Semiconductor etching process parameter optimization method and device and storage medium

Abstract

The invention relates to the technical field of data processing, and provides a semiconductor etching process parameter optimization method and device and a storage medium, and the method comprises the steps: obtaining the real-time data of a semiconductor etching process in the etching process, carrying out the preprocessing, obtaining a cleaned data set, extracting key features, generating an etching state weight matrix, and carrying out the calculation of the weight matrix; optimizing and adjusting the air pressure distribution parameter of the etching chamber to obtain an air pressure set value; acquiring temperature distribution data of the etching chamber, and calculating the air pressure set value and the temperature distribution data to obtain a radio frequency power set value; and optimizing the semiconductor etching process parameters by using the radio frequency power set value and the air pressure set value. The radio frequency power set value and the air pressure set value are obtained and applied through real-time data acquisition and processing, the overall performance of the etching process is improved, and the problems of low efficiency, insufficient precision and poor adaptability when the process complexity and the real-time performance requirement are high are solved.

Description

Technical Field

[0001] The present application relates to the technical field of data processing, and in particular to a method, device and storage medium for optimizing semiconductor etching process parameters. Background Art

[0002] The semiconductor etching process is a vital part of the semiconductor manufacturing process. With the increasing miniaturization and performance improvement of electronic devices, the semiconductor etching process has become more and more prominent in the manufacture of integrated circuits. The etching process forms a precise microstructure by removing excess materials, thereby realizing the functions of semiconductor devices. With the continuous development of semiconductor technology, higher requirements are placed on the accuracy and efficiency of the etching process.

[0003] Among the relevant technical means, etching process optimization is usually carried out based on experience and experiments. Operators determine the optimal parameter combination through repeated experiments and debugging, including adjusting etching process parameters such as gas pressure, RF power, temperature, and gas flow.

[0004] Regarding the above technical solution, although the etching process parameters can be optimized through experience and experiments, due to the lack of in-depth analysis of real-time data, it is unable to fully adapt to the complex and changing process environment, and there are certain limitations in process stability and consistency. There are problems of low efficiency, insufficient precision and poor adaptability. Summary of the invention

[0005] In order to improve the problems of low efficiency, insufficient precision and poor adaptability when the process complexity and real-time requirements are high, the present application provides a semiconductor etching process parameter optimization method, device and storage medium.

[0006] The present invention provides a method for optimizing semiconductor etching process parameters, comprising: acquiring real-time data of a semiconductor etching process during an etching process, preprocessing the real-time data to obtain a cleaned data set; extracting key features that affect the etching effect based on the cleaned data set, inputting the key features into a preset prediction model to obtain an etching residual thickness prediction value, an etching sidewall angle prediction value, and an etching rate deviation prediction value; generating an etching state weight matrix according to the etching residual thickness prediction value, the etching sidewall angle prediction value, and the etching rate deviation prediction value, optimizing and adjusting the gas pressure distribution parameters of an etching chamber based on the etching state weight matrix to obtain a gas pressure setting value; acquiring temperature distribution data of the etching chamber, calculating the gas pressure setting value and the temperature distribution data to obtain an etching uniformity correction parameter, adjusting the radio frequency power distribution based on the etching uniformity correction parameter to obtain a radio frequency power setting value; and optimizing semiconductor etching process parameters using the radio frequency power setting value and the gas pressure setting value.

[0007] As a preferred solution, the step of obtaining real-time data of the semiconductor etching process during the etching process, preprocessing the real-time data, and obtaining a cleaned data set includes: obtaining an original data set during the etching process; wherein the original data set includes plasma emission spectrum data, etching chamber pressure data, and etching layer material characteristic data; performing frequency domain transformation on the plasma emission spectrum data to obtain plasma excitation characteristic distribution, performing time series decomposition on the etching chamber pressure data to obtain a dynamic fluctuation component; calculating a short-period pressure fluctuation compensation value of the etching chamber based on the dynamic fluctuation component, and generating an etching condition adjustment factor based on the short-period pressure fluctuation compensation value, so as to achieve a good etching condition adjustment effect; The etching condition adjustment factor is corrected by using the etching layer material characteristic data to obtain a material adaptation correction value, and the etching surface charge accumulation parameter is calculated based on the material adaptation correction value to obtain a dynamic charge distribution of the etching process; a plasma energy density mapping matrix is ​​calculated based on the etching surface charge accumulation parameter and the plasma excitation characteristic distribution, and the plasma energy density mapping matrix is ​​fused with the short-period pressure fluctuation compensation value to obtain an etching stability factor and an etching rate compensation factor; the original data set is screened by using the etching stability factor and the etching rate compensation factor, and the screened data set is normalized to obtain a cleaned data set.

[0008] As a preferred solution, the step of extracting key features that affect the etching effect based on the cleaned data set, inputting the key features into a preset prediction model, and obtaining a predicted value of the etching residual thickness, a predicted value of the etching sidewall angle, and a predicted value of the etching rate deviation comprises: applying a data mining algorithm to extract key features that affect the etching effect from the cleaned data set; wherein the key features include etching chamber gas component change data, radio frequency power real-time fluctuation data, and etching electrode potential distribution data; calculating a radio frequency power offset factor through the radio frequency power real-time fluctuation data, and calculating an etching electrode dynamic response parameter through the etching electrode potential distribution data; calculating a plasma active particle density using the radio frequency power offset factor and the etching electrode dynamic response parameter, and jointly analyzing the plasma active particle density and the etching chamber gas component change data to obtain an etching uniformity feature group; and calculating a set of etching uniformity feature groups based on the etching uniformity feature group and the etching chamber environment parameters. The etching chamber environment parameters refer to the influencing parameters of the gas pressure parameters, the temperature parameters and the gas flow rate parameters on the plasma reaction; the material removal rate characteristic group is calculated by the etching region energy distribution matrix and the etching layer material characteristic parameters to obtain the etching reaction region distribution characteristics, and the etching residual thickness prediction value is calculated based on the etching reaction region distribution characteristics; wherein the etching layer material characteristic parameters refer to the physicochemical characteristics of the chemical composition, density parameters and reactivity parameters of the material; the etching side wall potential change is calculated based on the etching surface charge accumulation parameters, the etching side wall morphology parameters are calculated according to the etching side wall potential change and the plasma active particle density, and the etching side wall angle prediction value is generated according to the etching side wall morphology parameters; the etching rate adjustment coefficient is calculated based on the short-period pressure fluctuation compensation value, the etching rate adjustment coefficient is weightedly fused with the etching uniformity characteristic group to obtain the etching rate deviation prediction value.

[0009] As a preferred solution, the calculation formula for calculating the etching sidewall morphology parameters according to the etching sidewall potential change and the plasma active particle density is as follows: in, is the etching side wall morphology parameter, The potential change of the etching side wall is is the plasma active particle density, and is the constant obtained from experimental fitting.

[0010] As a preferred scheme, the step of generating an etching state weight matrix according to the etching residual thickness prediction value, the etching sidewall angle prediction value and the etching rate deviation prediction value, optimizing and adjusting the gas pressure distribution parameters of the etching chamber based on the etching state weight matrix, and obtaining the gas pressure setting value includes: calculating the etching layer thickness distribution weight using the residual thickness prediction value and the etching sidewall angle prediction value, calculating the etching rate adjustment factor by combining the etching layer thickness distribution weight with the etching rate deviation prediction value, and generating the etching state weight matrix based on the etching rate adjustment factor; calculating the etching layer thickness correction coefficient by using the etching state weight matrix, calculating the etching gas flow distribution parameter by combining the etching layer thickness correction coefficient with the etching condition adjustment factor, and generating the etching gas optimization parameter group according to the etching gas flow distribution parameter; calculating the chamber internal gas uniformity adjustment factor based on the etching gas optimization parameter group, calculating the etching chamber pressure correction value by using the chamber internal gas uniformity adjustment factor and the etching rate adjustment factor, and generating the gas pressure setting value according to the etching chamber pressure correction value.

[0011] As a preferred solution, the calculation formula for calculating the correction value of the etching chamber pressure by using the chamber internal gas uniformity adjustment factor and the etching rate adjustment factor is as follows: in, is the correction value of the etching chamber pressure, is the gas uniformity adjustment factor inside the chamber, is the rate of change of the gas pressure in the etching chamber with time, is the etching rate adjustment factor, is the correction factor in the process.

[0012] As a preferred embodiment, the steps of obtaining the temperature distribution data of the etching chamber, calculating the gas pressure setting value and the temperature distribution data to obtain an etching uniformity correction parameter, adjusting the RF power distribution based on the etching uniformity correction parameter, and obtaining the RF power setting value include: obtaining the temperature distribution data of the etching chamber through a sensor, calculating the etching uniformity correction parameter using the gas pressure setting value and the temperature distribution data, calculating the etching layer thermal response time parameter based on the etching uniformity correction parameter, and generating an etching area temperature compensation factor according to the etching layer thermal response time parameter; calculating the RF power correction value according to the etching area temperature compensation factor, calculating the RF power dynamic adjustment parameter using the RF power correction value in combination with the etching rate adjustment factor, and generating the RF power setting value according to the RF power dynamic adjustment parameter.

[0013] As a preferred embodiment, the step of optimizing the semiconductor etching process parameters using the RF power setting value and the gas pressure setting value includes: calculating a plasma energy density correction parameter based on the RF power setting value and the gas pressure setting value, and calculating an etching morphology optimization parameter group based on the plasma energy density correction parameter combined with the etching sidewall morphology parameter; and optimizing the semiconductor etching process parameters based on the etching morphology optimization parameter group.

[0014] The present application also provides a semiconductor etching process parameter optimization device, including: an acquisition module, used to acquire real-time data of a semiconductor etching process during an etching process, pre-process the real-time data, and obtain a cleaned data set; an extraction module, used to extract key features that affect the etching effect based on the cleaned data set, input the key features into a preset prediction model, and obtain an etching residual thickness prediction value, an etching sidewall angle prediction value, and an etching rate deviation prediction value; a generation module, used to generate an etching state weight matrix according to the etching residual thickness prediction value, the etching sidewall angle prediction value, and the etching rate deviation prediction value, and optimize and adjust the gas pressure distribution parameters of the etching chamber based on the etching state weight matrix to obtain a gas pressure setting value; a calculation module, used to acquire temperature distribution data of the etching chamber, calculate the gas pressure setting value and the temperature distribution data, obtain an etching uniformity correction parameter, and adjust the radio frequency power distribution based on the etching uniformity correction parameter to obtain a radio frequency power setting value; an optimization module, used to optimize the semiconductor etching process parameters using the radio frequency power setting value and the gas pressure setting value.

[0015] The present application also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the processor executes the semiconductor etching process parameter optimization method as described in any one of the above.

[0016] Compared with the prior art, the present application has the following beneficial effects: high efficiency and sufficient precision. The accuracy and reliability of data are ensured through the collection and processing of real-time data. The accuracy of etching parameter prediction is improved through the extraction of key features and the application of prediction models. The adaptability and stability of the etching process are enhanced through the generation and optimization adjustment of weight matrices. The uniformity and consistency of the etching effect are improved by calculating and applying etching uniformity correction parameters. The actual application of RF power setting values ​​and gas pressure setting values ​​significantly improves the overall performance of the etching process, and improves the problems of low efficiency, insufficient precision and poor adaptability when the process complexity and real-time requirements are high. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0018] The structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with this technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantive technical significance. Any structural modification, change in proportion or adjustment of size, without affecting the effects and purposes that can be achieved by the present invention, should still fall within the scope of the technical contents disclosed by the present invention.

[0019] Figure 1 It is a schematic flow chart of a method for optimizing semiconductor etching process parameters provided by an embodiment of the present invention; Figure 2 It is a schematic block diagram of the structure of a semiconductor etching process parameter optimization device provided by an embodiment of the present invention.

[0020] Description of reference numerals: 10. Semiconductor etching process parameter optimization device; 11. Acquisition module; 12. Extraction module; 13. Generation module; 14. Calculation module; 15. Optimization module. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0023] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0024] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0025] The technical solution of the present invention is further described below with reference to the accompanying drawings and through specific implementation methods.

[0026] Embodiment 1: like Figure 1 As shown, the present application provides a method for optimizing semiconductor etching process parameters, including steps S100 to S500.

[0027] Step S100: acquiring real-time data of a semiconductor etching process during the etching process, and preprocessing the real-time data to obtain a cleaned data set.

[0028] In this step, sensors and detection equipment are used to collect data in the etching process in real time, including parameters such as air pressure, temperature, RF power, and etching rate. Specifically, these real-time data are transmitted to the processing unit through the data acquisition system for preliminary filtering and denoising to remove outliers and interference data, thereby obtaining a cleaned data set.

[0029] For example, during the etching process, the changes in gas pressure and temperature are monitored in real time, and data points with excessive fluctuations are filtered out through algorithms to ensure the accuracy and reliability of the cleaned data set.

[0030] Step S200: extract key features that affect the etching effect based on the cleaned data set, input the key features into a preset prediction model, and obtain a predicted value of the etching residual thickness, a predicted value of the etching sidewall angle, and a predicted value of the etching rate deviation.

[0031] In this step, data analysis technology is used to extract characteristic parameters that have a significant impact on the etching effect from the cleaned data set, such as etching time, gas type and flow rate, RF power, etc. Specifically, feature engineering methods are used to standardize and normalize these features to facilitate model training and prediction.

[0032] For example, in the semiconductor etching process, etching time and RF power are extracted as key features, and the preset machine learning model is used for training and prediction to obtain the predicted value of etching residual thickness.

[0033] Step S300, generating an etching state weight matrix according to the etching residual thickness prediction value, the etching sidewall angle prediction value and the etching rate deviation prediction value, optimizing and adjusting the gas pressure distribution parameters of the etching chamber based on the etching state weight matrix to obtain a gas pressure setting value.

[0034] In this step, a weight matrix reflecting the etching state is generated according to the etching parameters obtained by the prediction model. Specifically, a multidimensional data matrix is ​​used to represent the weight distribution of different parameters on the etching effect, and the gas pressure of the etching chamber is optimized and adjusted in combination with the current process state.

[0035] For example, through weight matrix analysis, it is found that the gas pressure at a certain moment has a greater impact on the etching residual thickness, so the gas pressure parameters at that moment are appropriately adjusted to optimize the etching effect.

[0036] Step S400, obtaining temperature distribution data of the etching chamber, calculating the gas pressure setting value and the temperature distribution data to obtain an etching uniformity correction parameter, adjusting the RF power distribution based on the etching uniformity correction parameter, and obtaining the RF power setting value.

[0037] In this step, the temperature distribution of each area in the etching chamber is monitored in real time using a temperature sensor, and these data are calculated and analyzed together with the gas pressure setting value. Specifically, through the numerical analysis method, the parameters used to correct the etching uniformity are calculated, and the RF power distribution is adjusted to make the etching process more uniform.

[0038] For example, when the temperature of a certain area in the etching chamber is high, the RF power in this area is appropriately reduced to achieve uniform etching.

[0039] Step S500: Optimizing semiconductor etching process parameters using RF power setting values ​​and gas pressure setting values.

[0040] In this step, the optimized and adjusted RF power setting value and gas pressure setting value are applied to the actual semiconductor etching process. Specifically, by accurately controlling the RF power and gas pressure, the stability and accuracy of the etching process are ensured, and the uniformity of the etching effect is improved.

[0041] For example, in the actual etching process, strictly following the optimized parameters can significantly improve the consistency and repeatability of the etching process.

[0042] In this embodiment, by acquiring real-time data of the semiconductor etching process during the etching process and preprocessing the real-time data, a cleaned data set is obtained. Then, based on the cleaned data set, key features that affect the etching effect are extracted, and the key features are input into a preset prediction model to obtain a predicted value of the etching residual thickness, a predicted value of the etching sidewall angle, and a predicted value of the etching rate deviation. An etching state weight matrix is ​​generated based on these predicted values, and the gas pressure distribution parameters of the etching chamber are optimized and adjusted based on the etching state weight matrix to obtain a gas pressure setting value. Next, the temperature distribution data of the etching chamber is obtained, and the gas pressure setting value and the temperature distribution data are calculated to obtain an etching uniformity correction parameter, and the RF power distribution is adjusted based on the etching uniformity correction parameter to finally obtain the RF power setting value. The semiconductor etching process parameters are optimized using the RF power setting value and the gas pressure setting value.

[0043] Through the preprocessing of real-time data and the extraction of key features, this method can effectively improve the accuracy and stability of the etching process. The optimization adjustment based on the prediction model can not only adapt to the complex and changeable process environment, but also greatly improve the process consistency of the etching chamber. This solution not only improves work efficiency and reduces dependence on operator experience, but also ensures higher etching uniformity and accuracy, meets the requirements of high precision and high efficiency in the semiconductor manufacturing process, improves the problems of low efficiency, insufficient accuracy and poor adaptability when the process complexity and real-time requirements are high, and realizes high-precision, high-efficiency and high-stability semiconductor etching process parameter optimization.

[0044] Embodiment 2: In step S100, an original data set in the etching process is acquired; wherein the original data set includes plasma emission spectrum data, etching chamber pressure data and etching layer material characteristic data.

[0045] Real-time data acquisition is performed by using highly sensitive sensors and advanced acquisition systems. Specifically, the plasma emission spectrum data is spectrally analyzed to obtain the specific spectral line distribution of plasma excitation. A high-precision pressure sensor is used for the etching chamber pressure data to record real-time pressure changes and synchronize the data with a timestamp. For the material characteristic data of the etching layer, the chemical composition and physical properties of the etching layer are measured in detail by means of scanning electron microscopy (SEM) and energy spectrum analysis (EDS).

[0046] For example, during the etching process, laser induced spectroscopy (LIBS) is used to monitor the spectral distribution of the plasma in real time, while a high-frequency data acquisition system is used to synchronously record the pressure changes in the chamber. These data are transmitted to the control center via the network and combined with the composition and structural analysis results of the etched material to form a complete original data set.

[0047] The plasma emission spectrum data is transformed in the frequency domain to obtain the plasma excitation characteristic distribution, and the etching chamber pressure data is decomposed into time series to obtain the dynamic fluctuation component.

[0048] The plasma emission spectrum data is converted from the time domain to the frequency domain by Fourier transform. Specifically, the main spectral characteristic peaks are identified and the spectral energy of different bands is normalized. Wavelet transform is applied to the etching chamber pressure data to decompose the time series and extract the high-frequency and low-frequency components in order to better analyze the dynamic characteristics of pressure fluctuations.

[0049] For example, by performing frequency domain analysis on the spectral data of an etching chamber, it was found that the plasma excitation has a significant peak at a specific wavelength, indicating that the energy corresponding to this wavelength varies greatly. At the same time, by analyzing the pressure data of the etching chamber through wavelet analysis, the short-term components and long-term trends of the pressure fluctuations are obtained, which helps to understand the instantaneous pressure changes during the etching process.

[0050] The short-period pressure fluctuation compensation value of the etching chamber is calculated based on the dynamic fluctuation component, and the etching condition adjustment factor is generated based on the short-period pressure fluctuation compensation value. The etching condition adjustment factor is corrected using the etching layer material characteristic data to obtain the material adaptation correction value, and the etching surface charge accumulation parameter is calculated based on the material adaptation correction value to obtain the dynamic charge distribution of the etching process.

[0051] By establishing a pressure fluctuation compensation model, specifically, according to the short-period fluctuation component, the real-time pressure compensation value is calculated to adjust the pressure setting during the etching process. Using the material characteristic data of the etching layer, the working condition adjustment factor is corrected to adapt it to the etching requirements of different materials, and by calculating the surface charge accumulation parameters, the distribution change of the charge during the etching process is predicted.

[0052] For example, through experimental analysis of the etching layer of a certain material, it was found that the charge accumulation effect was obvious under a specific pressure. Based on this data, the operating factor was adjusted so that the etching process could maintain a stable charge distribution and ensure the uniformity of the etching effect.

[0053] The plasma energy density mapping matrix is ​​calculated based on the etching surface charge accumulation parameters and the plasma excitation characteristic distribution. The plasma energy density mapping matrix is ​​fused with the short-period pressure fluctuation compensation value to obtain the etching stability factor and etching rate compensation factor.

[0054] By integrating the charge accumulation parameters and plasma excitation characteristics, a plasma energy density mapping matrix is ​​constructed to represent the energy distribution under different etching conditions. The matrix is ​​combined with the short-period pressure fluctuation compensation value to calculate the etching stability factor, which is used to evaluate the stability of the etching process, and the etching rate compensation factor is calculated to optimize the etching rate.

[0055] For example, it is found through experiments that the etching rate changes significantly under a specific plasma energy density. Combined with the pressure compensation value, the etching stability factor is calculated so that the etching rate can be kept within the ideal range under changing etching conditions.

[0056] The original data set is screened using the etching stability factor and the etching rate compensation factor, and the screened data set is normalized to obtain a cleaned data set.

[0057] By applying a screening algorithm, specifically, data points in the original data set that do not meet the stability and rate requirements are removed, and the remaining data are normalized to be within the same magnitude range to facilitate subsequent analysis and model training.

[0058] For example, by screening an experimental data set, it is found that some data points are removed due to abnormal fluctuations. After normalization, the distribution of the cleaned data set in each parameter is more uniform, which is conducive to the subsequent feature extraction and prediction model application.

[0059] In step S200, a data mining algorithm is applied to extract key features that affect the etching effect from the cleaned data set; wherein the key features include etching chamber gas composition change data, RF power real-time fluctuation data, and etching electrode potential distribution data.

[0060] By constructing a multidimensional data analysis model, specifically, applying multivariate statistical analysis and machine learning algorithms, characteristic parameters closely related to the etching effect are extracted from the cleaned data set. For the gas component change data, gas chromatography and mass spectrometry analysis methods are used to identify the changes in gas composition under different working conditions. For the real-time fluctuation data of RF power, a high-frequency data acquisition system is used to record the fluctuation of RF power at different time points. For the etching electrode potential distribution data, a potentiometer and electric field sensor are used to measure the potential change on the electrode surface.

[0061] For example, by analyzing the data of a certain etching process, it was found that the change in the ratio of nitrogen and oxygen in the chamber significantly affects the etching rate. At the same time, the RF power fluctuates at a certain moment, resulting in uneven etching depth. The distribution of electrode potential also reflects the change of the electric field during the etching process. These characteristic parameters have a significant impact on the etching effect.

[0062] The RF power deviation factor is calculated through the real-time RF power fluctuation data, and the dynamic response parameters of the etching electrode are calculated through the etching electrode potential distribution data.

[0063] By establishing a radio frequency power fluctuation model, specifically, according to the real-time recorded radio frequency power data, the power offset at each time point is calculated, and these offsets are statistically analyzed to generate the radio frequency power offset factor. By performing curve fitting and dynamic response analysis on the etching electrode potential distribution data, the dynamic response parameters of the electrode under different working conditions are calculated.

[0064] For example, in a certain etching process, by calculating the RF power offset factor, it was found that the power fluctuation in a certain period of time was large, resulting in inconsistent etching depth. By analyzing the dynamic response data of the electrode, it was found that the response time of the electrode under different pressures and temperatures was significantly different. These parameters help optimize the etching process.

[0065] The plasma active particle density is calculated using the RF power offset factor and the dynamic response parameters of the etching electrode. The plasma active particle density and the etching chamber gas component change data are jointly analyzed to obtain the etching uniformity feature group.

[0066] By integrating the RF power offset factor and the electrode dynamic response parameters, a plasma physics model is constructed to calculate the plasma active particle density under different etching conditions. These particle density data are analyzed and correlated with the gas component change data to identify the etching uniformity feature group for etching process optimization.

[0067] For example, by numerically simulating the distribution of active particles in the plasma, combined with the actual measured RF power offset and electrode response parameters, the plasma active particle density is calculated. This is then combined with the gas composition change data for analysis to obtain a characteristic parameter group for etching uniformity.

[0068] An etching region energy distribution matrix is ​​constructed according to an etching uniformity feature group and etching chamber environmental parameters; wherein the etching chamber environmental parameters refer to parameters affecting plasma reaction such as gas pressure parameters, temperature parameters, and gas flow parameters.

[0069] By constructing an energy distribution model for the etching area, specifically, combining parameters such as gas pressure, temperature, and gas flow, the energy distribution in different areas is calculated. The matrix is ​​used to represent the energy distribution characteristics under different working conditions, providing a basis for subsequent process optimization.

[0070] For example, by simulating the energy distribution in the etching chamber and combining it with the actual gas pressure and temperature data, it was found that the energy density in a certain area was low, resulting in uneven etching. By adjusting the gas flow rate and RF power, the energy distribution was optimized and the consistency of the etching effect was improved.

[0071] The material removal rate characteristic group is calculated through the etching area energy distribution matrix and the etching layer material characteristic parameters, and the etching reaction area distribution characteristics are obtained. The etching residual thickness prediction value is calculated based on the etching reaction area distribution characteristics; wherein the etching layer material characteristic parameters refer to the physical and chemical properties of the material's chemical composition, density parameters, and reactivity parameters.

[0072] By establishing a material removal rate model, specifically, combining the etching area energy distribution matrix and the physical and chemical properties of the material, the removal rate of different materials during the etching process is calculated. Using these characteristic parameters, the etching residual thickness is predicted and the etching reaction area distribution characteristics are generated.

[0073] For example, by experimentally measuring the removal rates of different materials during etching, it was found that some materials are removed faster under high temperature conditions. Combined with energy distribution data, the residual thickness after etching can be predicted, the etching process parameters can be optimized, and the etching uniformity can be improved.

[0074] The etching sidewall potential change is calculated based on the etching surface charge accumulation parameters, the etching sidewall morphology parameters are calculated according to the etching sidewall potential change and the plasma active particle density, and the etching sidewall angle prediction value is generated according to the etching sidewall morphology parameters.

[0075] By establishing an etching sidewall potential change model, specifically, using the surface charge accumulation parameter to calculate the sidewall potential change, and combining the plasma active particle density data to calculate the morphological parameters of the etching sidewall, the angle change of the etching sidewall is predicted based on these morphological parameters, and the etching process is optimized.

[0076] For example, by experimentally measuring the etching sidewall morphology under different working conditions, it was found that the change in potential significantly affects the sidewall angle. Combined with the plasma particle density data, the sidewall angle under different conditions was predicted, the etching parameters were optimized, and the etching accuracy was improved.

[0077] The etching rate adjustment coefficient is calculated based on the short-period pressure fluctuation compensation value, and the etching rate adjustment coefficient is weightedly fused with the etching uniformity feature group to obtain the etching rate deviation prediction value.

[0078] By comprehensively analyzing the short-period pressure fluctuation compensation value, specifically, calculating the etching rate adjustment coefficient, and combining the etching uniformity feature group, the etching rate under different working conditions is predicted and adjusted. Using these predicted values, the etching process is optimized to ensure the consistency of the etching process.

[0079] For example, it is found through experiments that under certain pressure conditions, the etching rate deviates significantly from the expected value. By calculating the etching rate adjustment coefficient and combining the uniformity feature group, the etching rate is adjusted to optimize the etching process parameters.

[0080] Among them, the calculation formula for calculating the etching sidewall morphology parameters based on the etching sidewall potential change and the plasma active particle density is as follows: in, is the etching side wall morphology parameter, The potential change of the etching side wall is is the plasma active particle density, and is the constant obtained from experimental fitting.

[0081] By establishing a detailed mathematical model, specifically, using experimental data and numerical simulation results, the constants in the formula are fitted and , accurately predict the morphological parameters of the etched sidewalls and optimize the etching process based on these parameters.

[0082] For example, by fitting experimental data, the and The formula is used to predict the sidewall morphology under different conditions. Combined with the actual process parameters, the potential and particle density in the etching process are optimized to improve the etching accuracy.

[0083] In step S300, the residual thickness prediction value and the etching sidewall angle prediction value are used to calculate the etching layer thickness distribution weight, the etching rate adjustment factor is calculated by combining the etching layer thickness distribution weight with the etching rate deviation prediction value, and the etching state weight matrix is ​​generated based on the etching rate adjustment factor.

[0084] By establishing the distribution model of residual thickness and sidewall angle, specifically, the predicted values ​​are used to calculate the thickness distribution weights of each region. These weights represent the impact of different regions on the overall etching effect. Combined with the predicted value of etching rate deviation, the etching rate adjustment factor is calculated, and the etching state weight matrix is ​​generated based on these factors to further optimize the etching process.

[0085] For example, by calculating the thickness distribution weights of different regions in a certain etching process, it is found that the thickness changes in certain regions significantly affect the etching effect. Combined with the predicted value of the etching rate deviation, the etching process parameters are adjusted to improve the overall etching uniformity.

[0086] The etching layer thickness correction coefficient is calculated by the etching state weight matrix, and the etching gas flow distribution parameter is calculated by combining the etching layer thickness correction coefficient with the etching condition adjustment factor. The etching gas optimization parameter group is generated according to the etching gas flow distribution parameter.

[0087] By analyzing the etching state weight matrix, specifically, the thickness correction coefficients of different regions are calculated, which are used to correct the thickness changes of each region during the actual etching process. Combined with the working condition adjustment factor, the gas flow distribution parameters are optimized to generate the etching gas optimization parameter group to improve the uniformity and stability of the etching process.

[0088] For example, through experiments, it was found that the thickness of a certain area varied greatly. The thickness correction coefficient was used to make adjustments, optimize the gas flow distribution, and improve the consistency and accuracy of the etching effect.

[0089] The chamber internal gas uniformity adjustment factor is calculated based on the etching gas optimization parameter group, the etching chamber pressure correction value is calculated by the chamber internal gas uniformity adjustment factor and the etching rate adjustment factor, and the gas pressure setting value is generated according to the etching chamber pressure correction value.

[0090] By establishing a gas uniformity model, specifically, using the optimized parameter group to calculate the uniformity adjustment factor of the gas inside the etching chamber, combining the etching rate adjustment factor, calculating the chamber pressure correction value, and generating the gas pressure setting value based on these correction values, the uniformity of gas distribution and the stability of pressure during the etching process are ensured.

[0091] For example, by simulating the gas flow inside the chamber, it was found that the gas distribution in certain areas was uneven. The gas uniformity adjustment factor was used for optimization to generate a gas pressure setting value to improve the stability of the etching process.

[0092] The calculation formula for calculating the correction value of the etching chamber pressure by the chamber internal gas uniformity adjustment factor and the etching rate adjustment factor is as follows: in, is the correction value of the etching chamber pressure, is the gas uniformity adjustment factor inside the chamber, is the rate of change of the gas pressure in the etching chamber with time, is the etching rate adjustment factor, is the correction factor in the process.

[0093] By establishing a detailed pressure correction model, specifically, using experimental data and numerical simulation results, the constants in the formula are fitted and , accurately calculate the pressure correction values ​​of the etching chamber and optimize the etching process parameters based on these values.

[0094] For example, by fitting experimental data, the and The formula is used to calculate the pressure correction value under different conditions, optimize the gas pressure inside the chamber, and ensure the stability of the etching process.

[0095] In step S400, the temperature distribution data of the etching chamber is obtained through the sensor, the etching uniformity correction parameter is calculated using the gas pressure setting value and the temperature distribution data, and the etching layer thermal response time parameter is calculated based on the etching uniformity correction parameter, and the etching area temperature compensation factor is generated according to the etching layer thermal response time parameter.

[0096] The temperature distribution in the chamber is monitored in real time by a high-precision temperature sensor. Specifically, the gas pressure setting value and temperature data are used to calculate the etching uniformity correction parameters to ensure the temperature uniformity during the etching process. Based on these parameters, the thermal response time of the etched layer is calculated and the temperature compensation factor of the etched area is generated to optimize the thermal management of the etching process.

[0097] For example, by monitoring the temperature distribution in different areas during an etching process, it is found that temperature changes significantly affect etching uniformity. The temperature compensation factor is used to adjust and optimize the RF power setting to improve the consistency of the etching effect.

[0098] The RF power correction value is calculated according to the temperature compensation factor of the etching area, the RF power dynamic adjustment parameter is calculated using the RF power correction value combined with the etching rate adjustment factor, and the RF power setting value is generated according to the RF power dynamic adjustment parameter.

[0099] By establishing a radio frequency power correction model, specifically, using the temperature compensation factor to calculate the radio frequency power correction value, combined with the etching rate adjustment factor, the radio frequency power setting value is dynamically adjusted to ensure power stability and uniformity during the etching process.

[0100] For example, through experimental data analysis, it was found that the RF power fluctuated greatly during a certain etching process. Correction values ​​and adjustment factors were used for optimization to generate RF power setting values, thereby improving the stability and accuracy of the etching process.

[0101] In step S500, a plasma energy density correction parameter is calculated based on the RF power setting value and the gas pressure setting value, and an etching morphology optimization parameter group is calculated according to the plasma energy density correction parameter combined with the etching sidewall morphology parameter.

[0102] By comprehensively considering the RF power and gas pressure setting values, specifically, calculating the plasma energy density correction parameters, and combining the etching sidewall morphology parameters, the etching morphology parameter group is optimized to further improve the effect and accuracy of the etching process.

[0103] For example, by calculating the plasma energy density and combining it with the actual sidewall morphology data, the etching parameter group is optimized to ensure the morphological accuracy and consistency during the etching process.

[0104] The semiconductor etching process parameters are optimized based on the etching morphology optimization parameter group.

[0105] By utilizing the optimized morphological parameter group, specifically, adjusting the etching process parameters, the uniformity, accuracy and stability of the etching effect are improved, thereby ensuring high efficiency and high quality in the semiconductor manufacturing process.

[0106] For example, through experimental verification, a certain etching process was optimized using a morphology optimization parameter group, which improved the etching uniformity and accuracy, and effectively improved the performance and reliability of semiconductor devices.

[0107] In this embodiment, by acquiring the original data set in the etching process, including plasma emission spectrum data, etching chamber pressure data, and etching layer material characteristic data, the plasma emission spectrum data is transformed in the frequency domain to obtain the plasma excitation characteristic distribution, and the etching chamber pressure data is decomposed in time series to obtain the dynamic fluctuation component. The short-period pressure fluctuation compensation value of the etching chamber is calculated based on the dynamic fluctuation component, and the etching condition adjustment factor is generated. The etching condition adjustment factor is corrected using the etching layer material characteristic data to obtain the material adaptation correction value, and the etching surface charge accumulation parameter is calculated to obtain the dynamic charge distribution of the etching process. The plasma energy density mapping matrix is ​​calculated based on the etching surface charge accumulation parameter and the plasma excitation characteristic distribution to obtain the etching stability factor and the etching rate compensation factor. These factors are used to screen and normalize the original data set to obtain the cleaned data set.

[0108] Through data mining algorithms, key features are extracted from the cleaned data set, including etching chamber gas composition change data, RF power real-time fluctuation data, and etching electrode potential distribution data. The RF power offset factor and the dynamic response parameters of the etching electrode are calculated to obtain the etching uniformity feature group. The etching area energy distribution matrix is ​​constructed based on the etching uniformity feature group and the etching chamber environment parameters. The material removal rate feature group is calculated in combination with the material characteristic parameters to obtain the etching reaction area distribution characteristics and predict the etching residual thickness. The etching sidewall potential change and sidewall morphology parameters are calculated to predict the etching sidewall angle. The etching rate adjustment coefficient is calculated based on the short-period pressure fluctuation compensation value to generate the etching rate deviation prediction value.

[0109] The residual thickness prediction value and the etching side wall angle prediction value are used to calculate the etching layer thickness distribution weight and the etching rate adjustment factor, generate the etching state weight matrix, calculate the etching layer thickness correction coefficient and the etching gas flow distribution parameter, and generate the etching gas optimization parameter group. According to the etching gas optimization parameter group, the chamber internal gas uniformity adjustment factor and the etching rate adjustment factor are calculated to generate the gas pressure setting value. The temperature distribution data is obtained, the etching uniformity correction parameter and the thermal response time parameter are calculated, the temperature compensation factor is generated, the RF power correction value and the dynamic adjustment parameter are calculated, and the RF power setting value is generated. Finally, the etching process parameters are optimized through the RF power and gas pressure setting values ​​to improve the etching uniformity, accuracy and stability.

[0110] Embodiment 3: like Figure 2 As shown, the present application also provides a semiconductor etching process parameter optimization device 10, including an acquisition module 11, an extraction module 12, a generation module 13, a calculation module 14 and an optimization module 15.

[0111] The acquisition module 11 is mainly used to acquire real-time data of the semiconductor etching process during the etching process, and pre-process the real-time data to obtain a cleaned data set.

[0112] The extraction module 12 is mainly used to extract key features that affect the etching effect based on the cleaned data set, input the key features into a preset prediction model, and obtain the predicted value of the etching residual thickness, the predicted value of the etching sidewall angle, and the predicted value of the etching rate deviation.

[0113] The generation module 13 is mainly used to generate an etching state weight matrix according to the etching residual thickness prediction value, the etching side wall angle prediction value and the etching rate deviation prediction value, and optimize and adjust the gas pressure distribution parameters of the etching chamber based on the etching state weight matrix to obtain the gas pressure setting value.

[0114] The calculation module 14 is mainly used to obtain the temperature distribution data of the etching chamber, calculate the gas pressure setting value and the temperature distribution data, obtain the etching uniformity correction parameter, adjust the RF power distribution based on the etching uniformity correction parameter, and obtain the RF power setting value.

[0115] The optimization module 15 is mainly used to optimize the semiconductor etching process parameters by using the RF power setting value and the gas pressure setting value.

[0116] In this embodiment, the acquisition module 11, the extraction module 12, the generation module 13, the calculation module 14 and the optimization module 15 work together to achieve efficient optimization of semiconductor etching process parameters. The acquisition module 11 collects data in the etching process in real time and preprocesses the data to obtain a cleaned data set, which provides a reliable data basis for subsequent analysis. Based on the cleaned data set, the extraction module 12 uses a data mining algorithm to extract key features that affect the etching effect, and inputs these features into the prediction model to obtain the predicted value of the etching residual thickness, the predicted value of the etching side wall angle, and the predicted value of the etching rate deviation. These predicted values ​​provide a basis for the generation module 13 to generate an etching state weight matrix, and optimize and adjust the gas pressure distribution parameters based on the matrix to obtain the gas pressure setting value. The calculation module 14 further obtains the temperature distribution data of the etching chamber, combines the gas pressure setting value and the temperature distribution data, calculates the etching uniformity correction parameter, and adjusts the RF power distribution based on the parameter to generate the RF power setting value. The optimization module 15 optimizes the semiconductor etching process parameters using the RF power setting value and the gas pressure setting value, thereby improving the stability and precision of the etching process. The optimization device not only improves the efficiency of data processing, but also enhances the stability and consistency of the etching process, thereby achieving high-precision and high-efficiency etching parameter optimization.

[0117] It should be noted that technicians in the relevant technical field can clearly understand that for the convenience and conciseness of description, the specific working process of the above-described device and each module can refer to the corresponding process in the aforementioned semiconductor etching process parameter optimization method embodiment, and will not be repeated here.

[0118] Embodiment 4: The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the processor executes the semiconductor etching process parameter optimization method in Example 1.

[0119] In this embodiment, by storing a computer program for executing a semiconductor etching process parameter optimization method on a computer-readable storage medium, the processor can execute the optimization method described in Example 1 when running. Specifically, during the operation of the computer program, real-time data in the etching process is first obtained and preprocessed to obtain a cleaned data set. Then, the key features affecting the etching effect are extracted from the cleaned data set by a data mining algorithm, and these features are input into a preset prediction model to obtain a predicted value of the etching residual thickness, a predicted value of the etching sidewall angle, and a predicted value of the etching rate deviation. Based on these predicted values, an etching state weight matrix is ​​generated, and the gas pressure distribution parameters are optimized and adjusted to obtain a gas pressure setting value. Further obtain temperature distribution data, and calculate the etching uniformity correction parameter in combination with the gas pressure setting value, adjust the RF power distribution, and generate a RF power setting value. Finally, the semiconductor etching process parameters are optimized using the RF power setting value and the gas pressure setting value. Through this series of steps, the precise optimization of semiconductor etching process parameters is achieved, the stability and uniformity of the etching process are improved, and the manufacturing quality and performance of semiconductor devices are effectively improved.

[0120] The structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with this technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantive technical significance. Any structural modification, change in proportion or adjustment of size, without affecting the effects and purposes that can be achieved by the present invention, should still fall within the scope of the technical contents disclosed by the present invention.

[0121] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing semiconductor etching process parameters, characterized in that: include: Acquire real-time data of a semiconductor etching process during the etching process, and pre-process the real-time data to obtain a cleaned data set; Extracting key features that affect the etching effect based on the cleaned data set, inputting the key features into a preset prediction model to obtain a predicted value of the etching residual thickness, a predicted value of the etching sidewall angle, and a predicted value of the etching rate deviation; Generate an etching state weight matrix according to the etching residual thickness prediction value, the etching side wall angle prediction value and the etching rate deviation prediction value, and optimize and adjust the gas pressure distribution parameters of the etching chamber based on the etching state weight matrix to obtain a gas pressure setting value; Acquire temperature distribution data of the etching chamber, calculate the gas pressure setting value and the temperature distribution data to obtain an etching uniformity correction parameter, adjust the radio frequency power distribution based on the etching uniformity correction parameter, and obtain a radio frequency power setting value; The semiconductor etching process parameters are optimized using the radio frequency power setting value and the gas pressure setting value.

2. The method for optimizing semiconductor etching process parameters according to claim 1, characterized in that: The step of acquiring real-time data of the semiconductor etching process during the etching process, and preprocessing the real-time data to obtain a cleaned data set includes: Acquire the original data set in the etching process; wherein the original data set includes plasma emission spectrum data, etching chamber pressure data and etching layer material characteristic data; Performing frequency domain transformation on the plasma emission spectrum data to obtain plasma excitation characteristic distribution, and performing time series decomposition on the etching chamber pressure data to obtain dynamic fluctuation components; Calculating a short-period pressure fluctuation compensation value of the etching chamber based on the dynamic fluctuation component, and generating an etching condition adjustment factor based on the short-period pressure fluctuation compensation value, correcting the etching condition adjustment factor using the etching layer material characteristic data to obtain a material adaptation correction value, and calculating an etching surface charge accumulation parameter based on the material adaptation correction value to obtain a dynamic charge distribution during the etching process; Calculating a plasma energy density mapping matrix based on the etching surface charge accumulation parameter and the plasma excitation characteristic distribution, and fusing the plasma energy density mapping matrix with the short-period pressure fluctuation compensation value to obtain an etching stability factor and an etching rate compensation factor; The original data set is screened using the etching stability factor and the etching rate compensation factor, and the screened data set is normalized to obtain a cleaned data set.

3. The method for optimizing semiconductor etching process parameters according to claim 2, characterized in that: The step of extracting key features that affect the etching effect based on the cleaned data set, inputting the key features into a preset prediction model, and obtaining a predicted value of the etching residual thickness, a predicted value of the etching sidewall angle, and a predicted value of the etching rate deviation includes: Applying a data mining algorithm to extract key features that affect the etching effect from the cleaned data set; wherein the key features include etching chamber gas composition change data, radio frequency power real-time fluctuation data, and etching electrode potential distribution data; Calculating the radio frequency power deviation factor through the real-time radio frequency power fluctuation data, and calculating the dynamic response parameter of the etching electrode through the etching electrode potential distribution data; Calculating the plasma active particle density by using the radio frequency power offset factor and the etching electrode dynamic response parameter, and jointly analyzing the plasma active particle density and the etching chamber gas component change data to obtain an etching uniformity feature group; Constructing an etching region energy distribution matrix according to the etching uniformity feature group and etching chamber environmental parameters; wherein the etching chamber environmental parameters refer to parameters affecting the plasma reaction such as gas pressure parameters, temperature parameters, and gas flow parameters; The material removal rate characteristic group is calculated by the etching region energy distribution matrix and the etching layer material characteristic parameters to obtain the etching reaction region distribution characteristics, and the etching residual thickness prediction value is calculated based on the etching reaction region distribution characteristics; wherein the etching layer material characteristic parameters refer to the physical and chemical characteristics of the chemical composition, density parameters, and reactivity parameters of the material; Calculating the etching sidewall potential change based on the etching surface charge accumulation parameter, calculating the etching sidewall morphology parameter according to the etching sidewall potential change and the plasma active particle density, and generating the etching sidewall angle prediction value according to the etching sidewall morphology parameter; An etching rate adjustment coefficient is calculated based on the short-period pressure fluctuation compensation value, and the etching rate adjustment coefficient is weightedly fused with the etching uniformity feature group to obtain an etching rate deviation prediction value.

4. The method for optimizing semiconductor etching process parameters according to claim 3, characterized in that: The calculation formula for calculating the etching sidewall morphology parameters according to the etching sidewall potential change and the plasma active particle density is as follows: in, is the etching side wall morphology parameter, The potential change of the etching side wall is is the plasma active particle density, and is the constant obtained from experimental fitting.

5. The method for optimizing semiconductor etching process parameters according to claim 1, characterized in that: The step of generating an etching state weight matrix according to the etching residual thickness prediction value, the etching side wall angle prediction value and the etching rate deviation prediction value, and optimizing and adjusting the gas pressure distribution parameters of the etching chamber based on the etching state weight matrix to obtain the gas pressure setting value comprises: Calculating the etch layer thickness distribution weight using the residual thickness prediction value and the etch side wall angle prediction value, calculating the etch rate adjustment factor by combining the etch layer thickness distribution weight with the etch rate deviation prediction value, and generating an etch state weight matrix based on the etch rate adjustment factor; Calculating an etching layer thickness correction coefficient by using the etching state weight matrix, calculating an etching gas flow distribution parameter by using the etching layer thickness correction coefficient in combination with an etching condition adjustment factor, and generating an etching gas optimization parameter group according to the etching gas flow distribution parameter; The chamber internal gas uniformity adjustment factor is calculated based on the etching gas optimization parameter group, the etching chamber pressure correction value is calculated by the chamber internal gas uniformity adjustment factor and the etching rate adjustment factor, and the gas pressure setting value is generated according to the etching chamber pressure correction value.

6. The method for optimizing semiconductor etching process parameters according to claim 5, characterized in that: The calculation formula for calculating the correction value of the etching chamber pressure by using the chamber internal gas uniformity adjustment factor and the etching rate adjustment factor is as follows: in, is the correction value of the etching chamber pressure, is the gas uniformity adjustment factor inside the chamber, is the rate of change of the gas pressure in the etching chamber with time, is the etching rate adjustment factor, is the correction factor in the process.

7. The method for optimizing semiconductor etching process parameters according to claim 1, characterized in that: The step of obtaining the temperature distribution data of the etching chamber, calculating the gas pressure setting value and the temperature distribution data to obtain an etching uniformity correction parameter, and adjusting the RF power distribution based on the etching uniformity correction parameter to obtain the RF power setting value includes: Acquire temperature distribution data of the etching chamber through a sensor, calculate etching uniformity correction parameters using the gas pressure setting value and the temperature distribution data, calculate etching layer thermal response time parameters based on the etching uniformity correction parameters, and generate etching area temperature compensation factors according to the etching layer thermal response time parameters; The RF power correction value is calculated according to the etching area temperature compensation factor, the RF power dynamic adjustment parameter is calculated using the RF power correction value combined with the etching rate adjustment factor, and the RF power setting value is generated according to the RF power dynamic adjustment parameter.

8. The method for optimizing semiconductor etching process parameters according to claim 3, characterized in that: The step of optimizing the semiconductor etching process parameters by using the RF power setting value and the gas pressure setting value comprises: Calculating a plasma energy density correction parameter based on the RF power setting value and the gas pressure setting value, and calculating an etching morphology optimization parameter group according to the plasma energy density correction parameter combined with the etching sidewall morphology parameter; The semiconductor etching process parameters are optimized based on the etching morphology optimization parameter group.

9. A semiconductor etching process parameter optimization device, characterized in that: include: An acquisition module, used to acquire real-time data of the semiconductor etching process during the etching process, and pre-process the real-time data to obtain a cleaned data set; An extraction module is used to extract key features that affect the etching effect based on the cleaned data set, input the key features into a preset prediction model, and obtain a predicted value of the etching residual thickness, a predicted value of the etching sidewall angle, and a predicted value of the etching rate deviation; A generating module, used for generating an etching state weight matrix according to the etching residual thickness prediction value, the etching side wall angle prediction value and the etching rate deviation prediction value, and optimizing and adjusting the gas pressure distribution parameters of the etching chamber based on the etching state weight matrix to obtain a gas pressure setting value; A calculation module, used to obtain temperature distribution data of the etching chamber, calculate the gas pressure setting value and the temperature distribution data to obtain an etching uniformity correction parameter, adjust the RF power distribution based on the etching uniformity correction parameter, and obtain a RF power setting value; The optimization module is used to optimize the semiconductor etching process parameters by using the radio frequency power setting value and the gas pressure setting value.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the processor is enabled to execute the semiconductor etching process parameter optimization method as claimed in any one of claims 1 to 8.

Images