TFT gluing process parameter optimization method, device, equipment and storage medium

Abstract

The application discloses a TFT gluing process parameter optimization method and device, equipment and a storage medium. When a preset trigger event is detected, the gluing process parameters of multiple batches of TFTs and the TFT performance indexes of each batch are acquired. The performance indexes include threshold voltage uniformity. Based on the gluing process parameters and the TFT performance indexes of each batch, multiple sets of process parameter combinations are constructed. Each set of process parameter combinations is input into a TFT performance prediction model to obtain the threshold voltage uniformity and the confidence thereof of each set of process parameter combinations. Based on the threshold voltage uniformity and the confidence thereof corresponding to each set of process parameter combinations, the optimal process parameter combination is screened. The application does not require manual analysis of historical data of gluing process parameters, avoids the deviation caused by manual experience and subjective judgment, effectively solves the technical problems of low manual analysis efficiency, strong subjectivity and difficulty in accurately responding to multi-parameter coupling optimization in the prior art, and can efficiently, accurately and reliably optimize process parameters.

Description

Technical Field

[0001] This invention relates to the field of semiconductor processing technology, and in particular to a method, apparatus, equipment, and storage medium for optimizing TFT coating process parameters. Background Technology

[0002] In the current research and development and mass production control of oxide TFT coating processes, the range of values ​​for key process parameters such as photoresist thickness, coating speed, exposure energy density, and curing time directly affects the uniformity of TFT threshold voltage.

[0003] Currently, after each batch of TFT production is completed, technicians collect the process parameters used in production, as well as the performance indicators of the TFTs produced in that batch. Over time, technicians analyze the collected data to determine the photoresist coating process parameters that can achieve good performance indicators for the TFTs, thereby continuously optimizing the coating process parameters.

[0004] However, relying on technicians to manually analyze the collected data has two drawbacks. First, the data volume is enormous and the time required is long, making it impossible to update the optimal coating process parameters in a timely manner. Second, manual analysis is greatly affected by the experience and subjective judgment of technicians, making it difficult to accurately locate the optimal combination under the coupling of multiple parameters. Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, and storage medium for optimizing TFT coating process parameters, thereby achieving automatic optimization of coating process parameters and improving the efficiency and accuracy of process parameter optimization.

[0006] In a first aspect, the present invention provides a method for optimizing TFT coating process parameters, including: When a preset trigger event is detected, the coating process parameters of multiple batches of TFTs produced within a preset time window and the performance indicators of each batch of TFTs are obtained. The coating process parameters include photoresist thickness, coating speed, exposure energy density and curing time. The performance indicators include threshold voltage uniformity. Multiple sets of process parameter combinations are constructed based on the coating process parameters and TFT performance indicators used in each batch; Each set of process parameter combinations is input into a pre-trained TFT performance prediction model to obtain the threshold voltage uniformity and its confidence level corresponding to each set of process parameter combinations. The optimal combination of process parameters is selected based on the threshold voltage uniformity and confidence level of each combination.

[0007] Secondly, the present invention provides a TFT coating process parameter optimization device, comprising: The process parameter and performance index acquisition module is used to acquire the photoresist coating process parameters and the TFT performance index of each batch of TFTs produced within a preset time window when a preset trigger event is detected. The photoresist coating process parameters include photoresist thickness, coating speed, exposure energy density and curing time, and the performance index includes threshold voltage uniformity. The process parameter combination construction module is used to construct multiple sets of process parameter combinations based on the coating process parameters and TFT performance indicators used in each batch. The performance prediction module is used to input each combination of process parameters into a pre-trained TFT performance prediction model to obtain the threshold voltage uniformity and its confidence level corresponding to each combination of process parameters. The process parameter combination screening module is used to screen the optimal process parameter combination based on the threshold voltage uniformity and confidence level of each process parameter combination.

[0008] Thirdly, the present invention provides an electronic device, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the TFT coating process parameter optimization method according to the first aspect of the present invention.

[0009] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the TFT coating process parameter optimization method described in the first aspect of the present invention.

[0010] This invention obtains the coating process parameters and performance indicators of multiple batches of TFTs within a preset time window by triggering a preset event. Based on this, multiple sets of process parameter combinations are constructed and input into the TFT performance prediction model to obtain the threshold voltage uniformity and confidence level corresponding to each set of process parameter combinations. Furthermore, the optimal process parameter combination is selected based on the threshold voltage uniformity and confidence level corresponding to each set of process parameter combinations. This avoids the deviation caused by human experience and subjective judgment, and effectively solves the technical problems of low efficiency, strong subjectivity, and difficulty in accurately dealing with multi-parameter coupling optimization in the prior art. It can optimize process parameters efficiently, accurately, and reliably.

[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart of a TFT coating process parameter optimization method provided in Embodiment 1 of the present invention; Figure 2 This is a flowchart of a TFT coating process parameter optimization method provided in Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of a TFT coating process parameter optimization device provided in Embodiment 3 of the present invention; Figure 4 This is a schematic diagram of the structure of the electronic device provided in Embodiment 4 of the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] Example 1 Figure 1 This is a flowchart of a TFT coating process parameter optimization method provided in Embodiment 1 of the present invention. This embodiment is applicable to the optimization of TFT coating process parameters. The method can be executed by a TFT coating process parameter optimization device, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method for optimizing the TFT coating process parameters includes: S101, when a preset trigger event is detected, acquire the coating process parameters of multiple batches of TFTs produced within a preset time window and the performance indicators of each batch of TFTs.

[0016] The coating process parameters include photoresist thickness, coating speed, exposure energy density, and curing time. Among them, photoresist thickness refers to the thickness of the photoresist film layer coated on the surface of the TFT substrate using methods such as spin coating and slot coating. When the photoresist thickness deviates from the optimal range, if it is too thin, the substrate surface may be exposed too early during the etching process. If it is too thick, it may cause insufficient exposure energy, resulting in incomplete development or a decrease in pattern resolution, which in turn affects the morphological consistency and threshold voltage uniformity of the TFT channel region.

[0017] Coating speed refers to the moving speed of the coating nozzle or squeegee relative to the TFT substrate during the coating process. Coating speed is one of the key parameters that determines the uniformity of photoresist film thickness and the flatness of the coating surface. Too fast a speed will result in a thinner film and poorer uniformity, while too slow a speed will result in a thicker film and may cause edge accumulation effect. Coating speed significantly affects the uniformity of threshold voltage between TFT devices by affecting the consistency of the film thickness distribution.

[0018] Exposure energy density refers to the amount of ultraviolet light energy received per unit area of ​​substrate surface during the photolithography process. Too low an exposure energy density will result in underexposure of the pattern and excessively wide linewidth, while too high an exposure energy density may cause overexposure, excessively narrow linewidth, or deterioration of the film performance, thereby affecting the consistency of TFT channel pattern size and the uniformity of threshold voltage.

[0019] Curing time refers to the duration of heating and baking the photoresist film after exposure and development (i.e., hardening treatment). Insufficient curing time will result in insufficient film hardness and reduced etching resistance; excessive curing time may lead to over-hardening of the film, increased brittleness, or even thermal deformation, affecting the reliability of TFT devices and the uniformity of threshold voltage.

[0020] Performance metrics include threshold voltage uniformity, which refers to the difference in threshold voltage (V) between different TFT devices within the same batch. th The uniformity of threshold voltage indicates the degree of consistency among TFTs in the same batch. Better threshold voltage uniformity signifies smaller differences in electrical performance within the same batch, resulting in higher brightness uniformity and image quality for the display panel. In this embodiment, threshold voltage uniformity is quantified by statistical values ​​(such as standard deviation or range) of the threshold voltages of each TFT within the batch. A larger value indicates poorer uniformity (i.e., greater fluctuation), while a smaller value indicates better uniformity. For ease of standardized expression, this embodiment defines threshold voltage uniformity as an indicator where a larger value signifies better performance, i.e., using a appropriately converted score value.

[0021] Specifically, for the target TFT product, the trigger condition is set to trigger once every N batches produced, where N is a preset positive integer. The value of N can be set according to the actual production rhythm and optimization requirements, such as N=50, N=100 or N=200. When the number of production batches is detected to be an integer multiple of N, the coating process parameters and threshold voltage uniformity of the M batches of TFTs traced back from the current trigger time are obtained. The value of M should ensure that there is a sufficient sample size for subsequent statistical analysis, such as M=200, M=500 or M=1000.

[0022] Of course, the preset trigger event can also be the process parameter optimization operation triggered by the process engineer through the interactive interface, or the timed optimization task being triggered. This embodiment does not limit the preset trigger event.

[0023] By adopting an event-triggered data acquisition method, rather than continuously acquiring data, excessive consumption of data processing resources and frequent optimizations can be avoided.

[0024] S102, constructs multiple sets of process parameter combinations based on the coating process parameters and TFT performance indicators used in each batch.

[0025] In one embodiment, based on the acquired batch data, the high-contribution value ranges of each coating process parameter can be determined, and candidate process parameter combinations can be constructed using these high-contribution value ranges. Specifically, for each historical batch, firstly, based on whether its threshold voltage uniformity meets the preset standard, each process parameter of that batch is assigned a binary label (1 for meeting the standard, 0 for otherwise). Then, for the four parameters of photoresist thickness, coating speed, exposure energy density, and curing time, value ranges are divided according to preset rules, and the proportion of batches with a label value of 1 in each range is counted as the contribution of that range. The higher the contribution, the greater the probability that the batch falling into that range will produce a high-performance TFT. Further, the n value ranges with the largest contribution for each type of parameter are selected as candidate ranges. Then, one range is extracted from each of the four types of parameter candidate ranges for full combination, generating process parameter combinations that cover multiple parameter coupling situations for subsequent model prediction and optimization screening. The constructed process parameter combinations not only focus on parameter ranges that have performed well in the past, but also cover the coupling relationships between multiple parameters, providing high-quality data for accurately locating the globally optimal process parameter combinations.

[0026] S103, input each set of process parameter combinations into the pre-trained TFT performance prediction model to obtain the threshold voltage uniformity and its confidence level corresponding to each set of process parameter combinations.

[0027] The TFT performance prediction model is a pre-trained machine learning model. The model takes a set of coating process parameters (photoresist thickness, coating speed, exposure energy density, and curing time) as input and the predicted threshold voltage uniformity and its confidence level as output. For each set of process parameters constructed in step S102, the model is input for prediction to obtain the corresponding predicted threshold voltage uniformity value and its confidence level.

[0028] This embodiment uses a pre-trained TFT performance prediction model to quickly and in batches predict the performance of a large number of candidate process parameter combinations. The expected effect of each combination can be evaluated without actual production testing, thus improving the efficiency of process parameter optimization.

[0029] S104. The optimal combination of process parameters is selected based on the threshold voltage uniformity and confidence level of each combination of process parameters.

[0030] Specifically, the combinations are first sorted in descending order according to the predicted threshold voltage uniformity, and the top M combinations are taken as the initial screening results. Then, combinations with confidence levels below the preset threshold are removed from the M combinations, and J combinations with high prediction reliability are retained. Finally, for the J combinations, the product or weighted sum of the predicted threshold voltage uniformity and confidence level of each combination is calculated as a comprehensive score. The combination with the highest score is determined as the optimal coating process parameters, which takes into account both prediction performance and prediction reliability, and ensures that the selected combinations have both good expected performance and high prediction confidence.

[0031] This invention acquires the coating process parameters and performance indicators of multiple batches of TFTs within a preset time window by triggering a preset event. Based on this, multiple sets of process parameter combinations are constructed and input into the TFT performance prediction model to obtain the threshold voltage uniformity and confidence level corresponding to each set of process parameter combinations. Furthermore, the optimal process parameter combination is selected based on the threshold voltage uniformity and confidence level corresponding to each set of process parameter combinations. This eliminates the need for manual analysis of historical data of coating process parameters, avoids the bias caused by human experience and subjective judgment, and effectively solves the technical problems of low efficiency, strong subjectivity, and difficulty in accurately dealing with multi-parameter coupling optimization in the prior art. It can optimize process parameters efficiently, accurately, and reliably.

[0032] Example 2 Figure 2 This is a flowchart of a TFT coating process parameter optimization method provided in Embodiment 2 of the present invention. This embodiment optimizes the method based on Embodiment 1 described above. Figure 2 As shown, the method for optimizing the parameters of the TFT coating process includes: S201, when a preset trigger event is detected, acquire the coating process parameters of multiple batches of TFTs produced within a preset time window and the performance indicators of each batch of TFTs.

[0033] The specific implementation of step S201 in this embodiment is the same as that of step S101 in embodiment one, and will not be repeated here.

[0034] S202, for each batch, if the threshold voltage uniformity of the TFT is greater than or equal to the preset threshold, set the label value of each coating process parameter to 1; otherwise, set it to 0.

[0035] In one embodiment, the preset threshold is a threshold voltage uniformity standard pre-set according to product specifications. For each batch, it is determined whether the threshold voltage uniformity of the batch reaches or exceeds the preset threshold. If it reaches or exceeds the preset threshold, the tag values ​​of each coating process parameter (photoresist thickness, coating speed, exposure energy density, curing time) corresponding to the batch are all set to 1, indicating that the combination of process parameters used in the batch can produce a TFT with qualified performance. If it is lower than the preset threshold, the tag value is set to 0, indicating that the combination of process parameters used in the batch has failed to produce a TFT with qualified performance.

[0036] By converting continuous performance index values ​​into binary labels, the computational complexity of subsequent contribution statistics can be simplified, while making the statistical results more intuitive. The number of batches with a label value of 1 directly reflects the ability to produce qualified products within a certain parameter value range.

[0037] S203, for each type of adhesive coating process parameter, it is divided into multiple value ranges according to the preset value range division rules.

[0038] Specifically, for the four types of process parameters—photoresist thickness, coating speed, exposure energy density, and curing time—the values ​​are divided according to preset value range rules. These value range rules can be preset based on the physical characteristics of each parameter and the commonly used value ranges in actual production.

[0039] For example, the photoresist thickness can be divided into the following ranges: Thickness1 (0.7–1.2 μm), Thickness2 (1.2–2.0 μm), and Thickness3 (2.0–3.0 μm). The coating speed can be divided into Speed1 (low speed range), Speed2 (medium speed range), and Speed3 (high speed range). The exposure energy density and curing time are also divided into ranges according to similar rules.

[0040] By discretizing the continuously valued process parameters into several value intervals, the influence of different parameter intervals on TFT performance can be more clearly reflected, which facilitates subsequent statistical analysis and candidate interval selection.

[0041] S204, determine the contribution of each value range of the coating process parameter for each type based on the label value.

[0042] In one embodiment, for each range of values ​​for each type of coating process parameter, the number of batches of coating process parameters with a label value of 1 that fall into each range is counted. For each range, the ratio of the number of batches to the total number of batches is calculated as the contribution of the range.

[0043] Taking photoresist thickness as an example, assuming that in the M batches obtained, the Thickness1 interval is applied to 300 batches, of which 200 batches have a label value of 1 (i.e., threshold voltage uniformity meets the standard), the Thickness2 interval is applied to 400 batches, of which 300 batches have a label value of 1, and the Thickness3 interval is applied to 300 batches, of which 100 batches have a label value of 1, then the contribution of each value interval is as follows: Thickness1 is 200 / M, Thickness2 is 300 / M, and Thickness3 is 100 / M.

[0044] This embodiment uses an equal-weighted contribution rate, which is simple and intuitive to calculate and is suitable for scenarios where production conditions are relatively stable for each batch and there is no significant process drift.

[0045] In another embodiment, for each range of values ​​of the coating process parameters of each type, the process parameters and their label values ​​falling into each range are statistically analyzed. The label values ​​of each batch falling into each range are weighted, with different batches assigned different weight coefficients according to their production time. The later the production time, the higher the weight coefficient. The weighted contribution of each range is calculated based on the weighted label values, and the weighted contribution is used as the contribution of each range.

[0046] Specifically, the formula for calculating the weighted contribution is as follows: ; Where k is the total number of batches falling within the range of values. This is the weighting coefficient for the i-th batch, assigned according to production time, with later-produced batches having higher weighting coefficients. The weighting coefficients can be assigned linearly (e.g., 1 for the first batch, 2 for the second, and so on), exponentially, or using other preset increasing functions. (Label) i It is the label value (0 or 1) of the process parameters for the i-th batch.

[0047] Taking photoresist thickness as an example, the preset value range includes range A (0.7~1.2μm) and range B (1.2~2.0μm). The data for each of the five most recent batches (batch 1 to batch 5, produced from earliest to latest) are as follows: The table above uses linearly increasing weighting coefficients (1 for batch 1, 2 for batch 2, and 5 for batch 5) to calculate the weighted contributions of interval A and interval B respectively, as follows: Interval A (0.7–1.2 μm): The batches falling into this range are batch 1 (label 1, weight 1), batch 2 (label 1, weight 2), and batch 5 (label 1, weight 5), with a weighted contribution of (1×1+1×2+1×5) / (1+2+5)=1.0. Interval B (1.2–2.0 μm): The batches falling into this range are batch 3 (label 0, weight 3) and batch 4 (label 1, weight 4), with a weighted contribution of (0×3+1×4) / (3+4)≈0.57.

[0048] By employing a time-weighted contribution method, batches produced later in the production process have typically undergone prior debugging and optimization of their coating processes. The correlation between process parameters and TFT performance better represents the actual state of the current production line. By assigning higher weights to recent batches, the contribution evaluation results can be more closely aligned with current production realities, avoiding the dilution of recent effective information by a large amount of early historical data. Furthermore, as production continues, the weighting coefficients automatically shift towards the latest batches, enabling the contribution evaluation results to adaptively follow process drift or equipment status changes without requiring manual readjustment of the evaluation strategy.

[0049] S205. For each type of photoresist coating process parameter, select the n value ranges with the largest contribution to obtain n candidate photoresist thickness ranges, n candidate coating speed ranges, n candidate exposure energy density ranges, and n candidate curing time ranges, where n is a preset positive integer.

[0050] Specifically, n is a preset positive integer, such as n=2 or n=3. For the four types of process parameters, namely photoresist thickness, coating speed, exposure energy density and curing time, the contribution of each value interval calculated in step S204 is sorted from largest to smallest, and the n value intervals with the largest contribution are selected as candidate intervals.

[0051] This embodiment selects the top n value intervals with the highest contribution, which can significantly reduce the number of process parameter combinations that need to be generated subsequently, reduce computational complexity, and improve optimization efficiency. At the same time, since the selected intervals are all the intervals with the highest historical contribution, the candidate intervals themselves are guaranteed to have high quality.

[0052] S206: Extract one process parameter range from each of the n candidate photoresist thickness ranges, n candidate coating speed ranges, n candidate exposure energy density ranges, and n candidate curing time ranges, and combine them to generate multiple sets of process parameter combinations.

[0053] One interval is selected from each of the four types of candidate process parameter intervals for combination. Specifically, one process parameter interval can be extracted from each of the n candidate photoresist thickness intervals, n candidate coating speed intervals, n candidate exposure energy density intervals, and n candidate curing time intervals for full combination to generate n 4 Group process parameter combinations.

[0054] For example, when n=2, there are 2×2×2×2=16 sets of process parameter combinations, and when n=3, there are 3×3×3×3=81 sets of process parameter combinations. Each set of process parameter combinations includes a specific photoresist thickness range, a specific coating speed range, a specific exposure energy density range, and a specific curing time range.

[0055] This embodiment performs a full combination among the candidate intervals of each parameter. This full combination method can cover all possible coupling situations of the four process parameters, avoiding the omission of potential high-quality combinations. At the same time, since only the n intervals with the highest contribution are retained for each parameter, the total number of combinations is controlled within a reasonable range, ensuring computational efficiency.

[0056] S207, input each set of process parameter combinations into the pre-trained TFT performance prediction model to obtain the threshold voltage uniformity and its confidence level corresponding to each set of process parameter combinations.

[0057] In one optional embodiment, the training steps for the TFT performance prediction model are as follows: S1 uses the coating process parameters and TFT threshold voltage uniformity used in each historical batch as training data, where the TFT threshold voltage uniformity in each training data is used as the true label.

[0058] The coating process parameters (photoresist thickness, coating speed, exposure energy density, curing time) used in each historical batch and the threshold voltage uniformity of the TFT in that batch are used as training data, with the threshold voltage uniformity of the TFT in each training data serving as the true label.

[0059] S2, input each training data point into the pre-built initial TFT performance prediction model to obtain the predicted threshold voltage uniformity and its confidence level.

[0060] S3, calculate the loss value using the predicted threshold voltage uniformity and the actual threshold voltage uniformity; S4, determine whether the training stop condition is met (e.g., the loss value is lower than the preset threshold or the preset maximum number of iterations is reached); if yes, execute S5; otherwise, execute S6. S5. Stop training the initial TFT performance prediction model and obtain the trained TFT performance prediction model. S6, adjust the model parameters based on the loss value, and return to S2.

[0061] The initial TFT performance prediction model can be implemented using various machine learning models, such as Neural Network, Support Vector Regression (SVR), Random Forest, Gradient Boosting Decision Tree (GBDT), etc. The input layer of the model is a four-dimensional vector, corresponding to four coating process parameters, and the output layer is a two-dimensional vector, corresponding to the predicted threshold voltage uniformity and its confidence level. During training, the model not only learns the mapping relationship from process parameters to performance indicators, but also estimates the uncertainty of the prediction results through appropriate model structures (such as Bayesian neural networks or neural networks integrated with Dropout), thereby outputting the confidence level.

[0062] After being trained, the TFT performance prediction model can simultaneously output the predicted threshold voltage uniformity and the confidence level corresponding to the prediction result after a set of coating process parameters are input. The output of the confidence level enables the subsequent screening process to quantitatively evaluate the reliability of each prediction result, avoiding optimization decision errors caused by the low prediction accuracy of the model in certain parameter space regions.

[0063] S208, the optimal combination of process parameters is selected based on the threshold voltage uniformity and confidence level corresponding to each combination of process parameters.

[0064] Specifically, the process parameter combinations can be sorted from largest to smallest according to the threshold voltage uniformity. The top M process parameter combinations are selected, and J process parameter combinations with confidence levels greater than a preset confidence threshold are chosen from the M process parameter combinations. For each of the J process parameter combinations, the weighted sum of the threshold voltage uniformity and confidence level is calculated to obtain the score of each process parameter combination. The process parameter combination with the highest score is determined as the optimal process parameter combination.

[0065] Specifically, the scoring formula is as follows: S=U pred ×w1+C×w2; Among them, U pred This is the predicted value for threshold voltage uniformity, where C is the confidence level. w1 and w2 are the weighting coefficients for threshold voltage uniformity and confidence, respectively. They can be adjusted according to actual needs. For example, when more attention is paid to prediction performance, w1 > w2 can be set, and when more attention is paid to prediction reliability, w1 < w2 can be set. By default, w1 = w2 = 0.5, that is, the two are equally weighted.

[0066] By taking into account both prediction performance and prediction reliability through a weighted summation method, the advantages and disadvantages of each candidate process parameter combination can be evaluated more comprehensively and in a balanced manner, avoiding the one-sided pursuit of high prediction values ​​while ignoring prediction risks.

[0067] In this embodiment of the invention, when a preset trigger event is detected, the coating process parameters of multiple batches of TFTs and the performance indicators of each batch of TFTs are obtained. The coating process parameters are labeled and their contribution is calculated. Based on the contribution, value ranges are selected from the coating process parameters to construct multiple sets of process parameter combinations. The multiple sets of process parameter combinations are input into the TFT performance prediction model to obtain the threshold voltage uniformity and its confidence level corresponding to each set of process parameter combinations. Furthermore, the optimal process parameter combination is selected based on the threshold voltage uniformity and its confidence level corresponding to each set of process parameter combinations. This avoids the deviation caused by human experience and subjective judgment, and effectively solves the technical problems of low efficiency, strong subjectivity and difficulty in accurately dealing with multi-parameter coupling optimization in the prior art. It can optimize process parameters efficiently, accurately and reliably.

[0068] Example 3 Figure 3 This is a schematic diagram of a TFT coating process parameter optimization device provided in Embodiment 3 of the present invention. Figure 3 As shown, the TFT coating process parameter optimization device includes: The process parameter and performance index acquisition module 301 is used to acquire the photoresist coating process parameters and the TFT performance index of each batch of TFTs produced within a preset time window when a preset trigger event is detected. The photoresist coating process parameters include photoresist thickness, coating speed, exposure energy density and curing time, and the performance index includes threshold voltage uniformity. The process parameter combination construction module 302 is used to construct multiple sets of process parameter combinations based on the coating process parameters and TFT performance indicators used in each batch. The performance prediction module 303 is used to input each set of process parameter combinations into a pre-trained TFT performance prediction model to obtain the threshold voltage uniformity and its confidence level corresponding to each set of process parameter combinations. The process parameter combination screening module 304 is used to screen the optimal process parameter combination based on the threshold voltage uniformity and confidence level corresponding to each group of process parameter combinations.

[0069] The TFT coating process parameter optimization device provided in this embodiment of the invention can execute the TFT coating process parameter optimization method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0070] Example 4 Figure 4 A schematic diagram of an electronic device 40 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0071] like Figure 4 As shown, the electronic device 40 includes at least one processor 41 and a memory, such as a read-only memory (ROM) 42 or a random access memory (RAM) 43, communicatively connected to the at least one processor 41. The memory stores computer programs executable by the at least one processor. The processor 41 can perform various appropriate actions and processes based on the computer program stored in the ROM 42 or loaded from storage unit 48 into the RAM 43. The RAM 43 may also store various programs and data required for the operation of the electronic device 40. The processor 41, ROM 42, and RAM 43 are interconnected via a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0072] Multiple components in electronic device 40 are connected to I / O interface 45, including: input unit 46, such as keyboard, mouse, etc.; output unit 47, such as various types of monitors, speakers, etc.; storage unit 48, such as disk, optical disk, etc.; and communication unit 49, such as network card, modem, wireless transceiver, etc. Communication unit 49 allows electronic device 40 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0073] Processor 41 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 41 include, but are not limited to, central processing unit (CPU), graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 41 performs the various methods and processes described above, such as the TFT coating process parameter optimization method.

[0074] In some embodiments, the TFT coating process parameter optimization method can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 48. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 40 via ROM 42 and / or communication unit 49. When the computer program is loaded into RAM 43 and executed by processor 41, one or more steps of the TFT coating process parameter optimization method described above can be performed. Alternatively, in other embodiments, processor 41 can be configured to perform the TFT coating process parameter optimization method by any other suitable means (e.g., by means of firmware).

[0075] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0076] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0077] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0078] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0079] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0080] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0081] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0082] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for optimizing TFT coating process parameters, characterized in that, include: When a preset trigger event is detected, the coating process parameters of multiple batches of TFTs produced within a preset time window and the performance indicators of each batch of TFTs are obtained. The coating process parameters include photoresist thickness, coating speed, exposure energy density and curing time. The performance indicators include threshold voltage uniformity. Multiple sets of process parameter combinations are constructed based on the coating process parameters and TFT performance indicators used in each batch; Each set of process parameter combinations is input into a pre-trained TFT performance prediction model to obtain the threshold voltage uniformity and its confidence level corresponding to each set of process parameter combinations. The optimal combination of process parameters is selected based on the threshold voltage uniformity and confidence level of each combination.

2. The method according to claim 1, characterized in that, Obtain the coating process parameters and performance indicators of each batch of TFTs produced within a preset time window, including: For the target TFT product, the trigger condition is set to trigger once every N batches produced, where N is a preset positive integer; When the production batch quantity is detected to be an integer multiple of N, the coating process parameters and threshold voltage uniformity of the M batches of TFTs traced back from the current trigger time are obtained.

3. The method according to claim 1, characterized in that, Multiple sets of process parameter combinations are constructed based on the coating process parameters and TFT performance indicators used in each batch, including: For each batch, if the threshold voltage uniformity of the TFT is greater than or equal to the preset threshold, the label value of each coating process parameter is set to 1; otherwise, it is set to 0. For each type of adhesive application process parameter, it is divided into multiple value ranges according to the preset value range division rules; The contribution of each value range of the coating process parameter for each type is determined based on the label value; For each type of photoresist coating process parameter, the n value ranges with the largest contribution are selected to obtain n candidate photoresist thickness ranges, n candidate coating speed ranges, n candidate exposure energy density ranges, and n candidate curing time ranges, where n is a preset positive integer; One process parameter range is extracted from each of the n candidate photoresist thickness ranges, n candidate coating speed ranges, n candidate exposure energy density ranges, and n candidate curing time ranges, and these ranges are combined to generate the multiple sets of process parameter combinations.

4. The method according to claim 3, characterized in that, The contribution of each range of values ​​for each type of adhesive coating process parameter is determined based on the label value, including: For each range of values ​​for each type of adhesive coating process parameter, count the number of batches of adhesive coating process parameters that fall within each range and have a label value of 1. For each value interval, the ratio of the number of batches to the total number of batches is calculated as the contribution of the value interval.

5. The method according to claim 3, characterized in that, The contribution of each range of values ​​for each type of adhesive coating process parameter is determined based on the label value, including: For each type of adhesive coating process parameter, the process parameters falling into each range and their label values ​​are statistically analyzed. The label values ​​of each batch falling into each value range are weighted, with different batches assigned different weight coefficients according to their production time, and the later the production time of the batch, the higher the weight coefficient. The weighted contribution of each value interval is calculated based on the weighted label value, and the weighted contribution is used as the contribution of each value interval.

6. The method according to any one of claims 1-5, characterized in that, The TFT performance prediction model is trained through the following steps: The coating process parameters and TFT threshold voltage uniformity used in each historical batch are used as training data, and the TFT threshold voltage uniformity in each training data is used as the true label. Each training data point is input into a pre-built initial TFT performance prediction model to obtain the predicted threshold voltage uniformity and its confidence level. The loss value is calculated using the predicted threshold voltage uniformity and the actual threshold voltage uniformity. Determine whether the conditions for stopping training are met; If so, stop training the initial TFT performance prediction model and obtain the trained TFT performance prediction model; If not, adjust the model parameters based on the loss value and return to the step of inputting each training data point into the pre-built initial TFT performance prediction model.

7. The method according to any one of claims 1-5, characterized in that, The optimal combination of process parameters was selected based on the threshold voltage uniformity and confidence level of each combination, including: The process parameter combinations are sorted from largest to smallest according to the threshold voltage uniformity, and the top M process parameter combinations are selected. Select J process parameter combinations from the M process parameter combinations, each with a confidence level greater than a preset confidence threshold; For the J combinations of process parameters, calculate the weighted sum of threshold voltage uniformity and confidence level for each combination of process parameters to obtain the score for each combination of process parameters. The combination of process parameters with the highest score is determined as the optimal combination of process parameters.

8. A device for optimizing TFT coating process parameters, characterized in that, include: The process parameter and performance index acquisition module is used to acquire the photoresist coating process parameters and the TFT performance index of each batch of TFTs produced within a preset time window when a preset trigger event is detected. The photoresist coating process parameters include photoresist thickness, coating speed, exposure energy density and curing time, and the performance index includes threshold voltage uniformity. The process parameter combination construction module is used to construct multiple sets of process parameter combinations based on the coating process parameters and TFT performance indicators used in each batch. The performance prediction module is used to input each combination of process parameters into a pre-trained TFT performance prediction model to obtain the threshold voltage uniformity and its confidence level corresponding to each combination of process parameters. The process parameter combination screening module is used to screen the optimal process parameter combination based on the threshold voltage uniformity and confidence level of each process parameter combination.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the TFT coating process parameter optimization method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the TFT coating process parameter optimization method according to any one of claims 1-7.

Images