when the angular acceleration reaches or exceeds the critical threshold of 8000 rpm / s, the shear stress applied to the liquid colloidal reservoir instantly exceeds its yield limit, and the viscosity of the adhesive drops drastically to near the level of a pure solvent. The adhesive in its extremely low viscosity state achieves excellent self-leveling ability, rapidly and uniformly washing over and covering the four corners of the square substrate in a very short time, eliminating standing waves and streaks. Data from Examples 1-12 show that as long as the acceleration is ≥8000 rpm / s, regardless of whether the spin coating speed is 1000 rpm or 1600 rpm, the range can be firmly suppressed below the 1.50 nm red line (as low as about 1 nm). Furthermore, Examples 1-4 indicate that as long as the first-stage spreading speed meets the requirements, the first-stage spreading speed has almost no impact on the result. Examples 1-8 also show that under the current process method, as long as the adhesive amount meets the minimum requirement, the spin coating task can be completed without causing adhesive shortage; excess adhesive will eventually be ejected during the spin coating process. Examples 4, 9, and 10 demonstrate that, under optimal conditions during the first-stage spreading process, increasing acceleration positively improves the results, eliminating color streaks and improving the range of spectral peaks. Examples 7-12 further show that, in addition to optimizing the range of spectral peaks by increasing acceleration, adjusting the rotation speed in the second stage has a comprehensive impact on the film thickness; lower rotation speeds result in thicker films, and consequently, a higher spectral peak.

[0081] In the traditional process, after the spin coating stage, the substrate is usually stationary, waiting for the robotic arm to enter the cavity and remove it for inter-process transport. However, considering the special characteristics of pigment-based photoresists or similar photoresists with highly volatile solvents, the film surface is usually only initially formed after spin coating and has not yet been preliminarily shaped. Subsequently, as the robotic arm moves within the equipment channel, the film on the substrate surface often migrates locally due to the acceleration and deceleration of the robotic arm and fluctuations in the internal environment of the equipment. The film cannot maintain the uniform film formation state after spin coating. For conventional photoresists, due to slow solvent evaporation and good leveling properties, disturbances in subsequent steps can be compensated for, i.e., automatic leveling. For the pigment-based photoresists or similar photoresists mentioned above, the presence of solid pigment dispersions means that such characteristics are not present.

[0082] To mitigate the impact of subsequent operations on the uniformity of the film formed after spin coating, we incorporated an in-situ stabilization step. This involves a short waiting period, minimally affecting batch production capacity, to ensure the adhesive film on the substrate has undergone initial stabilization before proceeding with subsequent steps. During this time, subsequent handling or movement will not cause further changes in the adhesive film thickness. This process involves maintaining the substrate in its original spin coating environment, such as the spin coating chamber and its platform, for a third duration. In one approach, after the second spin coating stage, an in-situ step is added to keep the substrate stationary within the spin coating chamber for a third duration, typically 5-15 seconds. This allows the solvent in the adhesive film to evaporate in situ, promoting initial stabilization before subsequent handling steps begin.

[0083] To promote solvent evaporation in the coating film and prevent environmental differences within the spin coater chamber from affecting film uniformity, we set a low, gradual third rotation speed for the substrate. This third rotation speed, for example, 40-70 rpm, is maintained within the same process environment as in the previous steps—i.e., the same spin coater chamber. Since this step is typically only to avoid the impact of environmental differences on film thickness, and not to affect the coating thickness itself, this third rotation speed is significantly lower than the speed used in our spreading step. On one hand, low-speed rotation obviously accelerates solvent evaporation through convection. On the other hand, we cannot guarantee that the environment within the chamber will be circumferentially uniform due to the design of the FFU (air intake unit) and the exhaust path of the spin coater chamber. Therefore, the non-uniformity of the downward airflow in the circumferential direction could potentially cause film thickness variations on the substrate surface when it is stable in situ. Rotating the substrate avoids this problem, ensuring that the impact of environmental differences on film thickness is applied uniformly to the coating film on the substrate over time. To more realistically assess the impact of the in-situ stabilization step on the final product quality, the visual inspection and spectral data detection for each group in the table below were performed after all intermediate steps, including substrate spin coating, in-situ stabilization, and subsequent robotic transfer, and before proceeding to the next photolithography process. The experimental data were obtained in the same way as in Examples 1-12, except that the aforementioned data were measured after spin coating was completed and a 10-minute wait was made to ensure the film remained unchanged. In contrast, the data in the following examples and Comparative Example 7 were measured after the in-situ stabilization step was completed, followed by transfer and another 10-minute wait.

[0084] Example 13 adds an in-situ stabilization step of 80 rpm to Example 1, with a holding time of 10 seconds, followed by 10 minutes for spectral and visual inspection. The comparative example, based on Example 1, omits the in-situ stabilization step; after homogenization, the material is directly transported to the next station by a robotic arm or similar device within the equipment, and the results are measured after the same time interval. Examples 14-17, based on Example 13, design different process parameters for different situations and the in-situ stabilization speed, and repeat the experiments.

[0085] Table 4: Results of different in-situ stabilization speeds after adding the in-situ stabilization step

[0086]

[0087] Table 5: Spectral peak data along the diagonal after the spin coating process is completed following the addition of the in-situ stabilization step.

[0088]

[0089] Examples 14-17 present several different in-situ stabilization methods to test their effectiveness. Comparative Example 7 exhibits significant problems; its spectral peak curves are unevenly distributed along the two diagonals, showing a distribution significantly different from normal conditions. Analysis revealed this was due to lateral flow of the adhesive surface during handling. The pigment adhesive solvent evaporates rapidly, and the adhesive film is in a semi-fluid state immediately after homogenization. The acceleration and deceleration inertia forces during robotic arm handling, as well as possible robotic arm tilting, caused the adhesive to flow and rearrange, resulting in flow on the substrate. Due to the poor leveling properties of the pigment adhesive, this migration-induced thickness deviation cannot self-heal, leading to the clearly asymmetrical experimental results seen in Comparative Example 7.

[0090] After in-situ stabilization, the overall spectral peak curve basically maintained the state after homogenization, without drift or deformation. Furthermore, the range of spectral peaks met the process requirements. Based on this, we can see that, compared to Examples 14 and 16, the spectral uniformity of Examples 13 and 15 after the substrate was rotated was significantly better than that in the static state. In particular, the difference between the two diagonals was smaller. The brief dwell time provided a valuable window for the natural evaporation of the solvent, allowing a sufficiently strong stabilizing skin to quickly form on the adhesive film surface, completing the initial stabilization. At this point, the flow of the adhesive was constrained during robotic handling. The range of values ​​in Examples 14 and 16 was between 1.2 and 1.5 nm, demonstrating the effectiveness of static evaporation.

[0091] Further comparisons were made with Examples 13, 15, and 17. When the in-situ stabilization step employed a low-speed rotation of 40-70 rpm, the range data showed further convergence (mostly between 0.9-1.1 nm, generally better than static stabilization). The underlying reason lies in the slight physical asymmetry in the circumferential direction between the downward airflow (FFU supply) and the exhaust channel within the homogenization chamber. If the substrate evaporates statically, this airflow inhomogeneity and the difference in substrate level will manifest in the film thickness; however, through low-speed rotation of 40-70 rpm, the substrate continuously changes its orientation during evaporation and setting, perfectly homogenizing the slight differences in ambient airflow over time, thus approaching the physical limit of film thickness uniformity. It should be noted that Example 13 used a stable rotation speed of 80 rpm. Although the results were satisfactory, a higher rotation speed may introduce additional centrifugal force interference. Therefore, this invention locks the optimal third rotation speed range at 40-70 rpm.

[0092] It should be noted that, while addressing the film uniformity of photoresists with similar rheological properties through the above process steps, we also require protection for the spin coating equipment that can be used to perform the above process. This equipment is used to execute the spin coating method for photoresist described in any of the foregoing embodiments.

[0093] [Second Embodiment]

[0094] See Figure 7 This application also provides a spin coating apparatus 200. The spin coating method described in the above embodiments is applied to the spin coating apparatus 200. The spin coating apparatus 200 includes a rotating stage 210 for supporting a square substrate 100; a servo motor 220 connected to the rotating stage 210 and capable of providing an angular acceleration of not less than 8000 rpm / s to drive the rotating stage 210 to rotate; a dispensing device 230 for dispensing photoresist onto the square substrate 100; and a controller that controls the rotating stage 210 and the dispensing device 230.

[0095] The rotating stage 210, also commonly referred to as a chuck, carries and holds the substrate 100 via vacuum suction or other means, and is driven by a precision spindle motor to achieve precise speed control and high angular acceleration speed switching. The dispensing device 230 typically includes a movable swing arm and one or more nozzles for precisely dispensing photoresist onto the center of the substrate 100 located on the rotating stage 210, under the command of a controller. The process chamber provides a controlled process environment; its internal modules or other modules of the equipment, such as a temperature and humidity control unit and a gas circulation filter unit (FFU), provide a stable and clean controlled process environment for the photoresist homogenization process, preventing the formation of color patterns 110.

[0096] The controller can be an industrial computer, a programmable logic controller (PLC), or a dedicated control board containing a microprocessor. The controller is electrically connected to actuators such as the spindle motor driver of the rotary stage 210, the drive mechanism of the dispensing device 230, and fluid control valves. The controller integrates a memory and a processor, in which computer program instructions are stored or loaded. When the processor executes these instructions, the controller is configured to send control signals to each actuator strictly according to the logic and parameter sequence defined in the foregoing embodiments. For example, the controller controls the dispensing device 230 to complete dispensing; then immediately controls the spindle motor to accelerate the rotary stage 210 to a first rotational speed and maintain it for a first duration to complete the first spreading step; next, it seamlessly increases the speed to a second rotational speed and maintains it for a preset time to complete the second uniform dispensing step; in embodiments including an in-situ stabilization step, the controller also controls the spindle motor to decelerate to a standstill or a third rotational speed after the second uniform dispensing step and maintain it for a third duration. In this way, the spin coating equipment 200 can completely and accurately reproduce the spin coating process proposed in this invention, thereby solving the problem of film uniformity of photoresists with specific rheological properties.

[0097] Accordingly, the present invention also provides a computer-readable storage medium. This storage medium can be non-volatile, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid-state drive (SSD); or it can be volatile, such as random access memory (RAM). A set of computer programs or instruction codes are stored on this storage medium. When this set of computer programs or instruction codes is loaded and executed by the processor within the controller of the aforementioned spin coating apparatus 200, the processor interprets these instructions and generates corresponding control signals to drive the hardware components of the spin coating apparatus 200 (such as the rotating stage 210, the dispensing device 230, etc.) to execute all the steps of the photoresist spin coating method described in the foregoing embodiments. In other words, the method of the present invention can be implemented on the existing or future spin coating apparatus 200 hardware platform through software programming, and the computer-readable storage medium is the physical carrier of this solution.

[0098] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the invention; therefore, the scope of protection of the present invention should be determined by the scope defined in the claims.

Claims

1. A method of spin coating a photoresist, characterized by, The method for coating photoresist onto a square substrate, wherein the photoresist is a pigment adhesive containing a solid pigment dispersion, includes: First spreading step: When the preset amount of photoresist is dropped onto the center of the square substrate, the square substrate is rotated at a first speed for a first duration to drive the photoresist to spread and form a liquid colloidal reserve layer that does not yet cover the edge of the square substrate; wherein, the first speed is 100-200 rpm, and the radius of the liquid colloidal reserve layer is controlled between 0.3 and 0.6 times the radius of the inscribed circle of the square substrate; Second uniform coating step: After the first spreading step is completed, the square substrate is accelerated from the first rotation speed to the second rotation speed without interruption at an angular acceleration of not less than 8000 rpm / s, and maintained for a second duration, so that the liquid colloid reserve layer expands and covers the surface of the square substrate to form a film, the second rotation speed being not less than 1000 rpm. After the second spin coating step is completed, an in-situ stabilization step is performed. The in-situ stabilization step is used to allow the adhesive film to be initially shaped by solvent evaporation before the square substrate is removed from the current process environment. In the in-situ stabilization step, the square substrate rotates at a third rotation speed, which is 40-70 rpm.

2. The photoresist uniformizing method according to claim 1, wherein The first duration is 1 to 3 seconds, and the second duration is 3 to 5 seconds, to complete the planarization of the adhesive film before the solvent evaporates and the film begins to set.

3. The method for spin-coating photoresist according to claim 1, characterized in that, The second rotational speed is 1000-1600 rpm.

4. The method for spin-coating photoresist according to claim 1, characterized in that, The in-situ stabilization step is continued for a preset third duration, which is 5 to 15 seconds.

5. The method for spin-coating photoresist according to claim 1, characterized in that, In the first spreading step, the volume of photoresist dropped onto the center of the square substrate is 0.60 ml to 0.80 ml.

6. A spin coating device, characterized in that, The spin coating method according to any one of claims 1 to 5 is applied to the spin coating apparatus, wherein the spin coating apparatus comprises: A rotating platform is used to support the square substrate; A servo motor is connected to the rotating platform and is capable of providing an angular acceleration of not less than 8000 rpm / s to drive the rotating platform to rotate. A dispensing device for dispensing photoresist onto the square substrate; A controller that controls the rotating stage and the dispensing device.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it drives the controller of the spin coating equipment to perform the spin coating method for photoresist as described in any one of claims 1 to 5.

Images