Method and device for adjusting etching parameters in production of MOSFET (metal oxide semiconductor field effect transistor) chip

Abstract

The invention provides an etching parameter adjusting method and device in MOSFET chip production, and relates to the technical field of semiconductors, and the method comprises the steps: obtaining plasma density distribution data and an etching rate monitoring value of a wafer surface, and dynamically dividing the wafer surface into a plurality of segment independent control units; adjusting a radio frequency power distribution weight, and synchronously adjusting a reaction gas injection parameter; carrying out coupling analysis on the concentration gradient of the etching by-product and the matching degree of the surface morphology of the wafer; and when the deviation of the target value threshold is detected, temperature compensation and air pressure compensation are carried out. According to the method and the device, the technical problem of poor etching precision of the MOSFET chip caused by non-uniform etching rates of different regions in the etching process is solved, and the uniformity of the etching rates is improved through real-time monitoring and dynamic compensation, so that the etching precision of the MOSFET chip is improved.

Description

Technical Field

[0001] The present application relates to the field of semiconductor technology, and in particular to a method and device for adjusting etching parameters in the production of MOSFET chips. Background Art

[0002] MOSFETs are a crucial component of modern semiconductor electronic devices. With the continuous improvement of chip integration and performance, the production process for MOSFET chips is becoming increasingly complex. Precisely controlling the etch rate and ensuring its uniformity is particularly crucial for improving the production quality and yield of MOSFET chips. During the etching process, uneven distribution of reactive gases, differences in equipment performance, and temperature and pressure fluctuations can lead to uneven etch rates across different regions. This uneven etch rate results in inconsistent material removal across different areas of the chip surface, leading to unstable electrical performance of the MOSFET chip, reducing overall chip performance, impacting yield, and increasing production costs. Furthermore, byproducts produced during the etching process often affect the etching effect, further exacerbating the uneven etch rate.

[0003] In summary, the existing technology has technical problems such as differences in etching rates between different areas due to the complexity of the etching process and the limitations of the equipment, resulting in poor etching accuracy, which in turn affects the performance and yield of MOSFET chips. Summary of the Invention

[0004] The purpose of this application is to provide a method and device for adjusting etching parameters in the production of MOSFET chips, so as to solve the technical problem in the prior art that due to the complexity of the etching process and the limitations of the equipment, the etching rate varies between different areas, resulting in poor etching accuracy, which in turn affects the performance and yield of the MOSFET chip.

[0005] In view of the above problems, the present application provides a method and device for adjusting etching parameters in the production of MOSFET chips.

[0006] In the first aspect, the present application provides an etching parameter adjustment method in MOSFET chip production, which is implemented by an etching parameter adjustment device in MOSFET chip production, wherein the etching parameter adjustment method in MOSFET chip production includes: based on the MOSFET chip, obtaining ion density distribution data and etching rate monitoring values ​​on the wafer surface, and dynamically dividing the RF power source into multiple independent segment control units based on the ion density distribution data; adjusting the RF power distribution weight according to the regional etching rate deviation values ​​of the multiple independent segment control units, and synchronously adjusting the reaction gas injection parameters, wherein the reaction gas injection parameters include a multi-nozzle array angle and a nozzle flow ratio; based on the reaction gas injection parameters and the etching rate monitoring values, coupling analysis is performed on the matching degree between the concentration gradient of the etching by-products and the wafer surface morphology, and the etching matching state is configured; at the same time, the three-dimensional distribution of the etching depth is monitored in real time, and when it is detected that the local etching depth deviates from the target value threshold, the temperature compensation unit is triggered to perform gradient heating on the edge area of ​​the wafer, and the pressure fluctuation in the etching chamber is dynamically compensated according to the spectral analysis characteristics of the etching by-products.

[0007] Optionally, a circular probe array is arranged on the top of the etching chamber to capture the spatial distribution of plasma electron temperature and ion density; the spatial distribution of plasma electron temperature and ion density is subjected to noise reduction processing to generate a plasma density thermogram; density difference areas are marked in the plasma density thermogram to divide the priority zones for RF power regulation.

[0008] Optionally, taking the etching rate uniformity as the objective function, the adaptive power ratio of the multiple segment independent control units is determined, and a distributed power control architecture is constructed, wherein the distributed power control architecture has multiple edge fog nodes and a control center, and each edge fog node is used to collect the control parameters of the corresponding segment independent control unit, including plasma density, etching rate, and RF power.

[0009] Optionally, a rotatable microfluidic nozzle is provided, and the deflection angle of each microfluidic nozzle is dynamically adjusted in step units; the Cl2 / O2 mixed gas flow ratio of each microfluidic nozzle is determined based on the mass spectrometry analysis results of the etching by-products and the thickness deviation of the sidewall passivation layer.

[0010] Optionally, a wavelength tunable laser light source is connected to alternately scan the wafer surface with dual wavelengths to extract the phase offset of the interference fringes and invert the resolution three-dimensional distribution of the etching depth; based on the resolution inversion three-dimensional distribution, when it is detected that the etching depth gradient exceeds a preset proportional section of the upper limit of the etching depth, the ion energy compensation electrode is activated to balance the plasma sheath potential.

[0011] Optionally, a cavity resonance frequency offset of the etching chamber is introduced; and based on the pressure fluctuation range in the etching chamber and the cavity resonance frequency offset, the pumping speed of the turbo pump is adjusted.

[0012] Optionally, the stress wave signal on the wafer surface is collected, and the wavelet packet transform is used to extract the energy value of the characteristic frequency band related to the microcracks in the stress wave spectrum; when the energy value of the characteristic frequency band exceeds a preset energy threshold, the RF power in the control parameter is adjusted.

[0013] Optionally, in the etching termination stage, the residual polymer components at the bottom of the etching are identified; based on the residual polymer components, the characteristic absorption peak positions and relative intensity ratios of the residual CF bonds and the residual Si-O bonds are determined; and the RF power in the control parameters is feedback optimized through the characteristic absorption peak positions and relative intensity ratios of the residual CF bonds and the residual Si-O bonds.

[0014] Optionally, a thermal expansion sensor array is integrated at the bottom of the etching chamber to determine the local thermal resistance change rate of the fixture contact interface; the coherence coefficient between the deformation gradient distribution and the etching heat load is determined through the local thermal resistance change rate of the fixture contact interface and the deformation gradient distribution of the wafer support structure, and a machine learning model based on historical failure data is used to perform failure risk judgment, and the judgment results are synchronized to the control center.

[0015] In the second aspect, the present application also provides an etching parameter adjustment device in the production of MOSFET chips, which is used to execute the etching parameter adjustment method in the production of MOSFET chips as described in the first aspect, wherein the etching parameter adjustment device in the production of MOSFET chips includes: a dynamic segmentation module for obtaining the ion density distribution data and etching rate monitoring value on the wafer surface based on the MOSFET chip, and dynamically dividing the RF power source into multiple independent control units according to the ion density distribution data; a synchronous adjustment module for adjusting the RF power source according to the regional etching rate deviation values ​​of the multiple independent control units. The power distribution weight is adjusted and the reaction gas injection parameters are adjusted synchronously, and the reaction gas injection parameters include the multi-nozzle array angle and the nozzle flow ratio; the coupling analysis module is used to perform coupling analysis on the matching degree between the concentration gradient of the etching by-products and the wafer surface morphology based on the reaction gas injection parameters and the etching rate monitoring value, and configure the etching matching state; the dynamic compensation module is used to simultaneously monitor the three-dimensional distribution of the etching depth in real time. When it is detected that the local etching depth deviates from the target value threshold, the temperature compensation unit is triggered to perform gradient heating on the edge area of ​​the wafer, and the pressure fluctuation in the etching chamber is dynamically compensated according to the spectral analysis characteristics of the etching by-products.

[0016] One or more technical solutions provided in this application have at least the following beneficial effects:

[0017] Based on the MOSFET chip, the ion density distribution data and etching rate monitoring value of the wafer surface are obtained, and the RF power source is dynamically divided into multiple independent control units according to the ion density distribution data; the RF power distribution weight is adjusted according to the regional etching rate deviation values ​​of the multiple independent control units, and the reaction gas injection parameters are adjusted synchronously, and the reaction gas injection parameters include the multi-nozzle array angle and the nozzle flow ratio; based on the reaction gas injection parameters and the etching rate monitoring value, the concentration gradient of the etching by-products and the matching degree of the wafer surface morphology are coupled and analyzed, and the etching matching state is configured; at the same time, the three-dimensional distribution of the etching depth is monitored in real time. When it is detected that the local etching depth deviates from the target value threshold, the temperature compensation unit is triggered to perform gradient heating on the edge area of ​​the wafer, and the pressure fluctuation in the etching chamber is dynamically compensated according to the spectral analysis characteristics of the etching by-products. In other words, by dynamically dividing the RF power source and precisely adjusting the nozzle flow ratio of the reaction gas, the uniformity of the etching rate is significantly improved, the etching differences between different areas on the wafer surface are reduced, and real-time dynamic compensation of gas pressure and temperature is achieved, effectively reducing the impact of these factors on the etching depth. The RF power and reaction gas injection parameters are precisely controlled, the etching process is optimized, the accumulation of by-products is reduced, and precise control of the etching process is achieved. The uniformity of the etching rate is improved, thereby improving the etching accuracy and improving the efficiency and product quality of MOSFET chip production.

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, which can be implemented in accordance with the contents of the description, and to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically listed below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in this application or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely exemplary, and a person of ordinary skill in the art can obtain other drawings based on the provided drawings without creative work.

[0020] Figure 1 This is a flow chart of the etching parameter adjustment method in the production of MOSFET chips in this application;

[0021] Figure 2This is a schematic diagram of the structure of the etching parameter adjustment device in the production of MOSFET chips in this application.

[0022] Description of reference numerals: dynamic segmentation module 11 , synchronous adjustment module 12 , coupling analysis module 13 , dynamic compensation module 14 . DETAILED DESCRIPTION

[0023] This application solves the technical problem in the prior art that, due to the complexity of the etching process and the limitations of the equipment, the etching rate varies between different areas, resulting in poor etching accuracy, which in turn affects the performance and yield rate of the MOSFET chip by providing an etching parameter adjustment method and device in the production of MOSFET chips. By dynamically dividing the RF power source and accurately adjusting the nozzle flow ratio of the reaction gas, the uniformity of the etching rate is significantly improved, the etching difference between different areas of the wafer surface is reduced, and real-time dynamic compensation of gas pressure and temperature is achieved, effectively reducing the influence of these factors on the etching depth. The RF power and reaction gas injection parameters are accurately controlled, the etching process is optimized, the accumulation of by-products is reduced, and precise control of the etching process is achieved. The uniformity of the etching rate is improved, thereby improving the etching accuracy and improving the efficiency and product quality of MOSFET chip production.

[0024] Below, the technical solutions in this application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this application, rather than all the embodiments of this application. It should be understood that this application is not limited to the example embodiments described herein. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application. It should also be noted that, for the convenience of description, only the parts related to this application, rather than all of them, are shown in the accompanying drawings.

[0025] For example, see the attached Figure 1 The present application provides a method for adjusting etching parameters in MOSFET chip production, wherein the method for adjusting etching parameters in MOSFET chip production is performed by an etching parameter adjustment device in MOSFET chip production, and the method for adjusting etching parameters in MOSFET chip production specifically includes the following steps:

[0026] S100: Based on the MOSFET chip, the plasma density distribution data and the etching rate monitoring value on the wafer surface are obtained, and according to the plasma density distribution data, the radio frequency power source is dynamically divided into multiple independent control units.

[0027] Specifically, a MOSFET chip is a metal oxide semiconductor field effect transistor, a semiconductor component commonly used in integrated circuits, primarily for switching or amplifying electronic signals. In the etching process of MOSFET chips, uniform distribution of ion density is a key factor in ensuring etching quality and uniformity. Acquiring ion density distribution data on the wafer surface involves real-time monitoring of ion density in different areas of the wafer surface using sensors and probe technology. This data is typically acquired using a plasma density detector or electron beam probe. These detection instruments can provide precise ion density information, helping to analyze the etching conditions in different areas. Ion density distribution data represents the distribution of ions on the wafer surface. The higher the ion density, the faster the etching rate. Based on this ion density distribution data, it is possible to identify areas on the wafer surface with higher ion density (i.e., faster etching rate) and lower ion density (i.e., slower etching rate).

[0028] At the same time, the etching rate monitoring value is also collected in real time to evaluate the efficiency of the etching process. The etching rate monitoring value refers to the material removal rate of each area during the etching process, usually expressed in terms of how much material is removed per second (such as nanometers / second). According to the ion density distribution data, the RF power source is dynamically divided into multiple independent control units. Specifically, the entire RF power source no longer supplies power to the entire etching chamber in a uniform power manner, but dynamically adjusts the distribution of the power source according to the ion density in different areas, dividing it into multiple sections. The power of each section can be controlled independently, which means that the RF power is increased in areas with lower ion density to increase the excitation intensity of the ions and compensate for the low density in the area; while in areas with higher ion density, the RF power is reduced to avoid over-etching.

[0029] During the etching process, an RF power source is used to excite the gas and generate plasma to achieve etching. By dynamically dividing the RF power source into multiple independent control units, the etching process is precisely controlled and the uniformity of the etching rate is improved.

[0030] S200: adjusting the radio frequency power distribution weight according to the regional etching rate deviation values ​​of the multiple segment independent control units, and synchronously adjusting the reaction gas injection parameters, wherein the reaction gas injection parameters include a multi-nozzle array angle and a nozzle flow ratio.

[0031] Specifically, during the etching process of MOSFET chips, the etching chamber is divided into multiple areas, and each area has an independent control unit to accurately adjust the etching conditions of the area, such as RF power and gas injection parameters. The etching rate data of each area in the etching chamber is obtained through the etching rate monitoring unit, and the etching rate deviation value of each sub-area is calculated based on the actual measurement results of the etching rate. The deviation value represents the difference between the actual etching rate and the target rate of the area. If the etching rate in some areas is too high or too low, these deviations are identified and corresponding measures are taken to correct them. The etching rate deviation value refers to the difference between the actual etching rate and the expected etching rate in different areas. By detecting the etching rate of each sub-area, it is identified which areas have rate deviations, and corresponding measures are taken to adjust them.

[0032] According to the etching rate deviation of each area, the RF power distribution weight is dynamically adjusted. For areas with large etching rate deviations, the RF power of the area is increased to provide more energy to accelerate the etching rate of the area; for areas with small or too high etching rate deviations, the RF power is reduced to prevent excessive etching or uneven etching. In other words, based on the density difference in the plasma density thermogram, areas with low or high density are marked. Then, the etching chamber is divided into multiple sections, and each section has an independent control unit to adjust the RF power and reaction gas injection of the corresponding area. By comparing the deviation between the etching rate of each area and the target rate, the RF power distribution weight is dynamically adjusted according to the deviation value. If the etching rate of a certain area is lower than the target value, the RF power of the area is increased, otherwise, the RF power is reduced.

[0033] According to the regional etching rate deviation, the angle of the multi-nozzle array and the nozzle flow ratio are adjusted. By changing the angle of the nozzle, the distribution of the gas in the etching area is adjusted to ensure that the gas can evenly cover the entire area, and avoid excessive concentration of gas in certain areas and affecting the etching effect. For example, when the etching rate of a certain area is low, the nozzle angle is tilted toward that area to increase the gas supply. Different types of reaction gases (such as Cl2, O2, etc.) will have different effects on the etching process. According to the feedback information of the etching rate, the flow ratio of each gas is dynamically adjusted to ensure that the ratio of the reaction gas is most suitable for the current etching state. By precisely adjusting the gas flow and RF power distribution in each area, the etching rate of each area is ensured to be consistent, thereby achieving uniform etching of the entire wafer.

[0034] By adjusting RF power and gas injection parameters in real time, each area during the etching process can more uniformly achieve the predetermined etching rate target. Real-time adjustment of etching conditions in each area reduces etching non-uniformity and improves etching quality and efficiency. The combination of edge mist nodes and microfluidic nozzles enables more refined gas distribution and power regulation, further improving the flexibility and accuracy of the etching process.

[0035] S300: Based on the reaction gas injection parameters and the etching rate monitoring value, a coupling analysis is performed on the matching degree between the concentration gradient of the etching by-products and the wafer surface morphology, and an etching matching state is configured.

[0036] Specifically, the etch rate monitoring value is the etch rate data of each area, which reflects the etch rate of a certain area on the wafer surface and identifies which areas have too high or too low etch rates. The reaction gas injection parameters refer to factors such as the type, flow rate, ratio, nozzle angle, etc. of the gas injected into the etching chamber during the etching process, including the gas flow ratio and the nozzle array angle. By monitoring the etch rate on the wafer surface and combining the injection parameters of the reaction gas, the distribution of the gas in the etching chamber is simulated, and the by-product concentration distribution in different areas is predicted. Excessive or small by-product concentration gradients may lead to uneven etching and even affect the quality of the wafer surface.

[0037] By analyzing the concentration changes of byproducts in the etching chamber, a byproduct concentration gradient distribution map is drawn to identify non-uniformities in the etching process. The concentration gradient refers to the concentration variation of byproducts at different locations in the etching chamber. The change in concentration gradient will affect the etching process. Excessively high or low concentrations will have a negative impact on the etching effect.

[0038] Wafer surface topography refers to the shape of the wafer surface after the etching process, including indicators such as surface flatness, texture, and roughness. Surface topography is an important indicator of etching quality, and etching non-uniformity will affect the subsequent processing and final function of the wafer. The uniformity and accuracy of etching directly affect the quality of the surface topography, and excessive etching may cause surface damage or morphological distortion. A coupling analysis is performed on the matching degree of the by-product concentration gradient and the surface topography to determine the areas with non-uniform by-product concentration, and these areas are compared with the changes in the wafer surface topography. Changes in surface topography (such as surface roughness and unevenness) are often closely related to the non-uniformity of the etching rate, differences in gas distribution, and the concentration gradient of the by-products.

[0039] By calculating the correlation between the byproduct concentration gradient and the surface morphology, we can determine whether the current etching state is close to the ideal state. If the byproduct concentration gradient is large and the surface morphology is uneven, it means that the current etching process is not optimally matched and the etching conditions need to be adjusted.

[0040] According to the results of the coupling analysis, the reaction gas injection parameters (such as nozzle flow ratio, nozzle array angle, etc.) are dynamically adjusted to optimize the distribution of by-products. Based on the etching rate monitoring value, by-product concentration gradient and surface morphology data, the relevant control parameters (such as RF power, gas flow, nozzle angle, etc.) are adjusted to achieve the etching matching state. In other words, by adjusting various parameters, the etching rate is made more uniform, the by-product concentration gradient is controlled, and the wafer surface morphology reaches the optimal state, thereby improving the etching quality. The etching matching state refers to the process in which the reaction gas injection parameters, RF power, etc. are adjusted to achieve the best matching state between the by-product concentration gradient and the wafer surface morphology during the etching process, thereby ensuring the efficiency and uniformity of the etching process. The ideal etching matching state is that all factors are in the optimal balance state, thereby ensuring the efficiency and high quality of the etching process.

[0041] By analyzing the matching degree between by-product concentration and etching rate, parameters such as gas flow rate and nozzle angle are adjusted in real time to reduce the unevenness in the etching process, improve etching accuracy, avoid local over-etching or under-etching caused by uneven by-product concentration, improve the morphology quality of the wafer surface, and ensure the consistency of etching rates in different areas.

[0042] S400: At the same time, the three-dimensional distribution of the etching depth is monitored in real time. When it is detected that the local etching depth deviates from the target value threshold, the temperature compensation unit is triggered to perform gradient heating on the edge area of ​​the wafer, and the pressure fluctuation in the etching chamber is dynamically compensated according to the spectral analysis characteristics of the etching by-products.

[0043] Specifically, the three-dimensional distribution of etch depth refers to the spatial distribution of etch depth at different locations on the wafer surface during the etching process. Etch depth refers to the vertical depth to which material is removed during the etching process, and the three-dimensional distribution represents the variation in etch depth across the wafer surface, taking into account differences at different locations. This is obtained through advanced surface monitoring technologies, such as optical interferometers and scanning probe microscopes, which accurately measure the etch depth on the wafer surface and generate a three-dimensional distribution map of the etch depth, reflecting the removal of different areas of the wafer surface during the etching process.

[0044] When the etching depth deviation is detected in real-time monitoring, the control center automatically identifies the area that deviates from the target value and triggers a series of compensation actions. The deviation threshold is set based on the process requirements and equipment accuracy. Usually, a target depth range is set, and the allowable deviation range is defined for each area. When the etching depth of a certain area exceeds this range, the control center will initiate compensation measures. During the etching process, the edge area of ​​the wafer may cause deviations in the etching depth due to uneven heat. For example, the etching rate in the edge area of ​​the wafer may be slower than that in the center area, resulting in uneven etching depth. To correct this problem, the temperature compensation unit will perform gradient heating on the edge area.

[0045] By using a heating device (such as an electric heater or laser heating), the control center precisely adjusts the temperature of the edge area to keep the temperature within the ideal range, thereby avoiding changes in the etching rate caused by temperature fluctuations. Temperature compensation is achieved by calculating the heat load of different areas and performing local heating to maintain consistent temperature and ensure the stability of the etching process. The temperature compensation unit refers to a device or module used to adjust the surface temperature of the wafer during the etching process. By controlling the heating element or cooling element, the temperature of the wafer is kept stable during the etching process to prevent temperature fluctuations from adversely affecting the etching results. The purpose of heating the edge area is to correct the difference in etching depth caused by uneven temperature.

[0046] Etching byproducts are continuously generated during the etching process, and their concentration directly affects the uniformity of the etching rate. The concentration and type of byproducts are monitored and analyzed in real time. Different etching byproducts absorb specific wavelengths of light, and by measuring these absorption characteristics allows precise determination of their concentration. When abnormal byproduct concentrations are detected, the control center activates a dynamic pressure compensation mechanism. Through a real-time feedback mechanism, the temperature compensation unit, pressure regulation unit, and spectral analysis unit work together. During the etching process, even small fluctuations in pressure can affect plasma uniformity, leading to uneven etching depth. After real-time monitoring of pressure fluctuations, the turbo pump speed is dynamically adjusted based on the feedback signal. Simultaneously, temperature fluctuations in the edge area are adjusted by the temperature compensation unit to avoid etching rate variations caused by thermal inhomogeneity. By integrating real-time monitoring and dynamic compensation technology, the effects of temperature and pressure fluctuations, as well as byproduct concentration, during the etching process are effectively controlled, resulting in more uniform etching depth.

[0047] Furthermore, the present application S200 includes:

[0048] A circular probe array is arranged at the top of the etching chamber to capture the spatial distribution of plasma electron temperature and ion density; the spatial distribution of plasma electron temperature and ion density is subjected to noise reduction processing to generate a plasma density heat map; and density difference areas are marked in the plasma density heat map to divide the priority zones for RF power regulation.

[0049] Specifically, the etching chamber is an enclosed space in the semiconductor manufacturing process, where processing processes such as etching are carried out to remove materials from the wafer through media such as plasma to form circuit patterns. A ring probe array is arranged on the top of the etching chamber. These probes are used to capture the spatial distribution of electron temperature and ion density in the plasma in real time. The ring probe array is an array of multiple probes arranged on the top of the etching chamber, arranged in a ring, and is used to monitor the electron temperature and ion density of the plasma. The plasma electron temperature is the average energy or temperature of the electrons in the plasma, which usually affects the chemical activity and etching rate of the plasma; the spatial distribution of ion density is the distribution of ions in the plasma in space, and the ion density affects the reaction rate during the etching process.

[0050] Since actual measurements may be affected by noise and interference, such as electronic interference or unstable signals introduced by environmental factors, it is necessary to perform noise reduction on the collected data to generate more accurate plasma density thermograms. The purpose of noise reduction is to remove noise caused by measurement errors, environmental interference, and other factors, resulting in smoother and more accurate data.

[0051] The spatial distribution of plasma electron temperature and ion density is subjected to noise reduction. Kalman filtering is used to smooth the real-time data of plasma electron temperature and ion density. Noise fluctuations are reduced by calculating the average value within a certain range around the data point. The median of the data within the window is taken to eliminate extreme values, thereby reducing the impact of noise. After noise reduction, the original electron temperature and ion density data are more accurate and reflect the actual distribution of the plasma. Using the noise-reduced spatial distribution of plasma electron temperature and ion density, a plasma density heat map is generated. This visualizes the ion density data by spatial location, typically using color coding to indicate high and low density. For example, high-density areas are represented by red, and low-density areas by blue. This heat map provides a visual representation of the plasma distribution within the etching chamber, demonstrating differences in spatial distribution.

[0052] In the plasma density heat map, density difference areas are marked. These areas are caused by uneven etching rates due to differences in plasma density. The purpose of marking density difference areas is to distinguish high-density areas from low-density areas in the heat map and identify key areas that require power adjustment. For example, a threshold value can be preset based on actual needs to automatically identify areas where the density difference exceeds the preset threshold. In the plasma density heat map, based on the distribution of density difference areas, areas that require priority adjustment are marked. These marked areas are often the cause of uneven etching rates, so it is necessary to adjust the RF power allocation strategy and perform targeted optimization on these areas.

[0053] Prioritize RF power adjustment based on the marked areas. In low-density areas (such as the blue areas in the heat map), the plasma density is insufficient, resulting in a slow etching rate and uneven etching or over-etching. Therefore, the RF power should be increased to increase the ion density in these areas and enhance the etching reaction. In moderate-density areas (such as the yellow areas in the heat map), the ion density in these areas is within the normal range, and the RF power can be maintained at the standard value to ensure a normal etching rate. In high-density areas (such as the red areas in the heat map), the ion density in these areas is already high. Excessive RF power may lead to over-etching, so the RF power should be reduced to avoid over-activation of the plasma, which can lead to excessive etching. Priority should be set for RF power adjustment in different plasma density areas based on factors such as density differences and etching effects.

[0054] For example, assuming that the plasma density heat map obtained after noise reduction processing shows that the density of area A is 1.5×10 12 m -3 , which is a high-density area; the density of area B is 0.8×10 12 m -3 , is a low-density area, and the density of area C is 1.0×10 12 m -3 . By marking the density difference areas, area B (low-density area) and area A (high-density area) are marked as key adjustment areas. For area B, it is necessary to increase the RF power first to increase the ion density and ensure the etching rate. The RF power is set to increase by 10%-15% first; for area A, there is already a high ion density, and excessive power may cause over-etching. The RF power needs to be reduced by 5%-10% to prevent excessive etching; for area C, since the density is moderate, the RF power can be kept at a normal value and no special adjustment is required.

[0055] By placing a circular probe array on the top of the etching chamber and performing noise reduction processing on the data, the spatial distribution of the plasma can be accurately captured and analyzed, a plasma density thermogram can be generated, and the ion density and electron temperature distribution in different areas can be intuitively presented. The density difference areas can be marked and the RF power can be adjusted to prioritize the areas. The power distribution during the etching process can be precisely controlled, thereby effectively improving the uniformity of the etching process, reducing local over-etching or under-etching, and ultimately improving the production yield and chip performance.

[0056] Furthermore, the present application further comprises the following steps:

[0057] Taking the etching rate uniformity as the objective function, the adaptive power ratio of the multiple segment independent control units is determined, and a distributed power control architecture is constructed, wherein the distributed power control architecture has multiple edge fog nodes and a control center, and each edge fog node is used to collect the control parameters of the corresponding segment independent control unit, including plasma density, etching rate, and RF power.

[0058] Specifically, the etching rate uniformity is used as the objective function, that is, the uniformity of the etching rate in each area during the entire etching process. Etching rate uniformity is an important indicator in the etching process. Higher uniformity can avoid performance differences due to uneven etching on the chip surface and ensure the functional consistency of the chip. In other words, during the entire etching process, the etching rates of each area should be as close as possible to avoid over-etching or under-etching. Determine the adaptive power ratio of multiple independent control units in each section. The adaptive power ratio refers to ensuring that the power allocation of each area is adapted to the needs of the area when allocating the RF power source to different areas. The power allocation of each area will be dynamically adjusted according to the changes in the ion density and etching rate to achieve uniformity of the overall etching rate. During the etching process, the adaptive power ratio of each section is dynamically calculated based on the monitored values ​​of the ion density and etching rate of each area.

[0059] A distributed power control architecture is constructed, consisting of multiple edge fog nodes and a control center. Each edge fog node is responsible for collecting relevant data for its segment, including control parameters such as plasma density, etch rate, and RF power, and performing preliminary analysis and processing. The processed data is transmitted to the control center, which makes adjustment decisions based on the global data, instructing each edge node to adjust the local RF power to maintain etch rate uniformity.

[0060] Each edge fog node is responsible for collecting control parameters such as plasma density, etching rate, and RF power in the corresponding area, reflecting the etching status of each area and the power that needs to be adjusted. Each edge fog node performs preliminary processing on the collected data, including noise reduction and smoothing, to obtain the etching rate deviation of each segment, that is, the difference between the actual etching rate and the ideal target rate. Based on the deviation of the etching rate, the edge fog node will calculate the adaptive power ratio that needs to be adjusted, which is equivalent to the aforementioned. For example, if the etching rate in a certain area is too high (over-etching), the RF power in that area will need to be appropriately reduced; conversely, if the etching rate in a certain area is too low (under-etching), the RF power needs to be increased.

[0061] All edge fog nodes transmit their processed data to the control center, which integrates feedback from each edge fog node and performs global power optimization. The control center analyzes the power requirements and etching rates of each segment and adjusts the RF power allocation weights of each edge fog node to achieve uniformity throughout the etching process. After global optimization, the control center transmits the new power configuration and adjustment plan back to each edge fog node. Each edge fog node readjusts the RF power based on the new power allocation, further uniformizing the etching rate across each region. Throughout the entire process, the etching rate and ion density of each region are continuously monitored to ensure uniformity. If the rate deviation in certain areas remains large, the edge fog node will make further adjustments to ensure dynamic adaptation of the process.

[0062] Each edge fog node regularly collects the control parameters of the plasma density, etching rate and RF power of the corresponding segment and sends this data to the control center. The control center dynamically adjusts the RF power of each segment based on this data to maintain the uniformity of the etching rate and achieve precise control of the independent control unit in each segment, thereby improving the uniformity of the etching rate.

[0063] Furthermore, the present application further comprises the following steps:

[0064] A rotatable microfluidic nozzle is set up, and the deflection angle of each microfluidic nozzle is dynamically adjusted in step units. The Cl2 / O2 mixed gas flow ratio of each microfluidic nozzle is determined based on the mass spectrometry analysis results of the etching by-products and the thickness deviation of the sidewall passivation layer.

[0065] Specifically, a rotatable microfluidic nozzle is provided for precisely controlling the injection of fluids (such as gases or liquids), which is commonly used in the etching process in chip manufacturing. The nozzle can be rotated to change the angle of the nozzle, thereby precisely controlling the injection direction and coverage. A plurality of rotatable microfluidic nozzles are provided in the etching equipment, and the rotation angle of each nozzle can be dynamically adjusted in step units. By adjusting the angle of the nozzle, the injection direction of the gas can be precisely controlled, thereby achieving a more uniform gas distribution. The rotatability of the microfluidic nozzle enables it to adapt to the needs of different areas during the etching process, especially when faced with complex geometric shapes. The step unit is the minimum angular unit of nozzle rotation, usually a fixed value, such as 1 degree or 0.5 degrees. The deflection angle is the rotation angle of the microfluidic nozzle in the horizontal or vertical direction. By adjusting the deflection angle of the nozzle, the coverage and direction of the gas injection can be controlled to achieve a uniform etching effect.

[0066] The etching process generates various byproducts that may affect the etching effect. These byproducts are chemicals produced during the etching process due to gas reactions, material removal, and other processes. These can affect the etching effect and even cause unnecessary contamination or accumulation. The composition and concentration of these byproducts are monitored in real time using mass spectrometry, an analytical technique used to measure the mass and composition of substances.

[0067] In the etching chamber, the etching gas reacts with the wafer material, generating various byproducts. These byproducts are then directed to a mass spectrometer, where ionization reactions convert them into charged ions, and their masses and relative abundances are measured. The mass spectrometer converts the byproducts into ions using an ionization source and separates them according to their mass-to-charge ratio (m / z) using a mass analyzer (such as a quadrupole or time-of-flight analyzer). The analyzer generates a mass spectrum based on the mass-to-charge ratio of each ion, with distinct peaks corresponding to different byproduct components. The mass spectrometer detects the byproducts and generates mass spectrometry results, including the mass peaks (m / z values) and their intensities. The intensities of these peaks correlate with the byproduct concentrations, reflecting the amount of byproducts generated during the etching process. By analyzing the dominant peaks in the mass spectrum, it is possible to determine which byproducts are produced most frequently during the etching process. For example, if a peak is very intense, this indicates a high concentration of that byproduct, and adjustments to the reaction gas ratio or etching conditions may be necessary to reduce byproduct generation.

[0068] Based on the results of mass spectrometry analysis, the problem of by-product accumulation that may occur during the etching process or the insufficient removal efficiency of the by-products can be identified. For example, when the concentration of certain by-products is detected to be too high, it may mean that the etching process in a certain area is uneven, or the ratio of the etching gas is inappropriate. At the same time, the thickness of the sidewall passivation layer may be uneven during the etching process, resulting in the etching rate of the sidewall area being different from that of other areas during the etching process. The sidewall passivation layer thickness deviation (that is, the gap between the actual passivation layer and the target thickness) is monitored in real time to determine whether the etching is uniform. If a part of the passivation layer is too thick or too thin, it may indicate that the etching effect in that area is not ideal.

[0069] Based on mass spectrometry analysis results and passivation layer thickness deviations, the Cl2 / O2 mixed gas flow ratio at each microfluidic nozzle is dynamically adjusted to optimize the etching process. Different gas ratios affect etching rate and uniformity. For example, Cl2 is commonly used to remove silicon materials, while O2 helps remove byproducts. By optimizing the gas flow ratio, the efficiency and uniformity of the etching process can be ensured while preventing the accumulation of byproducts that negatively impact etching quality. Cl2 (chlorine) and O2 (oxygen) are commonly used etching gases, each used to etch different types of materials. The Cl2 / O2 mixed gas flow ratio refers to the specific ratio in which these two gases are mixed and injected into the etching chamber during the etching process. Adjusting the gas flow ratio can optimize the etching effect. For example, if the passivation layer thickness on a sidewall is detected to be thick, the O2 ratio can be increased to enhance the oxidation reaction and remove the excess passivation layer. Conversely, if the passivation layer is thin, the Cl2 ratio can be increased to enhance the etching reaction.

[0070] By continuously adjusting the deflection angle of the microfluidic nozzle, combined with feedback from mass spectrometry analysis and passivation layer thickness, the localized etching area can be precisely controlled during the etching process. The gas flow ratio (Cl2 / O2) of each nozzle is also dynamically adjusted based on real-time data to ensure etching uniformity and accuracy. Each area in the etching process can be independently adjusted according to its specific needs. This precise local adjustment can effectively reduce problems such as uneven etching, over-etching, or under-etching.

[0071] Furthermore, the present application S400 includes:

[0072] Connect a wavelength tunable laser light source, scan the wafer surface alternately with dual wavelengths, extract the phase offset of the interference fringes, and invert the resolution three-dimensional distribution of the etching depth; based on the resolution inversion three-dimensional distribution, when it is detected that the etching depth gradient exceeds the preset proportional section of the etching depth upper limit, activate the ion energy compensation electrode to balance the plasma sheath potential.

[0073] Specifically, a wavelength-tunable laser light source can change the output laser wavelength. By adjusting the wavelength, experiments can be conducted under different optical characteristics. For example, by selecting a specific wavelength to scan the surface of an object, different reflection, refraction, or scattering information can be obtained. Dual-wavelength alternating scanning is a technical means of using two different wavelengths of laser light to alternately scan the target surface. Lasers of different wavelengths can provide different detection depths and reflection characteristics. By alternating these two wavelengths, the accuracy of etching depth detection can be improved.

[0074] A wavelength-tunable laser light source is used to perform alternating dual-wavelength scanning on the wafer surface. By varying the laser wavelength, interference fringes are generated when the laser beam strikes the wafer surface, reflecting changes in surface height, i.e., the etching depth. By analyzing the phase offset of the interference fringes, etching depth information can be obtained at different locations on the wafer surface. Changes in the phase offset reflect changes in surface topography, enabling precise measurement of the etching depth in each area. The high-resolution three-dimensional distribution of etching depth is a spatial distribution map of etching depth derived from the results of laser interferometry during the etching process. This provides etching depth information at each location and can demonstrate depth variations across the entire wafer surface.

[0075] During the etching process, by real-time monitoring of the three-dimensional distribution of etching depth, it is detected which areas have an etching depth exceeding the preset upper limit. Once it is detected that the etching depth gradient in certain areas exceeds the set threshold (i.e., the depth variation is too large), the ion energy compensation electrode is triggered. The ion energy compensation electrode adjusts the plasma potential to change the energy of ions bombarding the wafer surface, thereby compensating for the unevenness caused by local etching that is too deep or too shallow. By controlling the plasma potential, the compensation electrode makes the activated ion energy more uniform, thereby affecting the etching rate of each area, avoiding local over-etching or under-etching caused by excessively high or low ion energy, and ensuring high uniformity of the etching process.

[0076] The plasma sheath potential is the potential difference between the plasma and the material surface, which typically affects the energy and movement of ions. Uneven plasma sheath potential during the etching process can lead to inconsistent local etching rates, affecting the overall etching quality. For example, suppose that during the etching process, local over-etching occurs on the wafer surface, causing the etching depth in a certain area to exceed the set upper limit. In this case, the laser scanning system will detect the gradient change in the etching depth in this area, and the inverted three-dimensional distribution map will show the abnormal depth in this area. The ion energy compensation electrode is automatically triggered to adjust the potential to reduce the impact of the ion energy in this area, thereby returning the etching depth in this area to the predetermined value and ensuring the uniformity of the etching process. Through alternating dual-wavelength laser scanning and interference fringe phase offset analysis, the etching depth is accurately measured, the depth changes during the etching process are monitored in real time, and the etching accuracy is optimized.

[0077] Furthermore, the present application further comprises the following steps:

[0078] A cavity resonance frequency offset of the etching chamber is introduced; and a pumping speed of the turbo pump is adjusted based on the pressure fluctuation range in the etching chamber and the cavity resonance frequency offset.

[0079] Specifically, during the etching process, the etching chamber may experience a shift in resonant frequency due to factors such as changes in air pressure and gas flow. The resonant frequency refers to the frequency of the natural vibration of the cavity system under specific conditions. If the resonant frequency of the cavity shifts, it may affect the etching process, such as causing plasma instability, thereby affecting the accuracy and quality of etching. The resonant frequency of the cavity is monitored in real time using devices such as vibration sensors or optical sensors. If the resonant frequency shifts, it means that the vibration mode of the cavity is affected by changes in air pressure or other factors. The cavity resonant frequency offset represents the difference between the cavity resonant frequency and the ideal value.

[0080] The pressure in the etching chamber may fluctuate due to factors such as gas consumption during the etching reaction and the injection of external gases. Excessive pressure fluctuations can affect plasma stability, leading to uneven etching rates. Therefore, it is necessary to monitor the pressure fluctuation range and adjust the turbo pump speed in real time to maintain stable pressure.

[0081] Based on the monitored cavity resonant frequency offset and pressure fluctuation range, the turbo pump's pumping speed is adjusted by increasing or decreasing the pump's operating speed to control the pressure level within the etching chamber. If a shift in the cavity's resonant frequency is detected, or if the pressure fluctuation is excessive, the turbo pump's pumping speed is adjusted to stabilize the pressure within the cavity. Increasing the turbo pump's pumping speed accelerates the discharge of gas from the cavity, helping to reduce pressure fluctuations and restore the stability of the cavity's resonant frequency; reducing the pumping speed slows the gas discharge rate, helping to maintain stable pressure within the cavity and thus avoiding plasma instability caused by excessively low pressure. By combining the cavity resonant frequency offset and pressure fluctuation range, the turbo pump's pumping speed is dynamically adjusted to suppress instability caused by pressure fluctuations and resonant frequency offset, thereby improving the stability and uniformity of the etching process.

[0082] The turbo pump is a commonly used gas extraction device in the etching chamber. It generates airflow through high-speed rotating turbine blades to extract waste gas or excess gas from the chamber. The pumping speed of the turbo pump refers to the volume of gas that can be extracted per second, usually measured in standard liters per minute (SLM) or liters per second (L / s). Adjusting the pumping speed of the turbo pump helps stabilize the air pressure and gas flow in the etching chamber. By introducing the monitoring of the cavity resonant frequency offset and the range of air pressure fluctuation, potential unstable factors in the etching process can be captured in real time, the pumping speed of the turbo pump can be dynamically adjusted, and the air pressure stability in the chamber can be effectively controlled, thereby suppressing the uneven etching problem caused by air pressure fluctuations and resonant frequency offset.

[0083] Furthermore, the present application further comprises the following steps:

[0084] The stress wave signal on the wafer surface is collected, and the energy value of the characteristic frequency band related to the microcracks in the stress wave spectrum is extracted using wavelet packet transform; when the energy value of the characteristic frequency band exceeds a preset energy threshold, the radio frequency power in the control parameter is adjusted.

[0085] Specifically, the stress wave signal on the wafer surface is collected, that is, the stress fluctuation on the surface of the wafer material due to external force, thermal stress or other factors, which is usually collected through sensors (such as piezoelectric sensors, accelerometers, etc.). The stress wave signal can reflect the changes in the wafer caused by microcracks, physical stress and other factors during etching or processing. The collected stress wave signal is input into the wavelet packet transform algorithm. By decomposing the signal into multiple frequency bands, the energy of the signal can be analyzed in different frequency ranges. Wavelet packet transform is a signal processing technology that decomposes the signal into multiple frequency bands and analyzes the signal at different scales to reveal the high-frequency and low-frequency characteristics in the signal as well as local abnormal signals. Through wavelet packet transform, the spectral information of the signal is extracted and the characteristic frequency bands related to microcracks are identified. Usually, the energy value of microcracks in the stress wave spectrum will be concentrated in certain specific frequency bands.

[0086] Wavelet packet transform is used to identify characteristic frequency bands associated with microcracks. These frequency bands are typically generated during crack propagation or material damage. By analyzing the energy values ​​of these frequency bands, the energy intensity of these characteristic frequency bands within the spectrum is determined. The stress wave spectrum represents the stress wave signal in the frequency domain. After wavelet packet transform, the resulting stress wave spectrum contains multiple frequency bands. The presence of microcracks typically produces stronger signals in certain frequency bands. By analyzing the energy values ​​of each frequency band in the spectrum, characteristic frequency bands associated with microcracks are identified. The energy values ​​of these characteristic frequency bands are then extracted. When microcracks occur on the wafer surface, stress wave signals within specific frequency ranges are typically generated. The energy values ​​within these frequency ranges can be used as signatures of the microcracks. By analyzing the energy values ​​of these characteristic frequency bands, the presence of microcracks on the wafer surface can be assessed.

[0087] A preset energy threshold is set to determine whether there are microcracks on the wafer surface. When the energy value of the characteristic frequency band exceeds this threshold, it means that microcracks may have appeared on the wafer surface, and the etching process needs to be adjusted at this time to avoid further damage or uneven etching. Once the energy value of the characteristic frequency band of the microcrack is detected to exceed the preset threshold, the RF power in the control parameters is adjusted, such as increasing the RF power or decreasing the RF power. For example, if microcracks exist and the etching process needs to be accelerated, the RF power is appropriately increased to increase the plasma density and etching rate; if the appearance of microcracks leads to excessive etching or damage, the RF power is reduced to control the progress and depth of etching to avoid further damage to the wafer surface. RF power is an important parameter for regulating the etching process. By adjusting its size, the characteristics of the plasma can be changed, thereby controlling the etching process.

[0088] Through real-time monitoring of stress wave signals on the wafer surface and wavelet packet transform analysis, potential problems can be discovered in a timely manner before microcracks form or expand. Based on the energy changes in the detected microcrack characteristic frequency bands, the RF power can be automatically adjusted to stabilize the etching process and reduce etching unevenness or quality fluctuations caused by wafer damage.

[0089] Furthermore, the present application further comprises the following steps:

[0090] During the etching termination stage, the residual polymer components at the bottom of the etching are identified; based on the residual polymer components, the characteristic absorption peak positions and relative intensity ratios of the residual CF bonds and the residual Si-O bonds are determined; and the RF power in the control parameters is feedback optimized through the characteristic absorption peak positions and relative intensity ratios of the residual CF bonds and the residual Si-O bonds.

[0091] Specifically, in the etching termination stage, that is, after the main etching task is completed on the wafer surface, the stage of identifying and removing residual substances (such as polymers, etching by-products, etc.) begins. Infrared spectroscopy is used to identify the residual polymer components at the bottom of the etching. By performing spectral analysis on the material at the bottom of the etching, possible polymer components such as fluorine-containing substances or silicon-containing substances are identified. The residues of these components will affect the subsequent etching effect and the cleanliness of the equipment. The residual polymer components are usually polymers produced by by-products in the etching reaction, chemical reactions of the reaction gases, or ion collisions. The CF bond refers to the chemical bond between carbon and fluorine, and the Si-O bond refers to the chemical bond between silicon and oxygen, namely fluorine-containing polymers (CF) and oxygen-containing silicon polymers (Si-O).

[0092] For the residual CF bonds and Si-O bonds, the characteristic absorption peak positions and their relative intensity ratios are obtained by infrared spectroscopy. For example, the absorption peak of the CF bond usually appears at about 1200 to 1400 cm -1The absorption peak of Si-O bond usually appears in the range of 1000 to 1100 cm -1 Range. Different intensity ratios reflect the relative amounts of CF and Si-O, helping to analyze the specific composition of the polymer. By analyzing the position of these absorption peaks and their intensity ratios, the composition ratio of the residue can be determined, and then it can be evaluated whether there are still too many fluorides or silicon oxides remaining at the bottom of the etching. In other words, within the wavelength range corresponding to the CF bond and Si-O bond, the absorption peak positions of these two bonds are found respectively. By accurately locating the wavelengths of these absorption peaks and comparing their relative intensity ratios, the chemical composition of the residue can be further understood. In the infrared spectrum, the characteristic absorption peaks of the CF bond and Si-O bond have a fixed wavelength range. By detecting the position of these peaks, the specific composition of the residue can be determined. The relative intensity ratio of the characteristic absorption peaks refers to the ratio of the absorption intensities of different components in spectral analysis. By measuring the absorption intensity ratio of the CF bond and the Si-O bond, the chemical composition and characteristics of the residue can be further analyzed.

[0093] The type and proportion of the residue present can be determined based on the position and relative intensity ratio of the characteristic absorption peaks for C—F bonds and Si—O bonds obtained through spectral analysis. Excessive C—F bonds indicate excessive fluorinated polymer residues; a high Si—O bond ratio suggests the presence of silicon-oxygen residues. If there are too many C—F bonds, the plasma activity can be reduced by lowering the RF power, thereby minimizing fluoride formation. If the Si—O bond ratio is high, the RF power can be increased to facilitate the removal of silicon-oxygen polymers, or the gas composition can be adjusted to optimize the reaction conditions.

[0094] For example, suppose that at the end of a certain etching process, the characteristic absorption spectrum of the residue at the bottom of the etching is obtained by FTIR analysis. The analysis results show that the characteristic absorption peak of the C-F bond is located at 1200 cm -1 , with an intensity of 0.8, and the characteristic absorption peak of Si-O bond is located at 1100 cm -1 , with an intensity of 0.3. This result shows that the intensity of CF is significantly higher than that of Si-O, indicating that the bottom residue is primarily composed of fluorinated polymers. Based on this analysis, feedback was provided to adjust the RF power: the original RF power was 200W, and the adjusted RF power was 150W (reducing the RF power to reduce CF residue). By reducing the RF power, the formation of fluorinated polymers was effectively reduced, ensuring that the bottom residue was fully removed.

[0095] By introducing the identification of residual polymer components at the etching termination stage and adjusting the RF power based on the characteristic absorption peak positions and relative intensity ratios of CF bonds and Si-O bonds, the residue removal effect during the etching process is significantly optimized, avoiding the problem of excessive polymer residue affecting subsequent processes, while improving the uniformity and accuracy of etching.

[0096] Furthermore, the present application further comprises the following steps:

[0097] A thermal expansion sensor array is integrated at the bottom of the etching chamber to determine the local thermal resistance change rate of the fixture contact interface; the coherence coefficient between the deformation gradient distribution and the etching heat load is determined through the local thermal resistance change rate of the fixture contact interface and the deformation gradient distribution of the wafer support structure, and a machine learning model based on historical failure data is used to determine the failure risk, and the determination results are synchronized to the control center.

[0098] Specifically, a group of thermal expansion sensor arrays are integrated at the bottom of the etching chamber, and these sensors are used to obtain the local thermal resistance change rate of the fixture contact interface in real time. Since the temperature changes more drastically during the etching process, these thermal expansion sensors can monitor the local thermal resistance changes caused by temperature changes, and then reflect the impact of the heat load on the equipment during the etching process. The thermal expansion sensor is used to monitor the temperature changes at the bottom of the etching chamber or the fixture contact interface in real time to help determine whether the heat conduction is uniform. The data obtained by the thermal expansion sensor is used to calculate the change rate of the local thermal resistance of the fixture contact interface. The change in local thermal resistance is related to the contact condition between the wafer and the fixture and the heat load generated during the etching process. If the thermal resistance changes significantly, it means that there is a problem with the contact interface, which affects the etching effect.

[0099] Sensors mounted on the support structure measure the deformation gradient distribution of the wafer support structure. Uneven thermal loads can cause deformation in the support structure, impacting wafer stability. Analysis of deformation gradients helps determine the impact of thermal load distribution on the support structure. The deformation gradient distribution refers to the degree and spatial distribution of deformation of the wafer support structure caused by thermal load or other factors during the etching process. This influences the contact between the wafer and the etching equipment, and thus, the etching accuracy.

[0100] By analyzing the correlation between the deformation gradient distribution and the thermal load, the coherence coefficient between the two is calculated. The coherence coefficient measures the correlation between the thermal load and the support structure deformation. A higher coherence coefficient means that the thermal load has a greater impact on the support structure deformation, affecting the uniformity and quality of the etching. Based on historical failure data (such as equipment failure records, failure modes, temperature fluctuations, etc. during previous etching processes), a machine learning model is trained to predict the potential failure risks during the etching process. By analyzing the correlation patterns appearing in the historical data and combining them with real-time monitored thermal expansion data and deformation gradient information, the potential failure risks during the etching process are determined.

[0101] Collect historical failure data related to etching equipment, including the time of equipment failure, failure type, operating environment, and process parameters at the time of failure. Perform data preprocessing on this historical failure data. Select a suitable machine learning model, such as a decision tree, to predict equipment failure. Train the decision tree algorithm on data such as equipment temperature, power, and air pressure to predict equipment failure. The goal of model training is to analyze historical data and identify key failure-causing factors from these input features. The trained machine learning model needs to be validated to ensure high predictive accuracy. Divide the dataset into training and validation sets, performing training and validation separately to avoid overfitting. Evaluate model performance using metrics such as precision, recall, and F1 score. Continuously adjust model parameters and optimize the algorithm to ensure that the model maintains good predictive capabilities under varying operating conditions. After training and validation, apply the model to actual production processes. By analyzing relevant patterns in the historical data and combining them with real-time thermal expansion data and deformation gradient information, potential failure risks in the etching process can be determined.

[0102] When the machine learning model determines that there is a potential failure risk in the etching process, the result will be sent synchronously to the control center. Based on the failure risk determination results, the control center can adjust relevant control parameters (such as RF power, reaction gas flow, temperature regulation, etc.) to avoid or mitigate the failure risk. For example, when it is detected that the support structure may be excessively deformed, the control center can trigger automatic adjustment of the support structure or change the thermal loading strategy. By integrating a thermal expansion sensor array and a deformation monitoring system at the bottom of the etching chamber, and combining the machine learning model to predict the failure risk in the etching process, the stability and reliability of the etching process can be significantly improved.

[0103] In summary, the etching parameter adjustment method in MOSFET chip production provided by this application has the following beneficial effects:

[0104] Based on the MOSFET chip, the ion density distribution data and etching rate monitoring value of the wafer surface are obtained, and the RF power source is dynamically divided into multiple independent control units according to the ion density distribution data; the RF power distribution weight is adjusted according to the regional etching rate deviation values ​​of the multiple independent control units, and the reaction gas injection parameters are adjusted synchronously, and the reaction gas injection parameters include the multi-nozzle array angle and the nozzle flow ratio; based on the reaction gas injection parameters and the etching rate monitoring value, the concentration gradient of the etching by-products and the matching degree of the wafer surface morphology are coupled and analyzed, and the etching matching state is configured; at the same time, the three-dimensional distribution of the etching depth is monitored in real time. When it is detected that the local etching depth deviates from the target value threshold, the temperature compensation unit is triggered to perform gradient heating on the edge area of ​​the wafer, and the pressure fluctuation in the etching chamber is dynamically compensated according to the spectral analysis characteristics of the etching by-products. In other words, by dynamically dividing the RF power source and precisely adjusting the nozzle flow ratio of the reaction gas, the uniformity of the etching rate is significantly improved, the etching differences between different areas on the wafer surface are reduced, and real-time dynamic compensation of gas pressure and temperature is achieved, effectively reducing the impact of these factors on the etching depth. The RF power and reaction gas injection parameters are precisely controlled, the etching process is optimized, the accumulation of by-products is reduced, and precise control of the etching process is achieved. The uniformity of the etching rate is improved, thereby improving the etching accuracy and improving the efficiency and product quality of MOSFET chip production.

[0105] Example 2: Based on the same inventive concept as the etching parameter adjustment method in the production of MOSFET chips in the above-mentioned Example 1, this application also provides an etching parameter adjustment device in the production of MOSFET chips, please refer to the attached Figure 2 , the etching parameter adjustment device in the production of MOSFET chips includes:

[0106] A dynamic segmentation module 11 is used to obtain ion density distribution data and etching rate monitoring values ​​on the wafer surface based on the MOSFET chip, and dynamically divide the RF power source into multiple independent control units according to the ion density distribution data; a synchronous adjustment module 12 is used to adjust the RF power distribution weight according to the regional etching rate deviation values ​​of the multiple independent control units, and synchronously adjust the reaction gas injection parameters, wherein the reaction gas injection parameters include the multi-nozzle array angle and the nozzle flow ratio; a coupling analysis module 13 is used to perform a coupling analysis on the matching degree between the concentration gradient of the etching by-products and the wafer surface morphology based on the reaction gas injection parameters and the etching rate monitoring values, and configure the etching matching state; a dynamic compensation module 14 is used to simultaneously monitor the three-dimensional distribution of the etching depth in real time. When it is detected that the local etching depth deviates from the target value threshold, the temperature compensation unit is triggered to perform gradient heating on the edge area of ​​the wafer, and the pressure fluctuation in the etching chamber is dynamically compensated according to the spectral analysis characteristics of the etching by-products.

[0107] Furthermore, the synchronous adjustment module 12 in the etching parameter adjustment device in the production of MOSFET chips is also used to: arrange a ring probe array on the top of the etching chamber to capture the spatial distribution of plasma electron temperature and ion density; perform noise reduction processing on the spatial distribution of plasma electron temperature and ion density to generate a plasma density thermogram; mark density difference areas in the plasma density thermogram to divide the priority zones for RF power adjustment.

[0108] Furthermore, the synchronous adjustment module 12 in the etching parameter adjustment device in the MOSFET chip production is also used to: determine the adaptive power ratio of the multiple segment independent control units with the etching rate uniformity as the objective function, and construct a distributed power control architecture, wherein the distributed power control architecture has multiple edge fog nodes and a control center, and each edge fog node is used to collect the control parameters of the corresponding segment independent control unit including plasma density, etching rate, and RF power.

[0109] Furthermore, the synchronous adjustment module 12 in the etching parameter adjustment device in the production of the MOSFET chip is also used to: set a rotatable microfluidic nozzle, and the deflection angle of each microfluidic nozzle is dynamically adjusted in step units; according to the mass spectrometry analysis results of the etching by-products and combined with the side wall passivation layer thickness deviation, determine the Cl2 / O2 mixed gas flow ratio of each microfluidic nozzle.

[0110] Furthermore, the dynamic compensation module 14 in the etching parameter adjustment device in the production of the MOSFET chip is also used to: connect a wavelength tunable laser light source, alternately scan the wafer surface with dual wavelengths, extract the phase offset of the interference fringes, and invert the resolution three-dimensional distribution of the etching depth; based on the resolution inversion three-dimensional distribution, when it is detected that the etching depth gradient exceeds the preset proportional section of the upper limit of the etching depth, activate the ion energy compensation electrode to balance the plasma sheath potential.

[0111] Furthermore, the dynamic compensation module 14 in the etching parameter adjustment device in the MOSFET chip production is also used to: introduce the cavity resonance frequency offset of the etching chamber; and adjust the pumping speed of the turbine pump based on the gas pressure fluctuation range in the etching chamber and the cavity resonance frequency offset.

[0112] Furthermore, the synchronous adjustment module 12 in the etching parameter adjustment device in the production of MOSFET chips is also used to: collect stress wave signals on the wafer surface, and use wavelet packet transform to extract the energy value of the characteristic frequency band related to microcracks in the stress wave spectrum; when the energy value of the characteristic frequency band exceeds a preset energy threshold, adjust the RF power in the control parameter.

[0113] Furthermore, the etching parameter adjustment device in the production of MOSFET chips also includes a termination feedback module, which is also used to: identify the residual polymer components at the bottom of the etching during the etching termination stage; determine the characteristic absorption peak positions and relative intensity ratios of the residual CF bonds and the residual Si-O bonds based on the residual polymer components; and perform feedback optimization on the RF power in the control parameters through the characteristic absorption peak positions and relative intensity ratios of the residual CF bonds and the residual Si-O bonds.

[0114] Furthermore, the etching parameter adjustment device in the production of MOSFET chips also includes a failure judgment module, which is also used to: integrate a thermal expansion sensor array at the bottom of the etching chamber to determine the local thermal resistance change rate of the fixture contact interface; determine the coherence coefficient between the deformation gradient distribution and the etching heat load through the local thermal resistance change rate of the fixture contact interface and the deformation gradient distribution of the wafer support structure, and use a machine learning model based on historical failure data to perform failure risk judgment, and synchronize the judgment results to the control center.

[0115] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 1 The etching parameter adjustment method and specific examples in the production of MOSFET chips in Example 1 are also applicable to the etching parameter adjustment device in the production of MOSFET chips in this embodiment. Through the above detailed description of the etching parameter adjustment method in the production of MOSFET chips, those skilled in the art can clearly understand the etching parameter adjustment device in the production of MOSFET chips in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here.

[0116] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

[0117] Obviously, for those skilled in the art, several improvements and modifications can be made to the present application without departing from the principles of the present application, and these improvements and modifications also fall within the scope of protection of the present application.

Claims

1. A method for adjusting etching parameters in MOSFET chip production, characterized in that: include: Based on the MOSFET chip, the plasma density distribution data and etching rate monitoring value on the wafer surface are obtained, and according to the plasma density distribution data, the radio frequency power source is dynamically divided into multiple independent control units; According to the regional etching rate deviation values ​​of the multiple segment independent control units, the RF power distribution weight is adjusted, and the reaction gas injection parameters are simultaneously adjusted, wherein the reaction gas injection parameters include the multi-nozzle array angle and the nozzle flow ratio; Based on the reaction gas injection parameters and the etch rate monitoring value, a coupling analysis is performed on the matching degree between the concentration gradient of the etching by-products and the wafer surface morphology, and an etching matching state is configured; At the same time, the three-dimensional distribution of etching depth is monitored in real time. When it is detected that the local etching depth deviates from the target value threshold, the temperature compensation unit is triggered to perform gradient heating on the edge area of ​​the wafer, and the pressure fluctuations in the etching chamber are dynamically compensated based on the spectral analysis characteristics of the etching by-products.

2. The etching parameter adjustment method in the production of MOSFET chips according to claim 1, characterized in that: Adjusting the radio frequency power distribution weight according to the regional etching rate deviation values ​​of the multiple segment independent control units further includes: A ring probe array is placed on the top of the etching chamber to capture the spatial distribution of plasma electron temperature and ion density; performing noise reduction processing on the plasma electron temperature and ion density spatial distribution to generate a plasma density thermodynamic map; Density difference areas are marked in the plasma density thermodynamic map to divide the priority zones for radio frequency power regulation.

3. The etching parameter adjustment method in the production of MOSFET chips according to claim 2, characterized in that: include: Taking the etching rate uniformity as the objective function, the adaptive power ratio of the multiple segment independent control units is determined, and a distributed power control architecture is constructed, wherein the distributed power control architecture has multiple edge fog nodes and a control center, and each edge fog node is used to collect the control parameters of the corresponding segment independent control unit, including plasma density, etching rate, and RF power.

4. The etching parameter adjustment method in the production of MOSFET chips according to claim 1, characterized in that: A rotatable microfluidic nozzle is provided, and the deflection angle of each microfluidic nozzle is dynamically adjusted in step units; According to the mass spectrometry analysis results of the etching by-products and the thickness deviation of the sidewall passivation layer, the Cl2 / O2 mixed gas flow ratio of each microfluidic nozzle is determined.

5. The etching parameter adjustment method in the production of MOSFET chips according to claim 4, characterized in that: Connect a wavelength tunable laser light source to alternately scan the wafer surface with dual wavelengths, extract the phase offset of the interference fringes, and invert the three-dimensional distribution of the etching depth resolution; Based on the resolution inversion three-dimensional distribution, when it is detected that the etching depth gradient exceeds a preset proportional section of the etching depth upper limit, the ion energy compensation electrode is activated to balance the plasma sheath potential.

6. The etching parameter adjustment method in the production of MOSFET chips according to claim 5, characterized in that: Dynamically compensating for pressure fluctuations in the etching chamber based on the spectral analysis characteristics of etching by-products, including: Introducing a cavity resonance frequency offset into the etching chamber; The pumping speed of the turbo pump is adjusted based on the pressure fluctuation range in the etching chamber and the resonance frequency offset of the chamber.

7. The etching parameter adjustment method in the production of MOSFET chips according to claim 3, characterized in that: Each edge fog node is used to collect control parameters of the corresponding segment independent control unit, including plasma density, etching rate, and RF power, as well as: Collect stress wave signals on the wafer surface and use wavelet packet transform to extract the energy value of characteristic frequency bands related to microcracks in the stress wave spectrum; When the energy value of the characteristic frequency band exceeds a preset energy threshold, the radio frequency power in the control parameter is adjusted.

8. The etching parameter adjustment method in the production of MOSFET chips according to claim 7, characterized in that: Also includes: During the etch stop phase, the residual polymer components at the bottom of the etch are identified; Based on the residual polymer components, determining the characteristic absorption peak positions and relative intensity ratios of the residual CF bonds and the residual Si-O bonds; The radio frequency power in the control parameter is feedback-optimized based on the characteristic absorption peak positions and relative intensity ratios of the residual CF bond and the residual Si-O bond.

9. The etching parameter adjustment method in the production of MOSFET chips according to claim 3, characterized in that: Also includes: An array of thermal expansion sensors is integrated at the bottom of the etching chamber to determine the rate of change of local thermal resistance at the fixture contact interface. The coherence coefficient between the deformation gradient distribution and the etching heat load is determined by the local thermal resistance change rate of the fixture contact interface and the deformation gradient distribution of the wafer support structure, and a machine learning model based on historical failure data is used to determine the failure risk, and the determination results are synchronized to the control center.

10. An etching parameter adjustment device for MOSFET chip production, characterized in that: The method for adjusting etching parameters in the production of MOSFET chips according to any one of claims 1 to 9 is implemented, wherein the etching parameter adjusting device in the production of MOSFET chips comprises: A dynamic segmentation module is used to obtain plasma density distribution data and etching rate monitoring values ​​on the wafer surface based on the MOSFET chip, and dynamically segment the radio frequency power source into multiple independent control units based on the plasma density distribution data; a synchronous adjustment module, configured to adjust the RF power distribution weight according to the regional etching rate deviation values ​​of the multiple segment independent control units, and synchronously adjust the reaction gas injection parameters, wherein the reaction gas injection parameters include the multi-nozzle array angle and the nozzle flow ratio; A coupling analysis module, configured to perform a coupling analysis on the matching degree between the concentration gradient of etching byproducts and the wafer surface topography based on the reaction gas injection parameters and the etching rate monitoring value, and configure an etching matching state; The dynamic compensation module is used to simultaneously monitor the three-dimensional distribution of etching depth in real time. When it is detected that the local etching depth deviates from the target value threshold, the temperature compensation unit is triggered to perform gradient heating on the edge area of ​​the wafer, and the pressure fluctuations in the etching chamber are dynamically compensated based on the spectral analysis characteristics of the etching by-products.

Images