Method and system for preparing photoresist for high-resolution display panel

Abstract

The invention provides a preparation method and system of photoresist for a high-resolution display panel, and relates to the field of photoresist preparation, and the method comprises the following steps: randomly generating preparation parameters of the photoresist in a preparation parameter range, carrying out photoresist preparation integrated prediction, and obtaining a plurality of photoresist bubble distributions; performing etching simulation randomly in the plurality of photoresist bubble distributions to obtain a plurality of etching fuzzy rates; and according to the distribution of the plurality of photoresist bubbles and the plurality of etching fuzzy rates, calculating to obtain photoetching fitness, carrying out iterative optimization on photoresist preparation parameters, and obtaining optimal photoresist preparation parameters for preparation. The method aims at solving the technical problem that the precision and details of a photoetching pattern are affected due to the fact that bubble interference cannot be effectively controlled through a traditional photoresist preparation method, the accuracy and reliability of photoresist preparation parameter setting can be remarkably improved through integrated prediction and algorithm optimization, then generation and distribution of bubbles are effectively controlled, and the accuracy and details of the photoetching pattern are effectively improved. And the precision, definition and details of the pattern are improved.

Description

Technical Field

[0001] The present invention relates to the field of photoresist preparation, and particularly to a method and system for preparing a photoresist for a high-resolution display panel. Background Art

[0002] Bubbles not only affect the uniformity of the photoresist during the preparation process, but also have a serious impact on the formation of the photolithographic pattern. Especially in high-resolution lithography, the presence of bubbles can lead to blurred pattern edges, loss of details, and even serious defects.

[0003] Bubbles in the photoresist are usually caused by factors such as uneven stirring, solvent evaporation, or external vibration during the preparation process. In traditional photoresist preparation processes, the generation of bubbles is often inevitable, and it is difficult to accurately control the distribution of bubbles. Although some methods such as ultrasonic stirring are applied to reduce the generation of bubbles, due to the randomness and complexity of bubbles, existing technologies are difficult to achieve accurate prediction and precise control in actual production. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for preparing a photoresist for a high-resolution display panel to solve the technical problem that the traditional photoresist preparation method cannot effectively control bubble interference, resulting in affecting the accuracy and details of the photolithographic pattern, including: In a first aspect, the present invention provides a method for preparing a photoresist for a high-resolution display panel, including: obtaining the range of preparation parameters during the photoresist preparation process, where the preparation parameters include at least one of photosensitive dose, resin amount, solvent amount, and stirring parameters; randomly generating the preparation parameters of the photoresist within the range of the preparation parameters, performing integrated prediction of photoresist preparation, and obtaining multiple photoresist bubble distributions; randomly performing etching simulation within each of the multiple photoresist bubble distributions to obtain multiple etching blur rates; calculating the photolithography fitness based on the multiple photoresist bubble distributions and multiple etching blur rates, performing iterative optimization of the photoresist preparation parameters, obtaining the optimal photoresist preparation parameters, and performing photoresist preparation.

[0005] Preferably, the method for preparing a photoresist for a high-resolution display panel further includes: obtaining the ranges of photosensitive dose, resin amount, solvent amount, and stirring parameters during the photoresist preparation process; combining the ranges of at least one of the photosensitive dose, resin amount, solvent amount, and stirring parameters to obtain the range of preparation parameters.

[0006] Preferably, the method for preparing a photoresist for a high-resolution display panel further includes: randomly generating preparation parameters of the photoresist within the range of the preparation parameters to obtain the preparation parameters; inputting the preparation parameters into a pre-trained photoresist preparation predictor, and predicting and outputting to obtain a plurality of photoresist bubble distributions, wherein the photoresist preparation predictor includes a plurality of pre-trained photoresist preparation prediction branches.

[0007] Preferably, the method for preparing a photoresist for a high-resolution display panel further includes: according to the photoresist preparation data within the historical time, collecting a set of sample preparation parameters, and collecting the bubble position distributions of the bubbles detected in the photoresists actually prepared with different sample preparation parameters, and labeling them as a set of sample photoresist bubble distributions, wherein each photoresist bubble distribution includes the bubble coordinates of a plurality of bubbles; combining the set of sample preparation parameters and the set of sample photoresist bubble distributions to obtain a set of sample preparation prediction data; randomly selecting K pieces of data with replacement within the set of sample preparation prediction data to obtain K pieces of preparation prediction training data; based on the generative adversarial network, constructing K photoresist preparation prediction branches, wherein each photoresist preparation prediction branch includes a generative network and an adversarial network; respectively using the K pieces of preparation prediction training data to train and test the K photoresist preparation prediction branches, and after passing the test, combining the K photoresist preparation prediction branches to obtain a photoresist preparation predictor.

[0008] Preferably, the method for preparing a photoresist for a high-resolution display panel further includes: obtaining the etching specification parameters for the current etching, wherein the etching specification parameters include the minimum etching pattern size; according to the etching specification parameters, randomly selecting etching positions within the plurality of photoresist bubble distributions respectively to obtain a plurality of etching position distributions, wherein each etching position distribution includes a plurality of etching position coordinates, and the distance between adjacent etching position coordinates is the minimum etching pattern size; respectively calculating the interval distances between the etching position coordinates and the nearest bubble position coordinates within the plurality of etching position distributions and the plurality of photoresist bubble distributions, and calculating the proportion of the etching position coordinates with the interval distance less than a preset distance threshold to obtain a plurality of etching blur rates.

[0009] Preferably, the method for preparing a photoresist for a high-resolution display panel further includes: constructing a resolution evaluation function, and based on the resolution evaluation function, calculating and obtaining the resolution fitness according to the plurality of photoresist bubble distributions and the plurality of etching blur rates; continuing to randomly generate preparation parameters within the range of the preparation parameters, and calculating the resolution fitness, and performing iterative optimization of the photoresist preparation parameters until optimization convergence; outputting the preparation parameters with the maximum resolution fitness to obtain the optimal photoresist preparation parameters, and performing photoresist preparation.

[0010] Preferably, the method for preparing a photoresist for a high-resolution display panel further includes: constructing a resolution evaluation function as follows: ; where RESfit is the resolution fitness, K is the number of distributions of photoresist bubbles, and are weights, is the preset number of bubbles, is a small real number, is the number of bubbles in the i-th distribution of photoresist bubbles, is the preset etching blur rate, is the i-th etching blur rate; obtaining the number of bubbles in the multiple distributions of photoresist bubbles, combining the multiple etching blur rates, and calculating the resolution fitness based on the resolution evaluation function.

[0011] In a second aspect, the present invention also provides a system for preparing a photoresist for a high-resolution display panel, which is used to execute the method for preparing a photoresist for a high-resolution display panel as described in the first aspect, and includes: a preparation parameter range acquisition module, which is used to acquire the range of preparation parameters during the preparation of the photoresist, where the preparation parameters include at least one of photosensitive dose, resin amount, solvent amount, and stirring parameters; a photoresist preparation prediction module, which is used to randomly generate the preparation parameters of the photoresist within the range of the preparation parameters, perform integrated prediction of photoresist preparation, and obtain multiple distributions of photoresist bubbles; a random etching simulation module, which is used to randomly perform etching simulation within the multiple distributions of photoresist bubbles respectively to obtain multiple etching blur rates; a preparation parameter optimization module, which is used to calculate the lithography fitness according to the multiple distributions of photoresist bubbles and the multiple etching blur rates, perform iterative optimization of the preparation parameters of the photoresist, obtain the optimal preparation parameters of the photoresist, and perform the preparation of the photoresist.

[0012] The embodiments of the present invention have the following advantages: By acquiring the range of preparation parameters during the preparation of the photoresist, where the preparation parameters include at least one of photosensitive dose, resin amount, solvent amount, and stirring parameters; then randomly generating the preparation parameters of the photoresist within the range of the preparation parameters, performing integrated prediction of photoresist preparation, and obtaining multiple distributions of photoresist bubbles; then randomly performing etching simulation within the multiple distributions of photoresist bubbles respectively to obtain multiple etching blur rates; further calculating the lithography fitness according to the multiple distributions of photoresist bubbles and the multiple etching blur rates, performing iterative optimization of the preparation parameters of the photoresist, and obtaining the optimal preparation parameters of the photoresist; finally, preparing the photoresist according to the optimal preparation parameters of the photoresist. That is to say, through integrated prediction and algorithm optimization, the accuracy and reliability of the setting of the preparation parameters of the photoresist can be significantly improved, thereby effectively controlling the generation and distribution of bubbles, reducing their impact on the lithography pattern, and thus improving the accuracy, clarity, and details of the pattern. Brief Description of the Drawings

[0013] Figure 1 is a flowchart of the steps of a method for preparing a photoresist for a high-resolution display panel according to the present invention; Figure 2 is a schematic structural diagram of a photoresist preparation system for a high-resolution display panel according to the present invention.

[0014] Description of the Reference Numerals: a preparation parameter range acquisition module 11, a photoresist preparation prediction module 12, a random etching simulation module 13, and a preparation parameter optimization module 14. Detailed Embodiments

[0015] By providing a method and system for preparing a photoresist for a high-resolution display panel, the present invention solves the technical problem that the traditional photoresist preparation method cannot effectively control the bubble interference, resulting in the influence on the accuracy and details of the lithography pattern. By integrating prediction and algorithm optimization, the accuracy and reliability of the photoresist preparation parameter setting can be significantly improved, thereby effectively controlling the generation and distribution of bubbles, reducing their influence on the lithography pattern, and thus improving the accuracy, clarity, and details of the pattern.

[0016] Next, the technical solutions in the present invention will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments of the present invention. It should be understood that the present invention is not limited by the example embodiments described herein. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention. Additionally, it should be noted that for the sake of description, only the parts related to the present invention are shown in the accompanying drawings rather than all of them.

[0017] Example 1, please refer to the attached Figure 1 , the present invention provides a method for preparing a photoresist for a high-resolution display panel, which is applied to a photoresist preparation system for a high-resolution display panel, and specifically includes the following steps: S10: Obtain the range of preparation parameters during the photoresist preparation process, where the preparation parameters include at least one of photosensitive dose, resin amount, solvent amount, and stirring parameters.

[0018] Furthermore, step S10 of the present invention further includes: S11: Obtain the ranges of photosensitive dose, resin amount, solvent amount, and stirring parameters during the photoresist preparation process; S12: Combine the ranges of at least one of the photosensitive dose, resin amount, solvent amount, and stirring parameters to obtain the range of preparation parameters.

[0019] Specifically, first, obtain the ranges of photosensitive dose, resin amount, solvent amount, and stirring parameters during the preparation of photoresist. Among them, the photosensitive dose refers to the concentration of photosensitizer in the photoresist, which determines the responsiveness of the photoresist to light during exposure. Excessive photosensitizer may cause the photoresist to overreact, resulting in blurred patterns or uneven photoresist, while too little photosensitizer may lead to incomplete exposure of the photoresist. Generally, the range of photosensitive dose is 0.5% to 10% (by the proportion of photosensitizer in the total amount of photoresist); the resin amount affects the curing performance and fluidity of the photoresist. Too high a concentration of resin will make the photoresist too viscous, resulting in uneven coating; too low a resin amount may make the photoresist too thin, affecting the stability and resolution of the pattern. Generally, the range of resin amount is 20% to 50% (by the proportion of resin in the total amount of photoresist); the solvent amount determines the fluidity and coating uniformity of the photoresist. The use of solvent not only helps to dissolve the resin and photosensitizer, but also can adjust the viscosity of the photoresist. Too much solvent will increase the volatility of the photoresist, which may lead to the formation of bubbles; too little solvent may cause the photoresist to be too thick and difficult to coat evenly. Generally, the range of solvent amount is 30% to 60% (by the proportion of solvent in the total amount of photoresist). Stirring is a key step in controlling bubble generation during the preparation of photoresist. Ultrasonic stirring is usually used to reduce the generation of large bubbles, but insufficient stirring will cause bubbles to remain, while excessive stirring may generate too many tiny bubbles, affecting the uniformity and performance of the photoresist. The range of stirring parameters includes the frequency range (usually in the range of 20 kHz to 40 kHz), the power range (generally between 100 W and 500 W), and the time (generally controlled within 5 to 30 minutes).

[0020] Next, combine the ranges of at least one of the photosensitive dose, resin amount, solvent amount, and stirring parameters, that is, at least select one of the photosensitive dose, resin amount, solvent amount, and stirring parameters as the preparation parameters. The preparation parameters can be set according to the actual preparation scenario to obtain the range of preparation parameters. For example, the preparation parameters can include the photosensitive dose and resin amount, or can also include the photosensitive dose, resin amount, and solvent amount.

[0021] S20: Randomly generate the preparation parameters of the photoresist within the range of the preparation parameters, perform integrated prediction of photoresist preparation, and obtain multiple photoresist bubble distributions.

[0022] Furthermore, step S20 of the present invention further includes: S21: Randomly generate the preparation parameters of the photoresist within the range of the preparation parameters to obtain the preparation parameters.

[0023] Specifically, any preparation parameter is randomly selected within the range of the preparation parameters. For example, assuming that the preparation parameters include photosensitive dose and resin amount, where the photosensitive dose ranges from 0.5% to 10% and the resin amount ranges from 20% to 50%, then any preparation parameter can be a photosensitive dose of 3% and a resin amount of 30%.

[0024] S22: Input the preparation parameters into a pre-trained photoresist preparation predictor to predict and obtain multiple photoresist bubble distributions, where the photoresist preparation predictor includes multiple pre-trained photoresist preparation prediction branches.

[0025] Furthermore, step S22 of the present invention further includes: S221: According to the photoresist preparation data within the historical time, collect a set of sample preparation parameters, and collect the bubble position distributions of the bubbles detected in the photoresist actually prepared with different sample preparation parameters, and label them as a set of sample photoresist bubble distributions, where each photoresist bubble distribution includes the bubble coordinates of multiple bubbles; S222: Combine the set of sample preparation parameters and the set of sample photoresist bubble distributions to obtain a set of sample preparation prediction data; S223: Randomly select K pieces of data with replacement within the set of sample preparation prediction data to obtain K pieces of preparation prediction training data; S224: Based on the generative adversarial network, construct K photoresist preparation prediction branches, where each photoresist preparation prediction branch includes a generative network and an adversarial network; S225: Respectively use the K pieces of preparation prediction training data to train and test the K photoresist preparation prediction branches. After passing the test, combine the K photoresist preparation prediction branches to obtain a photoresist preparation predictor.

[0026] Specifically, first, according to the photoresist preparation data within the historical time (such as within the most recent three months), collect the key preparation parameters in the photoresist preparation process. These data come from historical experiments or production processes to obtain multiple sample preparation parameters to construct a set of sample preparation parameters. Then, collect the bubble position distributions of the bubbles detected in the photoresist actually prepared with different sample preparation parameters, such as using technologies such as microscopes and scanning electron microscopes to detect the bubbles in the photoresist. Among them, the bubbles will be marked with different coordinate points (for example, positions on the X and Y coordinate systems), and the specific position of each bubble will form a data point. Record the bubble positions of each photoresist sample detected to form a bubble position distribution, and label it as a sample photoresist bubble distribution to construct a set of sample photoresist bubble distributions.

[0027] Next, combine the sample preparation parameters and the sample photoresist bubble distributions that have a corresponding relationship in the set of sample preparation parameters and the set of sample photoresist bubble distributions as a set of sample preparation prediction data, and obtain multiple sets of sample preparation prediction data to construct a sample preparation prediction data set. Then divide the sample preparation prediction data set into K equal parts to obtain K sample data sets, and iteratively select K times from the K sample data sets to construct the first preparation prediction training data, and use the same method to iteratively select K times to obtain K preparation prediction training data, where K is an integer greater than or equal to 10.

[0028] A generative adversarial network is a deep learning model composed of two parts, a generator and a discriminator. These two parts are jointly trained in an adversarial manner to achieve the goal of generating high-quality data. Then, based on the generative adversarial network, construct K photoresist preparation prediction branches. Each photoresist preparation prediction branch includes a generation network and an adversarial network. The generation network is used to generate fake sample data (e.g., the bubble distribution of the photoresist) to deceive the adversarial network; the adversarial network is used to determine whether the sample is "real" or "generated", that is, to distinguish between the real photoresist bubble distribution and the generated fake bubble distribution; these two parts interact through adversarial training, and finally the generation network can generate more and more real fake samples.

[0029] Next, use the K sets of preparation prediction training data to train and test the K photoresist preparation prediction branches respectively. During the training process, the generator receives the photoresist preparation parameters as input and generates the predicted photoresist bubble distribution; the discriminator receives the generated bubble distribution and the real bubble distribution, and calculates their authenticity probabilities respectively. The goal of the generator is to maximize the discriminator's judgment probability for fake data, so that the generated data can be considered "real" by the discriminator. The task of the discriminator is to judge the gap between the generated bubble distribution and the real data. After each forward propagation, by calculating the loss function of the discriminator, execute the backpropagation algorithm to update the weights of the discriminator; after the discriminator is updated, train the generator, and by calculating the loss function of the generator, execute backpropagation to update the weights of the generator. During the testing process, use a data set (test set) different from the training set, input the photoresist preparation parameters, and generate the bubble distribution through the trained photoresist preparation prediction branch; then, compare the bubble distribution generated by the generator with the real bubble distribution to evaluate the prediction accuracy of the model; the discriminator evaluates the authenticity of the generated bubbles and outputs its judgment result; if the generated bubble distribution cannot be distinguished as real or fake by the discriminator, the performance of the generator is better; if the output of the generator meets the set accuracy standard and the discriminator successfully distinguishes between real and fake data, then this branch is considered to pass the test. At this time, obtain K photoresist preparation prediction branches that pass the test, and integrate them to construct a photoresist preparation predictor.

[0030] By training K branches and independently training and testing each branch, the generalization ability of the model can be improved, and the integrated method is used to combine each branch to finally obtain a photoresist preparation predictor with strong robustness and high accuracy.

[0031] Finally, the preparation parameters are input into the pre-trained photoresist preparation predictor. Through forward propagation calculation, the predictor will generate corresponding photoresist bubble distributions according to the input preparation parameters and output multiple photoresist bubble distributions.

[0032] S30: Randomly perform etching simulations within the multiple photoresist bubble distributions to obtain multiple etching blur rates.

[0033] Furthermore, step S30 of the present invention further includes: S31: Obtain the etching specification parameters for the current etching, where the etching specification parameters include the minimum etching pattern size; S32: According to the etching specification parameters, randomly select etching positions within the multiple photoresist bubble distributions to obtain multiple etching position distributions, where each etching position distribution includes multiple etching position coordinates, and the distance between adjacent etching position coordinates is the minimum etching pattern size; S33: Calculate the interval distances between the etching position coordinates and the nearest bubble position coordinates within the multiple etching position distributions and the multiple photoresist bubble distributions, and calculate the proportion of the etching position coordinates with the interval distance less than the preset distance threshold to obtain multiple etching blur rates.

[0034] Specifically, first, obtain the etching specification parameters for the current etching, where the etching specification parameters include the minimum etching pattern size, and the minimum etching pattern size determines the size of the smallest pattern for each etching. For example, assume the minimum etching pattern size is 50 nm. Then, for each photoresist bubble distribution, randomly select an etching position in the bubble distribution and ensure that the distance between adjacent etching positions meets the minimum etching pattern size. For example, randomly select an etching position from the bubble distribution area, and then when selecting a new position, ensure that the distance from the previously selected etching position is at least the minimum etching pattern size to obtain multiple etching position distributions, where each etching position distribution includes multiple etching position coordinates.

[0035] Then, within the distributions of the multiple etching positions and the multiple photoresist bubble distributions respectively, calculate the interval distance between the etching position coordinates and the coordinates of the nearest bubble position, that is, for each etching position, find the bubble position closest to this etching position and calculate its distance; then count the proportion of the etching position coordinates with an interval distance less than a preset distance threshold (which can be set according to the standard preparation requirements, such as 100 nm). If the distance is less than this threshold, it is considered that this etching position is interfered by bubbles, resulting in a blurred pattern, and set the proportion of the etching positions affected by bubbles as the etching blur rate, obtaining multiple etching blur rates. By calculating the distance between each etching position and the bubble position and determining whether it is interfered by bubbles (the distance is less than the preset threshold), the blur rate of each etching position distribution can be calculated. This process helps to analyze and optimize the preparation process of the photoresist to improve the accuracy and details of the lithography pattern.

[0036] S40: According to the multiple photoresist bubble distributions and the multiple etching blur rates, calculate and obtain the lithography fitness, perform iterative optimization of the photoresist preparation parameters, obtain the optimal photoresist preparation parameters, and carry out photoresist preparation.

[0037] Furthermore, step S40 of the present invention further includes: S41: Construct a resolution evaluation function, and based on the multiple photoresist bubble distributions and the multiple etching blur rates, calculate and obtain the resolution fitness according to the resolution evaluation function.

[0038] Furthermore, step S41 of the present invention further includes: S411: Construct a resolution evaluation function as follows: ; where RESfit is the resolution fitness, K is the number of multiple photoresist bubble distributions, and are weights, is the preset number of bubbles, is a small real number, is the number of bubbles in the i-th photoresist bubble distribution, is the preset etching blur rate, is the i-th etching blur rate; S412: Obtain the multiple numbers of bubbles in the multiple photoresist bubble distributions, combine the multiple etching blur rates, and calculate and obtain the resolution fitness based on the resolution evaluation function.

[0039] Specifically, in the resolution evaluation function, RESfit is the resolution fitness. Among them, the larger the resolution fitness, the better the characterization of the preparation parameters; K is the number of multiple photoresist bubble distributions, and are weights, where is the bubble number weight, is the etching blur rate weight, and The sum of them is 1, and the weight value can be set according to the influence degree of the index on the resolution fitness. The greater the influence degree, the greater the weight; is the preset number of bubbles, is a small real number, which can be set according to actual requirements. For example, 10; is the number of bubbles in the i-th photoresist bubble distribution, is the preset etching blur rate, usually an ideal small value, representing the ideal situation without bubble interference, and can be set according to the preparation requirements, is the i-th etching blur rate. By constructing a resolution evaluation function, the influence of the bubble distribution and the etching blur rate on the resolution during the photoresist preparation process can be accurately quantified, thereby providing a specific numerical basis for optimizing the preparation parameters.

[0040] Obtain the multiple numbers of bubbles in the multiple photoresist bubble distributions, and then use the resolution evaluation function to calculate the resolution fitness according to the multiple numbers of bubbles and the multiple etching blur rates, which reflects the comprehensive influence of the number of bubbles and the etching blur rate on the pattern clarity.

[0041] S42: Continue to randomly generate preparation parameters within the range of the preparation parameters, calculate the resolution fitness, and perform iterative optimization of the photoresist preparation parameters until optimization convergence; S43: Output the preparation parameters with the maximum resolution fitness, obtain the optimal photoresist preparation parameters, and perform photoresist preparation.

[0042] Specifically, then randomly generate other preparation parameters within the range of the preparation parameters and calculate the resolution fitness; use the same method for iterative selection of the preparation parameters and fitness calculation until the predetermined number of iterations is reached. For example, set the number of iterations to 500 times. At this time, multiple preparation parameters and multiple resolution fitnesses are obtained. Finally, select the preparation parameters with the maximum resolution fitness as the optimal photoresist preparation parameters and perform photoresist preparation. By performing iterative optimization within the range of the preparation parameters to find the preparation parameters that make the pattern accuracy and clarity of the photoresist reach the best effect, the quality of photoresist preparation can be effectively improved.

[0043] In summary, a method for preparing a photoresist for a high-resolution display panel provided by the present invention has the following technical effects: By obtaining the range of preparation parameters in the photoresist preparation process, where the preparation parameters include at least one of photosensitive dose, resin amount, solvent amount, and stirring parameters; then randomly generating the preparation parameters of the photoresist within the range of the preparation parameters, performing integrated prediction of photoresist preparation to obtain multiple photoresist bubble distributions; then randomly performing etching simulation within each of the multiple photoresist bubble distributions to obtain multiple etching blur rates; further calculating the photolithography fitness based on the multiple photoresist bubble distributions and multiple etching blur rates, performing iterative optimization of the photoresist preparation parameters to obtain the optimal photoresist preparation parameters; and finally preparing the photoresist according to the optimal photoresist preparation parameters. That is to say, through integrated prediction and algorithm optimization, the accuracy and reliability of the setting of photoresist preparation parameters can be significantly improved, thereby effectively controlling the generation and distribution of bubbles, reducing their impact on the photolithography pattern, and thus improving the accuracy, clarity, and details of the pattern.

[0044] Embodiment 2. Based on the same inventive concept as the method for preparing a photoresist for a high-resolution display panel in the foregoing embodiment, the present invention further provides a system for preparing a photoresist for a high-resolution display panel. Please refer to the attached Figure 2 , including: a preparation parameter range acquisition module 11 for obtaining the range of preparation parameters in the photoresist preparation process, where the preparation parameters include at least one of photosensitive dose, resin amount, solvent amount, and stirring parameters; a photoresist preparation prediction module 12 for randomly generating the preparation parameters of the photoresist within the range of the preparation parameters, performing integrated prediction of photoresist preparation to obtain multiple photoresist bubble distributions; a random etching simulation module 13 for randomly performing etching simulation within each of the multiple photoresist bubble distributions to obtain multiple etching blur rates; a preparation parameter optimization module 14 for calculating the photolithography fitness based on the multiple photoresist bubble distributions and multiple etching blur rates, performing iterative optimization of the photoresist preparation parameters to obtain the optimal photoresist preparation parameters, and preparing the photoresist.

[0045] Further, the system for preparing a photoresist for a high-resolution display panel is further configured to: obtain the ranges of photosensitive dose, resin amount, solvent amount, and stirring parameters in the photoresist preparation process; combine the ranges of at least one of the photosensitive dose, resin amount, solvent amount, and stirring parameters to obtain the range of preparation parameters.

[0046] Further, the system for preparing a photoresist for a high-resolution display panel is further configured to: randomly generate the preparation parameters of the photoresist within the range of the preparation parameters to obtain the preparation parameters; input the preparation parameters into a pre-trained photoresist preparation predictor, and predict and output to obtain multiple photoresist bubble distributions, where the photoresist preparation predictor includes multiple pre-trained photoresist preparation prediction branches.

[0047] Further, the photoresist preparation system for a high-resolution display panel is further configured to: collect a set of sample preparation parameters according to the photoresist preparation data within a historical time period, and collect the bubble position distributions of the photoresists actually prepared with different sample preparation parameters for bubble detection, which are marked as a set of sample photoresist bubble distributions, where each photoresist bubble distribution includes the bubble coordinates of multiple bubbles; combine the set of sample preparation parameters and the set of sample photoresist bubble distributions to obtain a set of sample preparation prediction data; randomly select K pieces of data with replacement from the set of sample preparation prediction data to obtain K pieces of preparation prediction training data; based on a generative adversarial network, construct K photoresist preparation prediction branches, where each photoresist preparation prediction branch includes a generative network and an adversarial network; respectively use the K pieces of preparation prediction training data to train and test the K photoresist preparation prediction branches, and after passing the test, combine the K photoresist preparation prediction branches to obtain a photoresist preparation predictor.

[0048] Further, the photoresist preparation system for a high-resolution display panel is further configured to: obtain the etching specification parameters for the current etching, where the etching specification parameters include the minimum etching pattern size; according to the etching specification parameters, randomly select etching positions in the multiple photoresist bubble distributions respectively to obtain multiple etching position distributions, where each etching position distribution includes multiple etching position coordinates, and the distance between adjacent etching position coordinates is the minimum etching pattern size; respectively calculate the interval distances between the etching position coordinates and the nearest bubble position coordinates in the multiple etching position distributions and the multiple photoresist bubble distributions, and calculate the proportion of the etching position coordinates with the interval distance less than a preset distance threshold to obtain multiple etching blur rates.

[0049] Further, the photoresist preparation system for a high-resolution display panel is further configured to: construct a resolution evaluation function, and based on the multiple photoresist bubble distributions and multiple etching blur rates, calculate and obtain a resolution fitness according to the resolution evaluation function; continue to randomly generate preparation parameters within the range of the preparation parameters, calculate the resolution fitness, and perform iterative optimization of the photoresist preparation parameters until optimization convergence; output the preparation parameters with the maximum resolution fitness to obtain the optimal photoresist preparation parameters, and perform photoresist preparation.

[0050] Further, the photoresist preparation system for a high-resolution display panel is further configured to: construct a resolution evaluation function as follows: ; where RESfit is the resolution fitness, K is the number of multiple photoresist bubble distributions, and are weights, is the preset number of bubbles, is a small real number, is the number of bubbles within the i-th photoresist bubble distribution, is a preset etching blur rate, is the i-th etching blur rate; Obtain the multiple numbers of bubbles within the multiple photoresist bubble distributions, combine the multiple etching blur rates, and calculate and obtain a resolution fitness based on the resolution evaluation function.

[0051] In the present specification, each embodiment is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. The method and specific examples for preparing a photoresist for a high-resolution display panel in the foregoing Embodiment 1 are equally applicable to the system for preparing a photoresist for a high-resolution display panel in this embodiment. Through the foregoing detailed description of the method for preparing a photoresist for a high-resolution display panel, those skilled in the art can clearly understand the system for preparing a photoresist for a high-resolution display panel in this embodiment. Therefore, for the sake of simplicity of the specification, it will not be elaborated herein. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description in the method part.

[0052] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0053] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalent technologies, the present invention also intends to include these changes and modifications.

Claims

1. A method for preparing a photoresist for a high-resolution display panel, characterized in that: Methods include: Obtaining a range of preparation parameters in a photoresist preparation process, wherein the preparation parameters include at least one of a photosensitizer dosage, a resin dosage, a solvent dosage, and a stirring parameter; Randomly generating photoresist preparation parameters within the preparation parameter range, performing integrated prediction of photoresist preparation, and obtaining a plurality of photoresist bubble distributions; Randomly performing etching simulations in the plurality of photoresist bubble distributions to obtain a plurality of etching blur ratios; According to the multiple photoresist bubble distributions and multiple etching blur ratios, the photolithography adaptability is calculated, and the photoresist preparation parameters are iteratively optimized to obtain the optimal photoresist preparation parameters and perform photoresist preparation.

2. The method for preparing a high-resolution photoresist for a display panel according to claim 1, characterized in that: Obtain the range of preparation parameters during photoresist preparation, including: Obtain the range of photosensitizer dosage, resin amount, solvent amount and stirring parameters during photoresist preparation; The range of at least one parameter among the amount of photosensitizer, the amount of resin, the amount of solvent and the stirring parameter is combined to obtain the preparation parameter range.

3. The method for preparing a high-resolution photoresist for a display panel according to claim 1, characterized in that: The preparation parameters of the photoresist are randomly generated within the preparation parameter range, and the integrated prediction of the photoresist preparation is performed, including: Randomly generating preparation parameters of the photoresist within the preparation parameter range to obtain preparation parameters; The preparation parameters are input into a pre-trained photoresist preparation predictor, and a plurality of photoresist bubble distributions are obtained by prediction output, wherein the photoresist preparation predictor includes a plurality of pre-trained photoresist preparation prediction branches.

4. The method for preparing a high-resolution photoresist for a display panel according to claim 3, characterized in that: The pre-training step of the photoresist preparation predictor comprises: According to the photoresist preparation data in the historical time, a sample preparation parameter set is collected, and the bubble position distribution of the photoresist actually prepared with different sample preparation parameters for bubble detection is collected, and marked as a sample photoresist bubble distribution set, wherein each photoresist bubble distribution includes bubble coordinates of multiple bubbles; Combining the sample preparation parameter set and the sample photoresist bubble distribution set to obtain a sample preparation prediction data set; Randomly select K pieces of data with replacement from the sample preparation prediction data set to obtain K pieces of preparation prediction training data; Based on the generative adversarial network, K photoresist preparation prediction branches are constructed, wherein each photoresist preparation prediction branch includes a generative network and an adversarial network; The K pieces of preparation prediction training data are respectively used to train and test the K photoresist preparation prediction branches. After the test is qualified, the K photoresist preparation prediction branches are combined to obtain a photoresist preparation predictor.

5. The method for preparing a high-resolution photoresist for a display panel according to claim 1, characterized in that: Etching simulation is randomly performed in the plurality of photoresist bubble distributions to obtain a plurality of etching blur ratios, including: Obtaining etching specification parameters of the current etching, wherein the etching specification parameters include a minimum etching pattern size; According to the etching specification parameters, randomly selecting etching positions in the plurality of photoresist bubble distributions to obtain a plurality of etching position distributions, wherein each etching position distribution includes a plurality of etching position coordinates, and the distance between adjacent etching position coordinates is the minimum etching pattern size; In the multiple etching position distributions and the multiple photoresist bubble distributions, the interval distances between the etching position coordinates and the nearest bubble position coordinates are calculated respectively, and the proportion of the etching position coordinates whose interval distances are less than a preset distance threshold is calculated to obtain multiple etching blur rates.

6. The method for preparing a high-resolution photoresist for a display panel according to claim 1, characterized in that: According to the plurality of photoresist bubble distributions and the plurality of etching blur ratios, photolithography adaptability is calculated, photoresist preparation parameters are iteratively optimized, optimal photoresist preparation parameters are obtained, and photoresist preparation is performed, including: Constructing a resolution evaluation function, and calculating and obtaining a resolution fitness based on the resolution evaluation function according to the plurality of photoresist bubble distributions and the plurality of etching blur rates; Continue to randomly generate preparation parameters within the preparation parameter range, calculate resolution fitness, and perform iterative optimization of the photoresist preparation parameters until optimization converges; Output the preparation parameters with the greatest resolution adaptability, obtain the optimal photoresist preparation parameters, and perform photoresist preparation.

7. The method for preparing a high-resolution photoresist for a display panel according to claim 6, characterized in that: Constructing a resolution evaluation function, and calculating and obtaining resolution fitness based on the resolution evaluation function according to the plurality of photoresist bubble distributions and the plurality of etching blur rates, including: Construct the resolution evaluation function as follows: ； Among them, RESfit is the resolution fitness, K is the number of multiple photoresist bubble distributions, and is the weight, To preset the number of bubbles, is a small real number, is the number of bubbles in the ith photoresist bubble distribution, To preset the etching blur rate, is the i-th etching blur rate; The number of multiple bubbles in the multiple photoresist bubble distributions is obtained, and the resolution adaptability is calculated based on the resolution evaluation function in combination with the multiple etching blur ratios.

8. A high-resolution display panel photoresist preparation system, characterized in that: The steps for implementing the method for preparing a high-resolution photoresist for a display panel according to any one of claims 1 to 7 include: A preparation parameter range acquisition module is used to acquire the preparation parameter range in the photoresist preparation process, wherein the preparation parameter includes at least one of a photosensitizer dose, a resin amount, a solvent amount and a stirring parameter; A photoresist preparation prediction module, used to randomly generate photoresist preparation parameters within the preparation parameter range, perform integrated prediction of photoresist preparation, and obtain multiple photoresist bubble distributions; A random etching simulation module, used for randomly performing etching simulations in the plurality of photoresist bubble distributions to obtain a plurality of etching blur rates; The preparation parameter optimization module is used to calculate the photolithography fitness according to the multiple photoresist bubble distributions and multiple etching blur ratios, perform iterative optimization of the photoresist preparation parameters, obtain the optimal photoresist preparation parameters, and perform photoresist preparation.

Images