Deep silicon etching control method, deep silicon etching control system and etching equipment

Abstract

The present invention provides a deep silicon etching control method, a deep silicon etching control system and an etching device. The method obtains spectral signals of a deposition product and an etching product in one cycle of a Bosch process at a preset sampling frequency, calculates an average of the spectral signals of several deposition products to obtain an average of the spectral signals of the several deposition products to form a deposition product spectral line, calculates an average of the spectral signals of several etching products to obtain an average of the spectral signals of the several etching products to form an etching product spectral line, calculates a ratio of the two spectral lines to obtain a characteristic value, calculates a normalized deviation relative to a target ratio, compares the deviation with preset parameters, and adjusts the control parameters of the next cycle of the Bosch process according to the comparison result, thereby realizing dynamic adjustment of the control parameters in the Bosch process, achieving more precise etching, and reducing the disadvantages brought by fixed passivation and etching switching time, chemical gas flow rate and bias power.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor process technology, and in particular to a deep silicon etching control method, a deep silicon etching control system and etching equipment. Background Art

[0002] Deep Silicon Etching (DSE) is one of the key technologies in semiconductor device manufacturing and packaging, and usually adopts the Bosch process.

[0003] The Bosch process achieves high-depth-ratio silicon trench or hole structures by alternating deposition and etching, and uses fixed passivation and etching switching times, chemical gas flow rates, and bias power. However, as the etching depth increases, the etching behavior changes, mainly manifested in the following aspects:

[0004] 1. Change in etching rate: As the etching depth increases, the transmission efficiency of reactants and side reactants decreases, resulting in nonlinear attenuation of the etching rate;

[0005] 2. Changes in structure and sidewalls. Increased etching depth and changes in etching rate may lead to inconsistent bottom structure and sidewall roughness of silicon trenches or holes.

[0006] 3. Changes in etching uniformity. High depth ratio structures lead to significant differences in etching rates between the edge and center areas of the wafer.

[0007] Therefore, it is necessary to provide a novel deep silicon etching control method, deep silicon etching control system and etching equipment to solve the above problems existing in the prior art. Summary of the Invention

[0008] The purpose of the present invention is to provide a deep silicon etching control method, a deep silicon etching control system and an etching device to realize dynamic adjustment of the control parameters in the Bosch process, achieve more precise etching, and reduce the disadvantages brought about by fixed passivation and etching switching time, chemical gas flow rate and bias power.

[0009] To achieve the above objectives, in a first aspect, the present invention provides a deep silicon etching control method applied to the Bosch process, comprising:

[0010] Acquire the spectral signals of the deposition product and the etched product in one cycle of the Bosch process at a preset sampling frequency;

[0011] Dividing the spectral signals of a plurality of continuous deposition products into a group, and the spectral signals of two adjacent groups of deposition products share the spectral signals of a plurality of deposition products, averaging the spectral signals of a group of deposition products to obtain the spectral signal average of the plurality of deposition products, wherein the spectral signal average of the plurality of deposition products constitutes a deposition product spectral line;

[0012] Dividing the spectrum signals of a plurality of continuous etching products into a group, wherein the spectrum signals of two adjacent groups of etching products share the spectrum signals of a plurality of etching products, averaging the spectrum signals of a group of etching products to obtain the average value of the spectrum signals of the plurality of etching products, wherein the average value of the spectrum signals of the plurality of etching products constitutes an etching product spectrum line;

[0013] Calculating a ratio based on the deposition product spectrum line and the etching product spectrum line to obtain a characteristic value;

[0014] Obtaining a normalized deviation of the characteristic value relative to a target ratio;

[0015] The normalized deviation is compared with the preset parameters, and the control parameters of the next cycle of the Bosch process are adjusted according to the comparison result.

[0016] The beneficial effects of the deep silicon etching control method are as follows: the spectral signals of the deposition products and the spectral signals of the etching products in one cycle of the Bosch process are obtained at a preset sampling frequency, the spectral signals of several continuous deposition products are divided into one group, and the spectral signals of two adjacent groups of deposition products share the spectral signals of several deposition products, the spectral signals of one group of deposition products are averaged to obtain the spectral signal average of several deposition products, the spectral signal averages of several deposition products constitute the deposition product spectral line, the spectral signals of several continuous etching products are divided into one group, and the spectral signals of two adjacent groups of etching products share the spectral signals of several etching products. The method comprises the following steps: calculating an average of spectral signals of a group of etching products to obtain an average of spectral signals of several etching products, the average of spectral signals of several etching products constituting an etching product spectral line, calculating a ratio according to the deposition product spectral line and the etching product spectral line to obtain a characteristic value, calculating a normalized deviation of the characteristic value relative to a target ratio, comparing the normalized deviation with a preset parameter, and adjusting the control parameters of the next cycle of the Bosch process according to the comparison result, thereby realizing dynamic adjustment of the control parameters in the Bosch process, realizing more precise etching, and reducing the disadvantages brought by fixed passivation and etching switching time, chemical gas flow rate and bias power.

[0017] Optionally, the deposition product includes CF2, and the etching product includes SiF.

[0018] Optionally, the deep silicon etching control method also includes: determining the preset sampling frequency based on the time of one cycle of the Bosch process, the sampling frequency is 15ms / time to 25ms / time, the time of one cycle of the Bosch process is 2s to 5s, and the sampling frequency is proportional to the time of one cycle of the Bosch process.

[0019] Optionally, obtaining a normalized deviation of the characteristic value relative to a target ratio includes:

[0020] A target ratio is preset, a difference between the characteristic value and the target ratio is calculated, and then a ratio between the difference and the target ratio is calculated to obtain a normalized deviation of the characteristic value relative to the target ratio.

[0021] Optionally, the preset parameters include a first preset sub-parameter and a second preset sub-parameter, the first preset sub-parameter is greater than the second preset sub-parameter, and the control parameters include a DEP phase time, an ET1 phase time, an ET2 phase time, a chemical gas flow rate in the DEP phase, a chemical gas flow rate in the ET phase, and a bias power;

[0022] When the normalized deviation is greater than the first preset parameter, increasing the time of the DEP stage, decreasing the time of the ET1 stage, decreasing the time of the ET2 stage, increasing the chemical gas flow rate of the DEP stage, decreasing the chemical gas flow rate of the ET stage, and decreasing the bias power;

[0023] When the normalized deviation is less than the second preset parameter, reducing the time of the DEP stage, increasing the time of the ET1 stage, increasing the time of the ET2 stage, reducing the chemical gas flow rate of the DEP stage, increasing the chemical gas flow rate of the ET stage, and increasing the bias power;

[0024] When the normalized deviation is less than or equal to the first preset parameter and greater than or equal to the second preset parameter, the time of the DEP stage is increased, the time of the ET1 stage is reduced, the time of the ET2 stage is reduced, the chemical gas flow rate of the DEP stage is increased, the chemical gas flow rate of the ET stage is reduced, and the bias power is maintained.

[0025] Optionally, when the normalized deviation is greater than the first preset parameter, the increased time of the DEP stage, the reduced time of the ET1 stage, the reduced time of the ET2 stage, the increased chemical gas flow rate of the DEP stage, the reduced chemical gas flow rate of the ET stage, and the reduced bias power are respectively proportional to the difference between the normalized deviation and the first preset parameter;

[0026] When the normalized deviation is less than the second preset parameter, the reduced time of the DEP stage, the increased time of the ET1 stage, the increased time of the ET2 stage, the reduced chemical gas flow rate of the DEP stage, the increased chemical gas flow rate of the ET stage, and the increased bias power are respectively proportional to the difference between the normalized deviation and the second preset value;

[0027] When the normalized deviation is less than or equal to the first preset parameter and greater than or equal to the second preset parameter, the time increased in the DEP stage, the time reduced in the ET1 stage, the time reduced in the ET2 stage, the chemical gas flow increased in the DEP stage, and the chemical gas flow reduced in the ET stage are respectively inversely proportional to the difference between the normalized deviation and the first preset parameter.

[0028] Optionally, the first preset parameter is 0.075 to 0.125, and the second preset parameter is -0.075 to -0.125.

[0029] Optionally, the spectral signals of a plurality of continuous deposition products are divided into a group, and the spectral signals of two adjacent groups of deposition products share the spectral signals of a plurality of deposition products, and the spectral signals of a group of deposition products are averaged to obtain the average of the spectral signals of the plurality of deposition products, including:

[0030] According to the acquisition order of the spectral signals of the plurality of deposition products, the spectral signals of the plurality of deposition products are divided into one group, and the spectral signals of the deposition products of two adjacent groups share the spectral signals of the plurality of deposition products;

[0031] Taking the spectral signal of the current deposition product and the spectral signals of the m deposition products sorted after it as a group, an average is calculated to obtain the mean value of the spectral signals of the plurality of deposition products;

[0032] Among them, after the spectral signal of the last deposition product participates in the averaging for the first time, the step of averaging the spectral signals of the deposition products is completed, or, if the number of spectral signals of the deposition products sorted later is less than m, the average value of the missing number is calculated, or, after the number of spectral signals of the deposition products is supplemented, the average value is calculated, and m is a natural number greater than 1.

[0033] Optionally, the number of spectral signals of the deposition product is supplemented and then an average value is obtained, including:

[0034] The number of spectral signals of the deposition product is padded with data 0 so that the number of spectral signals of the deposition product for which an average value is calculated is m, and then the average value is calculated.

[0035] Optionally, the number of spectral signals of the deposition product is supplemented and then an average value is obtained, including:

[0036] The number of spectral signals of the deposition product is supplemented by the spectral signal of the current deposition product, so that the number of spectral signals of the deposition product for which the average value is calculated is m, and then the average value is calculated.

[0037] Optionally, the spectral signals of a plurality of continuous etching products are divided into a group, and the spectral signals of two adjacent groups of etching products share the spectral signals of a plurality of etching products, and the spectral signals of a group of etching products are averaged to obtain the average of the spectral signals of the plurality of etching products, including:

[0038] According to the acquisition order of the spectral signals of the etching products, the spectral signals of the etching products are divided into a group, and the spectral signals of the etching products of two adjacent groups share the spectral signals of the etching products;

[0039] Taking the spectrum signal of the current etching product and the spectrum signals of the m etching products sorted later as a group, an average value is calculated to obtain the average value of the spectrum signals of the plurality of etching products;

[0040] Among them, after the spectral signal of the last etching product participates in the averaging for the first time, the step of averaging the spectral signals of the etching products is completed, or, if the number of spectral signals of the etching products sorted later is less than m, the average value of the missing number is calculated, or, after the number of spectral signals of the etching products is supplemented, the average value is calculated, and m is a natural number greater than 1.

[0041] Optionally, the number of spectral signals of the etching products is supplemented and an average value is obtained, including:

[0042] The number of spectrum signals of the etching product is padded with data 0 so that the number of spectrum signals of the etching product for which an average value is calculated is m, and then the average value is calculated.

[0043] Optionally, the number of spectral signals of the etching products is supplemented and an average value is obtained, including:

[0044] The number of spectral signals of the etching product is supplemented by the spectral signal of the current etching product, so that the number of spectral signals of the etching product for which the average value is calculated is m, and then the average value is calculated.

[0045] In a second aspect, the present invention provides a deep silicon etching control system for executing the deep silicon etching control method described in any technical solution of the first aspect, comprising an acquisition unit, a mean calculation unit, a eigenvalue calculation unit, a normalized deviation calculation unit and a parameter adjustment unit, wherein the acquisition unit is used to acquire the spectral signal of the deposition product and the spectral signal of the etching product in one cycle of the Bosch process at a preset sampling frequency; the mean calculation unit is used to divide the spectral signals of a plurality of continuous deposition products into a group, and the spectral signals of two adjacent groups of deposition products share the spectral signals of a plurality of deposition products, and calculate the mean of the spectral signals of a group of deposition products to obtain the mean of the spectral signals of the plurality of deposition products, and the mean of the spectral signals of the plurality of deposition products constitutes the spectral signal of the deposition product. The method comprises the following steps: dividing the spectrum signals of a plurality of continuous etching products into a group, and the spectrum signals of two adjacent groups of etching products share the spectrum signals of a plurality of etching products, calculating the average of the spectrum signals of a group of etching products to obtain the average of the spectrum signals of a plurality of etching products, and the average of the spectrum signals of a plurality of etching products constitutes an etching product spectrum line; the characteristic value calculation unit is used to calculate the ratio according to the deposition product spectrum line and the etching product spectrum line to obtain the characteristic value; the normalized deviation calculation unit is used to calculate the normalized deviation of the characteristic value relative to the target ratio; the parameter adjustment unit is used to compare the normalized deviation with the preset parameter, and adjust the control parameter of the next cycle of the Bosch process according to the comparison result.

[0046] The beneficial effects of the deep silicon etching control system are: achieving more precise etching and reducing the disadvantages caused by fixed passivation and etching switching time, chemical gas flow rate and bias power.

[0047] In a third aspect, the present invention provides an etching device, comprising the deep silicon etching control system described in the second aspect.

[0048] The etching equipment has the beneficial effects of achieving more precise etching and reducing the disadvantages caused by fixed passivation and etching switching time, chemical gas flow rate and bias power. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 Flowchart of a deep silicon etching control method in some embodiments of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein should be the common meanings understood by people with ordinary skills in the field to which the present invention belongs. The words "including" and similar words used in this article mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects.

[0051] The Bosch process mainly includes a DEP stage, an ET1 stage, and an ET2 stage. The DEP stage is a deposition stage, the ET1 stage is a first etching stage, and the ET2 stage is a second etching stage. One DEP stage, one ET1 stage, and one ET2 stage constitute one cycle of the Bosch process.

[0052] The purpose of the DEP stage is to form a passivation layer on the sidewalls of the trench or hole to prevent erosion of the sidewalls during the subsequent etching process. The process is as follows: using octafluorocyclobutane (C4F8) as the main gas, it decomposes and deposits on the sidewalls of the trench or hole under the action of plasma to form a passivation layer (CF x ), such as CF2. Carbon difluoride can effectively prevent the sidewalls of the trench or hole from being further etched during the subsequent etching process, thereby ensuring the verticality and structural stability of the trench or hole.

[0053] The process of the ET1 stage is: using sulfur hexafluoride (SF6) as the etching gas, the main function of which is to bombard the bottom passivation layer of the groove or hole and remove the residual CF x , and perform local anisotropic etching, that is, perform local etching with strong directionality on the bottom of the trench or hole, and generate SiF.

[0054] The ET2 stage is the main etching stage, and the process is as follows: sulfur hexafluoride is used as the etching gas to isotropically etch the exposed silicon material and generate SiF. However, since the sidewalls of the grooves or holes are protected by a passivation layer, the uprightness and structural stability are maintained.

[0055] In view of the problems existing in the prior art, an embodiment of the present invention provides a deep silicon etching control method applied to the Bosch process. Figure 1 , the deep silicon etching control method comprises the following steps:

[0056] S1: Acquire the spectrum signals of the deposition product and the etching product in one cycle of the Bosch process at a preset sampling frequency.

[0057] In step S1, the deposition product includes but is not limited to CF2, and the etching product includes but is not limited to SiF. The spectral signal of the deposition product and the spectral signal of the etching product are obtained by a spectrometer. The resolution of the spectrometer used should be less than or equal to 0.1 nm, and the acquisition frequency is 100 Hz. The spectrometer can be specifically an atomic emission spectrometer (Optical Emission Spectrometer, OES). The spectral signal of the etching product obtained by the spectrometer can reflect the etching rate trend of the Bosch process in the ET1 stage and the ET2 stage. The spectral signal of the deposition product obtained by the spectrometer can reflect the generation efficiency of the passivation layer of the Bosch process in the DEP stage.

[0058] The spectrometer uses the argon (Ar) signal spectrum as a signal-to-noise ratio reference. In plasma equipment, Ar is typically used as the plasma ignition gas, and the gas flow rate used during ignition typically does not change with the etching step. Therefore, the Ar signal spectrum during ignition is used as a reference baseline for the signal-to-noise ratio to verify the stability of the reaction chamber.

[0059] In step S1, the preset sampling frequency is determined according to the time of one cycle of the Bosch process, and the preset sampling frequency is 15ms / time to 25ms / time, the time of one cycle of the Bosch process is 2s to 5s, and the sampling frequency is proportional to the time of one cycle of the Bosch process. Of course, a fixed preset sampling frequency of 20ms / time can also be determined according to the time of one cycle of the Bosch process. The spectrometer with a resolution less than or equal to 0.1nm and an acquisition frequency of 100Hz can capture more details during the changes in the plasma and achieve subtle monitoring of the etching reaction process, but at the same time it will generate a large amount of data, which requires extremely high storage and computing resources and is not conducive to long-term process monitoring. At the same time, due to the stability of the plasma and the reaction process, the preset sampling frequency of 20ms / time can ensure sensitive dynamic monitoring of the etching process without generating a large amount of data dynamic processing pressure.

[0060] Taking the preset sampling frequency of 20 ms / time as an example, when the time of one cycle of the Bosch process is 2 s, the spectrometer samples a total of 100 spectral signals of etching products and 100 spectral signals of deposition products.

[0061] Taking the preset sampling frequency of 20 ms / time as an example, when the time of one cycle of the Bosch process is 5 s, a total of 250 spectral signals of etching products and 250 spectral signals of deposition products are sampled by the spectrometer.

[0062] S2: Grouping the spectral signals of a plurality of continuous deposition products into a group, wherein the spectral signals of two adjacent groups of deposition products share the spectral signals of the plurality of deposition products, averaging the spectral signals of the one group of deposition products to obtain the spectral signal average of the plurality of deposition products, wherein the spectral signal averages of the plurality of deposition products constitute a deposition product spectral line; Grouping the spectral signals of a plurality of continuous etching products into a group, wherein the spectral signals of two adjacent groups of etching products share the spectral signals of the plurality of etching products, averaging the spectral signals of the one group of etching products to obtain the spectral signal average of the plurality of etching products, wherein the spectral signal averages of the plurality of etching products constitute an etching product spectral line. The spectral line can be formed by fitting the plurality of spectral signals.

[0063] In some embodiments, the spectral signals of several continuous deposition products are divided into a group, and the spectral signals of two adjacent groups of deposition products share the spectral signals of several deposition products, and the spectral signals of one group of deposition products are averaged to obtain the average of the spectral signals of several deposition products, including: according to the acquisition order of the spectral signals of the several deposition products, the spectral signals of several continuous deposition products are divided into a group, and the spectral signals of two adjacent groups of deposition products share the spectral signals of several deposition products; the spectral signal of the current deposition product and the spectral signals of the m deposition products sorted behind are taken as a group to obtain the average of the spectral signals of several deposition products, where m is a natural number greater than 1.

[0064] In some embodiments, the spectral signals of several consecutive etching products are divided into a group, and the spectral signals of two adjacent groups of etching products share the spectral signals of several etching products. The spectral signals of one group of etching products are averaged to obtain the average of the spectral signals of several etching products, including: according to the acquisition order of the spectral signals of several etching products, the spectral signals of several consecutive etching products are divided into a group, and the spectral signals of two adjacent groups of etching products share the spectral signals of several etching products; the spectral signal of the current etching product and the spectral signals of the m etching products sorted behind are taken as a group to obtain the average of the spectral signals of several etching products.

[0065] In the present application, the spectral signals of several continuous deposition products are divided into one group, and the spectral signals of two adjacent groups of deposition products are shared by the spectral signals of several deposition products. This can greatly reduce the impact of individual erroneous spectral signals on the results of averaging the spectral signals of one group of deposition products.

[0066] Furthermore, by sharing several spectral signals of the deposition products between two adjacent groups of spectral signals, the number of spectral signal means of the generated deposition products can be close to or the same as the number of spectral signals of the deposition products. After removing the influence of individual erroneous spectral signals, a spectral line, i.e., a deposition product spectral line, can be formed, so that the obtained deposition product spectral line is closer to the real deposition product spectral line, and the detection result is more accurate.

[0067] In the present application, the spectral signals of several continuous etching products are divided into one group, and the spectral signals of two adjacent groups of etching products are shared by the spectral signals of several etching products. This can greatly reduce the impact of individual erroneous spectral signals on the result of averaging the spectral signals of one group of etching products.

[0068] Furthermore, by sharing several spectral signals of the etching products between two adjacent groups of spectral signals of the etching products, the number of average values ​​of the spectral signals of the generated etching products can be close to or the same as the number of spectral signals of the etching products. After removing the influence of individual erroneous spectral signals, a spectral line, i.e., an etching product spectral line, can be formed, so that the obtained etching product spectral line is closer to the real etching product spectral line, making the detection result more accurate.

[0069] In step S2, according to the order in which the spectral signals of the etching products are acquired, the spectral signals of the etching products are sequentially used as the spectral signals of the current etching products, and according to the order in which the spectral signals of the deposition products are acquired, the spectral signals of the deposition products are sequentially used as the spectral signals of the current deposition products.

[0070] Preferably, m is 4. In the Bosch process, the minimum process time of a single stage is 500ms. If it is lower than 500ms, gas switching or etching process instability may occur. Therefore, selecting 5 data points in 100ms can ensure monitoring sensitivity without generating excessive data processing.

[0071] In some embodiments, in step S2, after the spectral signal of the last deposition product participates in the averaging for the first time, the step of averaging the spectral signal of the deposition product ends; after the spectral signal of the last etching product participates in the averaging for the first time, the step of averaging the spectral signal of the etching product ends.

[0072] The spectral signal of the deposition product is a CF2 spectral signal, and the spectral signal of the etching product is a SiF spectral signal. The CF2 spectral signal and the SiF spectral signal have strong anti-interference properties and can ensure the accuracy of the data.

[0073] For example, the SiF spectrum signals include a1, a2, a3, a4, a5, a6, a7, a8, a9, and a10, where a1 is the first SiF spectrum signal and a10 is the last SiF spectrum signal. For example, if m is 4, a1, a2, a3, a4, and a5 are grouped together, with an average value of A1; a2, a3, a4, a5, and a6 are grouped together, with an average value of A2; a3, a4, a5, a6, and a7 are grouped together, with an average value of A3; a4, a5, a6, a7, and a8 are grouped together, with an average value of A4; a5, a6, a7, a8, and a9 are grouped together, with an average value of A5; and a6, a7, a8, a9, and a10 are grouped together, with an average value of A6.

[0074] For example, the CF2 spectrum signals include b1, b2, b3, b4, b5, b6, b7, b8, b9, and b10, where b1 is the first CF2 spectrum signal and b10 is the last CF2 spectrum signal. Taking m as 4 as an example, b1, b2, b3, b4, and b5 are grouped together, with an average value of B1; b2, b3, b4, b5, and b6 are grouped together, with an average value of B2; b3, b4, b5, b6, and b7 are grouped together, with an average value of B3; b4, b5, b6, b7, and b8 are grouped together, with an average value of B4; b5, b6, b7, b8, and b9 are grouped together, with an average value of b5; and b6, b7, b8, b9, and b10 are grouped together, with an average value of B6.

[0075] In some embodiments, in step S2, if the number of spectral signals of the deposition product ranked later is less than m, the missing number is averaged; and if the number of spectral signals of the etching product ranked later is less than m, the missing number is averaged. This can reduce the amount of data for calculation.

[0076] For example, the plurality of SiF spectrum signals include a1, a2, a3, a4, a5, a6, a7, a8, a9, and a10, where a1 is the first SiF spectrum signal and a10 is the last SiF spectrum signal. Taking m as 4 as an example, a1, a2, a3, a4, and a5 are a group, and the average value is A1; a2, a3, a4, a5, and a6 are a group, and the average value is A2; a3, a4, a5, a6, and a7 are a group, and the average value is A3; a4, a5, a6, a7, and a8 are a group, and the average value is A4; a5, a6, a7, a8, and a9 are a group, and the average value is A5; a6, a7, a8, a9, and a10 are a group, and the average value is A6; a7, a8, a9, and a10 are a group, and the average value is A7; a8, a9, and a10 are a group, and the average value is A8; a9 and a10 are a group, and the average value is A9.

[0077] For example, the plurality of CF2 spectrum signals include b1, b2, b3, b4, b5, b6, b7, b8, b9, and b10, where b1 is the first CF2 spectrum signal and b10 is the last CF2 spectrum signal. Taking m as 4 as an example, b1, b2, b3, b4, and b5 are a group, and the average value is B1; b2, b3, b4, b5, and b6 are a group, and the average value is B2; b3, b4, b5, b6, and b7 are a group, and the average value is B3; b4, b5, b6, b7, and b8 are a group, and the average value is B4; b5, b6, b7, b8, and b9 are a group, and the average value is b5; b6, b7, b8, b9, and b10 are a group, and the average value is B6; b7, b8, b9, and b10 are a group, and the average value is B7; b8, b9, and b10 are a group, and the average value is B8; b9 and b10 are a group, and the average value is B9.

[0078] In some embodiments, in step S2, if the number of spectral signals of the etched products ranked later is less than m, the number of spectral signals of the etched products is supplemented and an average value is calculated; if the number of spectral signals of the deposition products ranked later is less than m, the number of spectral signals of the deposition products is supplemented and an average value is calculated. This ensures that the number of spectral signals of the etched products obtained is equal to the number of spectral signals of the etched products, and that the number of average values ​​of the spectral signals of the deposition products obtained is equal to the number of spectral signals of the deposition products obtained, thereby improving the accuracy of subsequent adjustment of control parameters for the next cycle of the Bosch process and avoiding inaccuracies caused by missing data.

[0079] In some embodiments, the number of spectral signals of the etching products is supplemented and then the average value is obtained, including: supplementing the number of spectral signals of the etching products with data 0 so that the number of spectral signals of the etching products for which the average value is obtained is m, and then the average value is obtained; the number of spectral signals of the deposition products is supplemented and then the average value is obtained, including: supplementing the number of spectral signals of the deposition products with data 0 so that the number of spectral signals of the deposition products for which the average value is obtained is m, and then the average value is obtained.

[0080] For example, the plurality of SiF spectrum signals include a1, a2, a3, a4, a5, a6, a7, a8, a9, and a10, where a1 is the first SiF spectrum signal and a10 is the last SiF spectrum signal. Taking m as 4 as an example, a1, a2, a3, a4, and a5 are a group, and the average value is A1; a2, a3, a4, a5, and a6 are a group, and the average value is A2; a3, a4, a5, a6, and a7 are a group, and the average value is A3; a4, a5, a6, a7, and a8 are a group, and the average value is A4; a5, a6, a7, a8, and a9 are a group, and the average value is A5; a6, a7, a8, a9, and a10 are a group, and the average value is A6; a7, a8, a9, a10, and 0 are a group, and the average value is A7; a8, a9, a10, 0, 0 are a group, and the average value is A8; a9, a10, 0, 0, 0 are a group, and the average value is A9; a10, 0, 0, 0, 0 are a group, and the average value is A10.

[0081] For example, the plurality of CF2 spectrum signals include b1, b2, b3, b4, b5, b6, b7, b8, b9, and b10, where b1 is the first CF2 spectrum signal and b10 is the last CF2 spectrum signal. Taking m as 4 as an example, b1, b2, b3, b4, and b5 are a group, and the average value is B1; b2, b3, b4, b5, and b6 are a group, and the average value is B2; b3, b4, b5, b6, and b7 are a group, and the average value is B3; b4, b5, b6, b7, and b8 are a group, and the average value is B4; b5, b6, b7, b8, and b9 are a group, and the average value is b5; b6, b7, b8, b9, and b10 are a group, and the average value is B6; b7, b8, b9, b10, and 0 are a group, and the average value is B7; b8, b9, b10, 0, 0 are a group, and the average value is B8; b9, b10, 0, 0, 0 are a group, and the average value is B9; b10, 0, 0, 0, 0 are a group, and the average value is B10.

[0082] In some embodiments, averaging the number of spectral signals of the etching product after supplementing the number of spectral signals includes: supplementing the number of spectral signals of the etching product with the spectral signal of the current etching product, so that the number of spectral signals of the etching product for which the average value is calculated is m, and then averaging the number; and averaging the number of spectral signals of the deposition product after supplementing the number of spectral signals includes: supplementing the number of spectral signals of the deposition product with the spectral signal of the current deposition product, so that the number of spectral signals of the deposition product for which the average value is calculated is m, and then averaging the number. Supplementing the number of spectral signals of the etching product with the spectral signal of the current etching product, and supplementing the number of spectral signals of the deposition product with the spectral signal of the current deposition product, makes the obtained average value more consistent with the true average value at the current time point, reducing the error introduced by additional data.

[0083] For example, the SiF spectrum signals include a1, a2, a3, a4, a5, a6, a7, a8, a9, and a10, where a1 is the first SiF spectrum signal and a10 is the last SiF spectrum signal. Taking m as 4 as an example, a1, a2, a3, a4, and a5 are grouped together, and the average value is A1; a2, a3, a4, a5, and a6 are grouped together, and the average value is A2; a3, a4, a5, a6, and a7 are grouped together, and the average value is A3; a4, a5, a6, a7, and a8 are grouped together, and the average value is A4; a5, a6, a7, a8, and a9 are grouped together, and the average value is A5; a 6, a7, a8, a9, and a10 are a group, and the average value is A6; a7, a8, a9, a10, and a7 are a group, and the average value is A7; a8, a9, a10, a8, and a8 are a group, and the average value is A8; a9, a10, a9, a9, and a9 are a group, and the average value is A9; a10, a10, a10, a10, and a10 are a group, and the average value is A10.

[0084] For example, the CF2 spectrum signals include b1, b2, b3, b4, b5, b6, b7, b8, b9, and b10, where b1 is the first CF2 spectrum signal and b10 is the last CF2 spectrum signal. Taking m as 4 as an example, b1, b2, b3, b4, and b5 are a group, and the average value is B1; b2, b3, b4, b5, and b6 are a group, and the average value is B2; b3, b4, b5, b6, and b7 are a group, and the average value is B3; b4, b5, b6, b7, and b8 are a group, and the average value is B4; b5, b6, b7, b8, and b9 are a group, and the average value is b5; b 6, b7, b8, b9, and b10 are a group, and the average value is B6; b7, b8, b9, b10, and b7 are a group, and the average value is B7; b8, b9, b10, b8, and b8 are a group, and the average value is B8; b9, b10, b9, b9, and b9 are a group, and the average value is B9; b10, b10, b10, b10, and b10 are a group, and the average value is B10.

[0085] S3: calculating a ratio based on the deposition product spectrum line and the etching product spectrum line to obtain a characteristic value.

[0086] Step S3 can be expressed as the eigenvalue calculation formula: R = E 440 (SiF*) / E 321 (CF*), where R represents the eigenvalue, E 440 (SiF*) represents the SiF spectral line, E 321 (CF*) represents the CF2 spectral line. The ratio of the SiF spectral line to the CF2 spectral line can be used to determine dynamic changes during the etching process. The ratio of the SiF spectral line to the CF2 spectral line, i.e., the characteristic value, can be any one of the intensity ratio, wavelength ratio, and slope ratio between the SiF spectral line and the CF2 spectral line. When the spectral line has multiple slopes, the corresponding slopes of the spectral lines are compared over time.

[0087] S4: Calculate the normalized deviation of the characteristic value relative to the target ratio.

[0088] In step S4, the normalized deviation of the characteristic value relative to the target ratio is obtained, including: presetting the target ratio, obtaining the difference between the characteristic value and the target ratio, and then obtaining the ratio between the difference and the target ratio to obtain the normalized deviation of the characteristic value relative to the target ratio. It can be expressed as a normalized deviation calculation formula: D = (RR*) / R*, where D represents the normalized deviation and R* represents the target ratio. Due to the different etching rates in deep silicon etching, process conditions and requirements of different aspect ratios, the spectral signals obtained by spectrometers under different process conditions will also be very different, so the target ratio is not specifically limited here.

[0089] In some embodiments, the target ratio is determined by preliminary experiments, including:

[0090] According to the specific process requirements of the sample to be etched, the control parameters of the reaction chamber are adjusted, and the etching rate within one cycle of the Bosch process is determined by scanning the slices with a scanning electron microscope (SEM) or a transmission electron microscope (TEM).

[0091] Obtaining the spectral signals of the etching products and the deposition products in the Bosch process through a spectrometer, and calculating the ratio of the spectral signals of the etching products to the spectral signals of the deposition products to obtain a characteristic value;

[0092] A corresponding relationship between the etching rate and the characteristic value is established, and then a target ratio is determined.

[0093] S5: Comparing the normalized deviation with preset parameters, and adjusting the control parameters of the next cycle of the Bosch process based on the comparison results. Multiple normalized deviations correspond to different time points, and the control parameters of the next cycle of the Bosch process at different time points are adjusted based on the comparison results to achieve dynamic adjustment.

[0094] When there are multiple slopes of the spectral line, the corresponding slopes between the spectral lines are compared according to time to obtain several eigenvalues. At this time, the several eigenvalues ​​are compared with the preset parameters. If more than 50% of the eigenvalues ​​are greater than the preset parameters, it is judged that the eigenvalue is greater than the preset parameters. If more than 50% of the eigenvalues ​​are less than the preset parameters, it is judged that the eigenvalue is less than the preset parameters.

[0095] In step S5, the preset parameters include a first preset sub-parameter and a second preset sub-parameter, the first preset sub-parameter is greater than the second preset sub-parameter, and the control parameters include the time of the DEP stage, the time of the ET1 stage, the time of the ET2 stage, the chemical gas flow rate of the DEP stage, the chemical gas flow rate of the ET stage, and the bias power. The bias power controls the energy and direction of the plasma bombardment. An increase in the bias power means that the energy of the ion bombardment increases and the anisotropy increases. A decrease in the bias power means that the energy of the plasma bombardment decreases and the isotropy increases.

[0096] When the normalized deviation is greater than the first preset parameter, increasing the time of the DEP stage, decreasing the time of the ET1 stage, decreasing the time of the ET2 stage, increasing the chemical gas flow rate of the DEP stage, decreasing the chemical gas flow rate of the ET stage, and decreasing the bias power;

[0097] When the normalized deviation is less than the second preset parameter, reducing the time of the DEP stage, increasing the time of the ET1 stage, increasing the time of the ET2 stage, reducing the chemical gas flow rate of the DEP stage, increasing the chemical gas flow rate of the ET stage, and increasing the bias power;

[0098] When the normalized deviation is less than or equal to the first preset parameter and greater than or equal to the second preset parameter, the time of the DEP stage is increased, the time of the ET1 stage is reduced, the time of the ET2 stage is reduced, the chemical gas flow rate of the DEP stage is increased, the chemical gas flow rate of the ET stage is reduced, and the bias power is maintained.

[0099] In step S5, when the normalized deviation is greater than the first preset parameter, the increased time of the DEP stage, the decreased time of the ET1 stage, the decreased time of the ET2 stage, the increased chemical gas flow rate of the DEP stage, the decreased chemical gas flow rate of the ET stage, and the decreased bias power are respectively proportional to the difference between the normalized deviation and the first preset parameter;

[0100] When the normalized deviation is less than the second preset parameter, the reduced time of the DEP stage, the increased time of the ET1 stage, the increased time of the ET2 stage, the reduced chemical gas flow rate of the DEP stage, the increased chemical gas flow rate of the ET stage, and the increased bias power are respectively proportional to the difference between the normalized deviation and the second preset parameter;

[0101] When the normalized deviation is less than or equal to the first preset parameter and greater than or equal to the second preset parameter, the time increased in the DEP stage, the time reduced in the ET1 stage, the time reduced in the ET2 stage, the chemical gas flow increased in the DEP stage, and the chemical gas flow reduced in the ET stage are respectively inversely proportional to the difference between the normalized deviation and the first preset parameter.

[0102] Optionally, the first preset parameter is 0.075 to 0.125, and the second preset parameter is -0.075 to -0.125, which can ensure that the control parameters of the next cycle of the Bosch process are adjusted accurately with a small error. If the first preset parameter and the second preset parameter exceed this range, the control parameters of the next cycle of the Bosch process will be adjusted inaccurately and the error will become larger.

[0103] In some specific embodiments, the first preset parameter is 0.1, and the second preset parameter is -0.1.

[0104] When the normalized deviation is greater than 0.1 and less than or equal to 0.2, the time of the DEP stage is increased by 5% to 10%, the time of the ET1 stage is reduced by 2.5% to 5%, the time of the ET2 stage is reduced by 2.5% to 5%, the chemical gas flow rate of the DEP stage is increased by 50 to 200 sccm, the chemical gas flow rate of the ET stage is reduced by 100 to 300 sccm, and the bias power is reduced by 0 to 10 W.

[0105] When the normalized deviation is greater than 0.1 and less than or equal to 0.2, the time of the DEP stage is increased by 5% to 10%, the time of the ET1 stage is reduced by 2.5% to 5%, the time of the ET2 stage is reduced by 2.5% to 5%, the chemical gas flow rate of the DEP stage is increased by 50 to 200 sccm, the chemical gas flow rate of the ET stage is reduced by 100 to 300 sccm, and the bias power is reduced by 0 to 10 W.

[0106] When the normalized deviation is less than -0.1 and greater than or equal to -0.2, the time of the DEP stage is reduced by 5% to 10%, the time of the ET1 stage is increased by 2.5% to 5%, the time of the ET2 stage is increased by 2.5% to 5%, the chemical gas flow rate of the DEP stage is reduced by 50 to 150 sccm, the chemical gas flow rate of the ET stage is increased by 10 to 200 sccm, and the bias power is increased by 0 to 15 W.

[0107] When the normalized deviation is less than or equal to 0.1 and greater than or equal to -0.1, the time of the DEP stage is increased by 0% to 5%, the time of the ET1 stage is reduced by 0% to 2.5%, the time of the ET2 stage is reduced by 0% to 2.5%, the chemical gas flow rate of the DEP stage is increased by 0% to 50 sccm, the chemical gas flow rate of the ET stage is reduced by 0% to 100 sccm, and the bias power is maintained.

[0108] An embodiment of the deep silicon etching control system of the present invention is used to implement the method described in any embodiment of the above-mentioned deep silicon etching control method, including an acquisition unit, a mean calculation unit, a characteristic value calculation unit, a normalized deviation calculation unit and a parameter adjustment unit, wherein the acquisition unit is used to obtain the spectral signal of the deposition product and the spectral signal of the etching product in one cycle of the Bosch process at a preset sampling frequency; the mean calculation unit is used to divide the spectral signals of several consecutive deposition products into a group, and the spectral signals of two adjacent groups of deposition products share the spectral signals of several deposition products, and calculate the mean of the spectral signals of one group of deposition products to obtain the mean of the spectral signals of several deposition products, and the mean of the spectral signals of several deposition products constitutes the deposition product spectral line, and, dividing the spectrum signals of several continuous etching products into one group, and the spectrum signals of two adjacent groups of etching products share the spectrum signals of several etching products, averaging the spectrum signals of one group of etching products to obtain the average of the spectrum signals of several etching products, and the average of the spectrum signals of several etching products constitutes the etching product spectrum line; the characteristic value calculation unit is used to calculate the ratio according to the deposition product spectrum line and the etching product spectrum line to obtain the characteristic value; the normalized deviation calculation unit is used to calculate the normalized deviation of the characteristic value relative to the target ratio; the parameter adjustment unit is used to compare the normalized deviation with the preset parameters, and adjust the control parameters of the next cycle of the Bosch process according to the comparison result. The beneficial effects of this embodiment are equivalent to the beneficial effects of any embodiment of the deep silicon etching control method, and will not be repeated here.

[0109] An embodiment of the etching equipment of the present invention includes the deep silicon etching control system described in the above embodiment of the deep silicon etching control system. The beneficial effects of this embodiment are equivalent to those of any embodiment of the deep silicon etching control method, and are not repeated here.

[0110] While the embodiments of the present invention have been described in detail above, it will be apparent to those skilled in the art that various modifications and variations of these embodiments are possible. However, it should be understood that such modifications and variations are within the scope and spirit of the present invention as set forth in the claims. Furthermore, the invention described herein is susceptible to other embodiments and may be practiced or implemented in a variety of ways.

Claims

1. A deep silicon etching control method, characterized in that: Applicable to Bosch processes, including: Acquire the spectral signals of the deposition product and the etched product in one cycle of the Bosch process at a preset sampling frequency; Dividing the spectral signals of a plurality of continuous deposition products into a group, and the spectral signals of two adjacent groups of deposition products share the spectral signals of a plurality of deposition products, averaging the spectral signals of a group of deposition products to obtain the spectral signal average of the plurality of deposition products, wherein the spectral signal average of the plurality of deposition products constitutes a deposition product spectral line; Dividing the spectrum signals of a plurality of continuous etching products into a group, wherein the spectrum signals of two adjacent groups of etching products share the spectrum signals of a plurality of etching products, averaging the spectrum signals of a group of etching products to obtain the average value of the spectrum signals of the plurality of etching products, wherein the average value of the spectrum signals of the plurality of etching products constitutes an etching product spectrum line; Calculating a ratio based on the deposition product spectrum line and the etching product spectrum line to obtain a characteristic value; Obtaining a normalized deviation of the characteristic value relative to a target ratio; The normalized deviation is compared with the preset parameters, and the control parameters of the next cycle of the Bosch process are adjusted according to the comparison result.

2. The deep silicon etching control method according to claim 1, wherein: The deposition product includes CF2, and the etching product includes SiF.

3. The deep silicon etching control method according to claim 1 or 2, characterized in that: Also includes: The preset sampling frequency is determined according to the time of one cycle of the Bosch process, the sampling frequency is 15ms / time to 25ms / time, the time of one cycle of the Bosch process is 2s to 5s, and the sampling frequency is proportional to the time of one cycle of the Bosch process.

4. The deep silicon etching control method according to claim 1 or 2, characterized in that: Obtaining a normalized deviation of the characteristic value relative to a target ratio includes: A target ratio is preset, a difference between the characteristic value and the target ratio is calculated, and then a ratio between the difference and the target ratio is calculated to obtain a normalized deviation of the characteristic value relative to the target ratio.

5. The deep silicon etching control method according to claim 1 or 2, characterized in that: The preset parameters include a first preset sub-parameter and a second preset sub-parameter, the first preset sub-parameter is greater than the second preset sub-parameter, and the control parameters include the time of the DEP stage, the time of the ET1 stage, the time of the ET2 stage, the chemical gas flow rate of the DEP stage, the chemical gas flow rate of the ET stage, and the bias power; When the normalized deviation is greater than the first preset parameter, increasing the time of the DEP stage, decreasing the time of the ET1 stage, decreasing the time of the ET2 stage, increasing the chemical gas flow rate of the DEP stage, decreasing the chemical gas flow rate of the ET stage, and decreasing the bias power; When the normalized deviation is less than the second preset parameter, reducing the time of the DEP stage, increasing the time of the ET1 stage, increasing the time of the ET2 stage, reducing the chemical gas flow rate of the DEP stage, increasing the chemical gas flow rate of the ET stage, and increasing the bias power; When the normalized deviation is less than or equal to the first preset parameter and greater than or equal to the second preset parameter, the time of the DEP stage is increased, the time of the ET1 stage is reduced, the time of the ET2 stage is reduced, the chemical gas flow rate of the DEP stage is increased, the chemical gas flow rate of the ET stage is reduced, and the bias power is maintained.

6. The deep silicon etching control method according to claim 5, characterized in that: When the normalized deviation is greater than the first preset parameter, the increased time of the DEP stage, the reduced time of the ET1 stage, the reduced time of the ET2 stage, the increased chemical gas flow rate of the DEP stage, the reduced chemical gas flow rate of the ET stage, and the reduced bias power are respectively proportional to the difference between the normalized deviation and the first preset parameter; When the normalized deviation is less than the second preset parameter, the reduced time of the DEP stage, the increased time of the ET1 stage, the increased time of the ET2 stage, the reduced chemical gas flow rate of the DEP stage, the increased chemical gas flow rate of the ET stage, and the increased bias power are respectively proportional to the difference between the normalized deviation and the second preset value; When the normalized deviation is less than or equal to the first preset parameter and greater than or equal to the second preset parameter, the time increased in the DEP stage, the time reduced in the ET1 stage, the time reduced in the ET2 stage, the chemical gas flow increased in the DEP stage, and the chemical gas flow reduced in the ET stage are respectively inversely proportional to the difference between the normalized deviation and the first preset parameter.

7. The deep silicon etching control method according to claim 6, characterized in that: The first preset parameter is 0.075 to 0.125, and the second preset parameter is -0.075 to -0.

125.

8. The deep silicon etching control method according to claim 1 or 2, characterized in that: The spectral signals of a plurality of continuous deposition products are divided into a group, and the spectral signals of two adjacent groups of deposition products share the spectral signals of a plurality of deposition products, and the spectral signals of a group of deposition products are averaged to obtain the average of the spectral signals of the plurality of deposition products, including: According to the acquisition order of the spectral signals of the plurality of deposition products, the spectral signals of the plurality of deposition products are divided into one group, and the spectral signals of the deposition products of two adjacent groups share the spectral signals of the plurality of deposition products; Taking the spectral signal of the current deposition product and the spectral signals of the m deposition products sorted after it as a group, an average is calculated to obtain the mean value of the spectral signals of the plurality of deposition products; Among them, after the spectral signal of the last deposition product participates in the averaging for the first time, the step of averaging the spectral signals of the deposition products is completed, or, if the number of spectral signals of the deposition products sorted later is less than m, the average value of the missing number is calculated, or, after the number of spectral signals of the deposition products is supplemented, the average value is calculated, and m is a natural number greater than 1.

9. The deep silicon etching control method according to claim 8, characterized in that: The number of spectral signals of the deposition products is supplemented and the average value is calculated, including: The number of spectral signals of the deposition product is padded with data 0 so that the number of spectral signals of the deposition product for which an average value is calculated is m, and then the average value is calculated.

10. The deep silicon etching control method according to claim 8, wherein: The number of spectral signals of the deposition products is supplemented and the average value is calculated, including: The number of spectral signals of the deposition product is supplemented by the spectral signal of the current deposition product, so that the number of spectral signals of the deposition product for which the average value is calculated is m, and then the average value is calculated.

11. The deep silicon etching control method according to claim 1 or 2, characterized in that: The spectral signals of a plurality of continuous etching products are divided into a group, and the spectral signals of two adjacent groups of etching products share the spectral signals of a plurality of etching products, and the spectral signals of a group of etching products are averaged to obtain the average value of the spectral signals of the plurality of etching products, including: According to the acquisition order of the spectral signals of the etching products, the spectral signals of the etching products are divided into a group, and the spectral signals of the etching products of two adjacent groups share the spectral signals of the etching products; Taking the spectrum signal of the current etching product and the spectrum signals of the m etching products sorted later as a group, an average value is calculated to obtain the average value of the spectrum signals of the plurality of etching products; Among them, after the spectral signal of the last etching product participates in the averaging for the first time, the step of averaging the spectral signals of the etching products is completed, or, if the number of spectral signals of the etching products sorted later is less than m, the average value of the missing number is calculated, or, after the number of spectral signals of the etching products is supplemented, the average value is calculated, and m is a natural number greater than 1.

12. The deep silicon etching control method according to claim 11, characterized in that: The number of spectral signals of the etching products is supplemented and the average value is calculated, including: The number of spectrum signals of the etching product is padded with data 0 so that the number of spectrum signals of the etching product for which an average value is calculated is m, and then the average value is calculated.

13. The deep silicon etching control method according to claim 11, characterized in that: The number of spectral signals of the etching products is supplemented and the average value is calculated, including: The number of spectral signals of the etching product is supplemented by the spectral signal of the current etching product, so that the number of spectral signals of the etching product for which the average value is calculated is m, and then the average value is calculated.

14. A deep silicon etching control system, characterized in that: The method is used to execute the deep silicon etching control method according to any one of claims 1 to 13, comprising an acquisition unit, a mean value calculation unit, a characteristic value calculation unit, a normalized deviation calculation unit, and a parameter adjustment unit, wherein the acquisition unit is used to acquire the spectral signal of the deposition product and the spectral signal of the etching product in one cycle of the Bosch process at a preset sampling frequency; the mean value calculation unit is used to divide the spectral signals of a plurality of continuous deposition products into a group, and the spectral signals of the deposition products of two adjacent groups share the spectral signals of a plurality of deposition products, and calculate the mean value of the spectral signal of a group of deposition products to obtain to the average of the spectral signals of the plurality of deposition products, the average of the spectral signals of the plurality of deposition products constituting a deposition product spectral line, and dividing the spectral signals of the plurality of continuous etching products into one group, and the spectral signals of two adjacent groups of etching products sharing the spectral signals of the plurality of etching products, calculating the average of the spectral signals of the one group of etching products to obtain the average of the spectral signals of the plurality of etching products, the average of the spectral signals of the plurality of etching products constituting an etching product spectral line; the eigenvalue calculation unit is used to calculate a ratio according to the deposition product spectral line and the etching product spectral line to obtain a eigenvalue; The normalized deviation calculation unit is used to obtain the normalized deviation of the characteristic value relative to the target ratio; the parameter adjustment unit is used to compare the normalized deviation with the preset parameters and adjust the control parameters of the next cycle of the Bosch process according to the comparison result.

15. An etching device, characterized in that: It includes the deep silicon etching control system as claimed in claim 14.

Images