Method of cleaning a semiconductor chamber and semiconductor apparatus

Abstract

The application discloses a cleaning method of a semiconductor chamber and a semiconductor device, comprising: cleaning the semiconductor chamber by using a first cleaning process, performing first filtering processing on first tail gas detection data collected during execution of the first cleaning process to obtain first filtered data, switching to a second cleaning process to clean the semiconductor chamber if the first filtered data meets a preset process switching condition, performing second filtering processing on second tail gas detection data collected during execution of the second cleaning process within a preset time period after switching to the second cleaning process to obtain second filtered data, and performing third filtering processing on the second tail gas detection data after the preset time period to obtain third filtered data. The second filtering processing removes burrs to a smaller extent than the first filtering processing and the third filtering processing. The cleaning condition of the semiconductor chamber is determined according to the first filtered data, the second filtered data and the third filtered data.

Description

Technical Field

[0001] This application relates to the field of semiconductor manufacturing technology, and in particular to a method for cleaning a semiconductor chamber and a semiconductor device. Background Technology

[0002] After a certain period of operation, semiconductor deposition equipment will accumulate deposits on the inner walls of its chambers. Regular cleaning, or cleaning under specified conditions, can extend the chamber's lifespan. The cleaning duration is typically determined by an endpoint detection device installed on the chamber's roughing line. This endpoint detection device measures the concentration of cleaning products generated during chemical cleaning in the exhaust gas from the chamber. When there is a large amount of film to be cleaned in the chamber, the concentration of cleaning products generated during chemical cleaning in the exhaust gas is high. As cleaning progresses, the concentration of cleaning products generated during chemical cleaning in the exhaust gas gradually decreases. Near the end of cleaning, the concentration of cleaning products generated during chemical cleaning in the exhaust gas becomes so low that it is difficult for the endpoint detection device to detect.

[0003] Because the chamber is continuously filled with cleaning reaction gas and gas in a plasma state, and because the chamber is in dynamic equilibrium, the exhaust velocity is not completely constant. Furthermore, the electromagnetic induction phenomenon generated by the nearby power source affects the signal; therefore, the curve output by the endpoint detection device fluctuates.

[0004] The current approach uses a sliding window for filtering, which introduces latency and results in a large computational load due to the large number of values ​​involved. Furthermore, traditional filtering methods indiscriminately smooth the curves, classifying any spikes as noise fluctuations and thus eliminating them. However, when the cleaning process changes, meaningful waveform fluctuations will inevitably occur due to the process change. Indiscriminate curve smoothing makes it impossible to observe the changes in cleaning product concentration caused by the process change, making it impossible to monitor the switching fluctuations. In particular, when the slopes of the curves before and after the cleaning process change are similar, if the switching fluctuations cannot be monitored, the timing of the cleaning process change will be difficult to detect, making it difficult to determine the completion status of the cleaning process from the filtered data. Summary of the Invention

[0005] The purpose of this application is to provide a method and apparatus for cleaning a semiconductor chamber, in order to solve the problem of over-filtering in the current cleaning process of semiconductor chambers.

[0006] To solve the above-mentioned technical problems, the embodiments of this application are implemented as follows: On one hand, embodiments of this application provide a method for cleaning a semiconductor chamber, including: The semiconductor chamber is cleaned using the first cleaning process; The first exhaust gas detection data collected during the first cleaning process is subjected to a first filtering process to obtain the first filtered data; If the first filtered data meets the preset process switching conditions, the process is switched to the second cleaning process to clean the semiconductor chamber. Within a preset time period after switching to the second cleaning process, the second exhaust gas detection data collected during the execution of the second cleaning process is subjected to a second filtering process to obtain the second filtered data. After the preset time period, the second exhaust gas detection data is subjected to a third filtering process to obtain third filtered data; wherein, the degree to which the second filtering process removes burrs is less than the degree to which the first filtering process removes burrs, and less than the degree to which the third filtering process removes burrs. The cleaning status of the semiconductor chamber is determined based on the first filtered data, the second filtered data, and the third filtered data.

[0007] On the other hand, embodiments of this application provide a semiconductor device, including a semiconductor chamber, an endpoint detection device, and a control unit; wherein, The semiconductor chamber is used to perform the first cleaning process and the second cleaning process; The endpoint detection device is used to detect the exhaust gas in the semiconductor chamber during the execution of the first cleaning process to obtain first exhaust gas detection data; and to detect the exhaust gas in the semiconductor chamber during the execution of the second cleaning process to obtain second exhaust gas detection data. The control unit includes a memory and a processor. The memory stores a computer program, and the processor is used to call and execute the computer program from the memory to implement the above-described semiconductor chamber cleaning method.

[0008] In another aspect, embodiments of this application provide a semiconductor chamber cleaning apparatus, including a processor; and a memory arranged to store computer-executable instructions configured to be executed by the processor to implement the above-described semiconductor chamber cleaning method.

[0009] In another aspect, embodiments of this application provide a storage medium for storing computer-executable instructions, which, when executed by a processor, implement the aforementioned semiconductor chamber cleaning method.

[0010] The technical solution of this application embodiment involves cleaning the semiconductor chamber using a first cleaning process. First exhaust gas detection data collected during the first cleaning process undergoes a first filtering process to obtain first filtered data. If the first filtered data meets preset process switching conditions, the process switches to a second cleaning process to clean the semiconductor chamber. During a preset time period after switching to the second cleaning process, second exhaust gas detection data collected during the second cleaning process undergoes a second filtering process to obtain second filtered data. After the preset time period, the second exhaust gas detection data undergoes a third filtering process to obtain third filtered data. The degree to which the second filtering process removes burrs is less than that of the first filtering process and less than that of the third filtering process. Therefore, the cleaning status of the semiconductor chamber is determined based on the first filtered data, the second filtered data, and the third filtered data. It is evident that, on the one hand, this technical solution employs different filtering processes before and after switching to the next cleaning process, with the post-switching filtering removing fewer burrs than the pre-switching filtering. This preserves the burr details in the exhaust gas detection data during the cleaning process switch, resulting in filtered data that more closely reflects the actual situation during the process switch. This facilitates monitoring the cleaning process switch based on the filtered data. On the other hand, this technical solution can apply different filtering processes to exhaust gas detection data collected at two different time points within a cleaning process. Since the variation range of exhaust gas detection data collected immediately upon entering the cleaning process differs from that collected after a certain duration of the cleaning process, different filtering processes can more effectively filter the exhaust gas detection data. This results in filtered data that more closely reflects the actual process conditions, avoids over-filtering, and is beneficial for determining the completion status of the semiconductor chamber cleaning based on the filtered data. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a schematic flowchart of a method for cleaning a semiconductor chamber according to an embodiment of this application; Figure 2 This is a graph of a filtering result according to an embodiment of this application; Figure 3 This is a graph showing another filtering result according to an embodiment of this application; Figure 4 This is a schematic flowchart of a method for cleaning a semiconductor chamber according to another embodiment of this application; Figure 5 This is a schematic block diagram of a semiconductor device according to an embodiment of this application; Figure 6 This is a schematic block diagram of a semiconductor device according to another embodiment of this application; Figure 7 This is a schematic block diagram of a semiconductor device according to another embodiment of this application; Figure 8 This is a schematic block diagram of a semiconductor device according to another embodiment of this application; Figure 9 This is a schematic diagram of a semiconductor chamber cleaning device according to an embodiment of this application. Detailed Implementation

[0013] This application provides a method and apparatus for cleaning a semiconductor chamber, which addresses the problem of over-filtering in the current semiconductor chamber cleaning process.

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

[0015] Figure 1 This is a schematic flowchart illustrating a method for cleaning a semiconductor chamber according to an embodiment of this application, as shown below. Figure 1 As shown, the method includes: S102, the semiconductor chamber is cleaned using the first cleaning process.

[0016] S104, the first exhaust gas detection data collected during the first cleaning process is subjected to a first filtering process to obtain the first filtered data.

[0017] S106, if the first filtered data meets the preset process switching conditions, switch to the second cleaning process to clean the semiconductor chamber.

[0018] Optionally, the preset process switching condition may be that the first filtered data is less than a first preset threshold; or, the first filtered data is greater than a second preset threshold; or, the fluctuation range of the first filtered data is within a preset data fluctuation range.

[0019] S108, within a preset time period after switching to the second cleaning process, the second exhaust gas detection data collected during the execution of the second cleaning process is subjected to a second filtering process to obtain the second filtered data.

[0020] The preset time period can be the homogenization period for the cleaning reaction gas. The homogenization period can be determined based on factors such as the size of the semiconductor chamber, the introduction rate of the cleaning reaction gas, and historical exhaust gas detection data of the cleaning process. During the homogenization period, the cleaning reaction gas is gradually introduced into the semiconductor chamber and evenly dispersed to various locations within the semiconductor chamber. Simultaneously, the introduced cleaning reaction gas reacts with the deposits within the semiconductor chamber to generate gaseous products.

[0021] In practical applications, the gas equalization process of the first cleaning process in the semiconductor chamber can be avoided by controlling the opening timing of the valve between the endpoint detection device and the vacuum pipeline of the semiconductor chamber.

[0022] S110, after a preset time period, performs a third filtering process on the second exhaust gas detection data to obtain the third filtered data.

[0023] The second filtering process removes burrs to a lesser extent than the first filtering process, and also less than the third filtering process. In other words, the second filtered data retains more burr details than the first filtered data, and more than the third filtered data. This facilitates the accurate monitoring of fluctuations during cleaning process switching.

[0024] The time for performing the third filtering process can include the period from after a preset time period until the end of the second cleaning process.

[0025] S112, determine the cleaning status of the semiconductor chamber based on the first filtered data, the second filtered data, and the third filtered data.

[0026] In practical applications, the cleaning process for a semiconductor chamber can include multiple steps. This technical solution uses a scenario where two cleaning processes are employed to clean the semiconductor chamber as an example to illustrate the semiconductor chamber cleaning method provided in this application embodiment. The first cleaning process is the first cleaning process, and the second cleaning process is the second cleaning process. Different cleaning processes use different cleaning reaction gases, resulting in different gaseous products. The first and second tail gas detection data can be output voltages generated by an endpoint detection device that match the concentration of the gaseous products.

[0027] It is understandable that when multiple cleaning processes are used to clean the semiconductor chamber, the filtering processes in the second to the last cleaning processes can refer to the filtering steps in the second cleaning process described above. For example, in a scenario where three cleaning processes are used to clean the semiconductor chamber, the filtering processes in the first and second cleaning processes can refer to steps S102-S110 described above. The filtering process in the third cleaning process can refer to steps S106-S110 described above, specifically including: switching to the third cleaning process to clean the semiconductor chamber when the third filtered data meets the preset process switching conditions; performing a fourth filtering process on the third exhaust gas detection data collected during the execution of the third cleaning process within a preset time period after switching to the third cleaning process to obtain the fourth filtered data; and performing a fifth filtering process on the third exhaust gas detection data after the preset time period to obtain the fifth filtered data. Optionally, the preset process switching conditions can be that the third filtered data is less than a third preset threshold; or, the third filtered data is greater than a fourth preset threshold; or, the fluctuation range of the third filtered data is within a preset data fluctuation range.

[0028] The fourth and second filtering processes can be the same or different, and the fifth and third filtering processes can be the same or different. The fourth filtering process removes glitch to a lesser extent than the third filtering process, and also less than the fifth filtering process.

[0029] The technical solution of this application embodiment involves cleaning the semiconductor chamber using a first cleaning process. First exhaust gas detection data collected during the first cleaning process undergoes a first filtering process to obtain first filtered data. If the first filtered data meets preset process switching conditions, the process switches to a second cleaning process to clean the semiconductor chamber. During a preset time period after switching to the second cleaning process, second exhaust gas detection data collected during the second cleaning process undergoes a second filtering process to obtain second filtered data. After the preset time period, the second exhaust gas detection data undergoes a third filtering process to obtain third filtered data. The degree to which the second filtering process removes burrs is less than that of the first filtering process and less than that of the third filtering process. Therefore, the cleaning status of the semiconductor chamber is determined based on the first filtered data, the second filtered data, and the third filtered data. It is evident that, on the one hand, this technical solution employs different filtering processes before and after switching to the next cleaning process, with the post-switching filtering removing fewer burrs than the pre-switching filtering. This preserves the burr details in the exhaust gas detection data during the cleaning process switch, resulting in filtered data that more closely reflects the actual situation during the process switch. This facilitates monitoring the cleaning process switch based on the filtered data. On the other hand, this technical solution can apply different filtering processes to exhaust gas detection data collected at two different time points within a cleaning process. Since the variation range of exhaust gas detection data collected immediately upon entering the cleaning process differs from that collected after a certain duration of the cleaning process, different filtering processes can more effectively filter the exhaust gas detection data. This results in filtered data that more closely reflects the actual process conditions, avoids over-filtering, and is beneficial for determining the completion status of the semiconductor chamber cleaning based on the filtered data.

[0030] In one embodiment, the filtering algorithms for the first, second, and third filtering processes are the same. The first and second filtering processes use different filtering parameters, and the second and third filtering processes use different filtering parameters.

[0031] The filtering parameters used in the first and third filtering processes can be the same or different. Optionally, the filtering algorithm can be a Kalman filter. Because the Kalman filter algorithm has the characteristics of low computational load and low computational latency, it can achieve real-time filtering of exhaust gas detection data with less computing power, thus improving the processing speed of exhaust gas detection data. A sufficiently good effect can be obtained using a first-order Kalman filter algorithm. Furthermore, the adjustable parameters Q and R in the Kalman filter algorithm correspond to physical quantities, namely the prediction noise covariance matrix and the observation noise covariance matrix, which can handle different semiconductor chambers or different types and principles of endpoint detection devices.

[0032] In this embodiment, using the same filtering algorithm to implement three filtering processes simplifies the control logic of the filtering process and improves its accuracy and efficiency. Furthermore, for two adjacent filtering processes, filtering algorithms with different filtering parameters can be used to filter the collected exhaust gas detection data, achieving different filtering effects. This not only makes the obtained filtered data more closely match the actual process conditions, avoiding over-filtering, but also facilitates understanding the cleaning status of the semiconductor chamber based on the filtered data (such as cleaning process switching status, cleaning process completion status, etc.).

[0033] Furthermore, the filtering algorithms for the first, second, and third filtering processes can be different. Specifically, the first and second filtering processes may use different filtering algorithms, the second and third filtering processes may use different filtering algorithms, and the first and third filtering processes may use the same or different filtering algorithms. The selection of filtering algorithms will not be elaborated upon in this embodiment, as long as the desired filtering outcome is achieved. Filtering outcomes may include: obtaining the desired frequency, removing unwanted frequencies, allowing observation of cleaning process switching, and facilitating the determination of the process endpoint, etc.

[0034] In one embodiment, the first filtering process and the third filtering process use the same filtering parameters.

[0035] In this embodiment, the filtering process is completed by using a filtering algorithm with the same filtering parameters as the first cleaning process after a preset time period of the second cleaning process. This reduces the number of parameters that need to be controlled, which helps to simplify the control logic of the filtering process and improve the accuracy and efficiency of the filtering process.

[0036] In one embodiment, when the filtering algorithm is a Kalman filter algorithm, the first exhaust gas detection data collected during the first cleaning process is subjected to a first filtering process to obtain the first filtered data (i.e., S104), which can be executed as follows: Steps A1-A2: Step A1: Update the filtering parameters of the Kalman filter algorithm using the first parameter value to obtain the first Kalman filter algorithm.

[0037] The filtering parameters of the Kalman filter algorithm can include the state transition matrix, the observation matrix, and the prediction noise covariance matrix. The first parameter value is the value of the above three filtering parameters.

[0038] This step provides an algorithmic basis for filtering the first exhaust gas detection data collected during the first cleaning process, enabling the filtering process to be executed accurately, thereby achieving accurate filtering of the first exhaust gas detection data and making the filtering results more consistent with the actual process conditions.

[0039] Step A2: For the kth first exhaust gas detection data collected during the execution of the first cleaning process, the first Kalman filter algorithm is used, and the first filter data and first filter error corresponding to the (k-1)th first exhaust gas detection data are used to determine the first filter data and first filter error corresponding to the kth first exhaust gas detection data.

[0040] Where k is an integer greater than or equal to 2. The first filtering error is determined based on the first exhaust gas detection data and its corresponding first filtered data. The first filtering error can be the difference between the first exhaust gas detection data and the first filtered data corresponding to the first exhaust gas detection data.

[0041] Optionally, the first filtered data can be a first posterior estimate, and the first filtered error can be a first posterior error covariance. The first parameter value may include a first state transition matrix, a first observation matrix, and a first prediction noise covariance matrix. Step A2 can be performed as follows: Steps A21-A22: Step A21: For the kth first exhaust gas detection data collected during the execution of the first cleaning process, determine the first prior estimate and the first prior error covariance of the k-1th first exhaust gas detection data based on the first posterior estimate and the first posterior error covariance corresponding to the (k-1)th first exhaust gas detection data, as well as the first state transition matrix and the first prediction noise covariance matrix.

[0042] Specifically, the first prior estimate of the k-th first exhaust gas detection data can be determined based on the first posterior estimate and the first state transition matrix corresponding to the (k-1)-th first exhaust gas detection data. Furthermore, the first prior error covariance of the k-th first exhaust gas detection data can be determined based on the first posterior error covariance, the first state transition matrix, and the first prediction noise covariance matrix corresponding to the (k-1)-th first exhaust gas detection data.

[0043] The first prior estimate of the kth first exhaust gas detection data and the first prior error covariance The calculation formulas are shown in formulas (1) and (2): (1) in, This is the first state transition matrix. This is the first posterior estimate corresponding to the (k-1)th first exhaust gas detection data. Optionally, the first state transition matrix can be a constant.

[0044] (2) in, Let the first posterior error covariance be the first exhaust gas detection data corresponding to the (k-1)th data point. Let be the first prediction noise covariance matrix.

[0045] Step A22: Based on the first prior estimate and first prior error covariance, the first observation matrix, the first observation noise covariance matrix, and the kth first exhaust gas detection data, determine the first posterior estimate and first posterior error covariance corresponding to the kth first exhaust gas detection data.

[0046] Specifically, the Kalman gain corresponding to the k-th first exhaust gas detection data can be determined based on the first prior error covariance, the first observation matrix, and the first observation noise covariance matrix of the k-th first exhaust gas detection data. Therefore, the first posterior estimate corresponding to the k-th first exhaust gas detection data can be determined based on the Kalman gain, the first prior estimate, the first observation matrix, and the k-th first exhaust gas detection data. Furthermore, the first posterior error covariance corresponding to the k-th first exhaust gas detection data can be determined based on the Kalman gain, the first observation matrix, and the first prior error covariance of the k-th first exhaust gas detection data.

[0047] Kalman gain corresponding to the kth first exhaust gas detection data The first posterior estimate corresponding to the kth first exhaust gas detection data and the first posterior error covariance The calculation formulas are shown in formulas (3) to (5): (3) in, Let be the first observation matrix. Optionally, the first observation matrix can be a constant. The first observation noise covariance matrix is... The first influencing factor is the covariance. Covariance represents the volatility, and therefore, the Kalman gain represents the compensation for this volatility, resulting in a smooth curve. The strength of the filter, i.e., the smoothness of the curve, is greatly influenced by the term... Make a decision. yes Multiply It represents Item and The proportional relationship between the items. In this embodiment, -1 < <1 and -1< <1, as can be seen from this formula, the degree of filtering K It is not subject to and The influence of absolute values, but generally conforms to K = / ( + ), that is K =1 / (1+ However, due to covariance... P The numerical ratio itself is very small, too large or Will cover P Due to the influence of [the factor], it is not recommended to choose excessively large values.

[0048] (4) in, This is the kth first exhaust gas detection data.

[0049] (5) In this embodiment, by clearly defining the filtering process for the second first exhaust gas detection data to the last first exhaust gas detection data collected during the first cleaning process, it is beneficial to ensure the accurate execution of the filtering process.

[0050] For example, if the first posterior estimate corresponding to the (k-1)th first exhaust gas detection data is... 3. First posterior error covariance The first state transition matrix is ​​0.0000001. 1. First observation matrix 1. First prediction noise covariance matrix The first observation noise covariance matrix is ​​0.01. If the value is 0.1 and the kth first exhaust gas detection data is 4, then, according to the above formulas (1) to (5), the first prior estimate of the kth first exhaust gas detection data can be calculated. =1*3=3, First prior error covariance =1 2* 0.0000001 + 0.01 = 0.0100001, the Kalman gain corresponding to the kth first exhaust gas detection data. =0.0100001*1 (0.0100001*1 2 +0.1) -1 =0.090909, the first posterior estimate corresponding to the kth first exhaust gas detection data. =3 + 0.090909 (4 - 1 * 3) = 3.090909, where the first posterior error covariance corresponds to the kth first exhaust gas detection data. = (1 - 0.090909 * 1) * 0.0100001 = 0.0090910009091.

[0051] It is understood that the first cleaning process is the same as the first cleaning process. In this embodiment, the first exhaust gas detection data collected during the execution of the first cleaning process is not filtered. The first exhaust gas detection data is only used for the filtering process formula (1) of the second exhaust gas detection data. Assignment. The first posterior error covariance of the first exhaust gas detection data. The preset values ​​are then used to accurately filter the second set of exhaust gas detection data.

[0052] In one embodiment, during a preset time period after switching to the second cleaning process, the second exhaust gas detection data collected during the execution of the second cleaning process is subjected to a second filtering process to obtain second filtered data (i.e., S108), which can be executed as follows: Steps B1-B3: Step B1: Within a preset time period after switching to the second cleaning process, update the filtering parameters of the Kalman filter algorithm using the second parameter value to obtain the second Kalman filter algorithm.

[0053] The second parameter value is the value of the three filtering parameters (state transition matrix, observation matrix, and prediction noise covariance matrix) of the Kalman filter algorithm mentioned above.

[0054] This step provides an algorithmic basis for filtering the second exhaust gas detection data collected within the preset time period of the second cleaning process, enabling the filtering process to be executed accurately, thereby achieving accurate filtering of the second exhaust gas detection data and making the filtering results more consistent with the actual process conditions.

[0055] Step B2: For the first second exhaust gas detection data collected within the preset time period, the second Kalman filter algorithm is used, and based on the first filter data and the first filter error corresponding to the last first exhaust gas detection data collected during the first cleaning process, the second filter data and the second filter error corresponding to the first second exhaust gas detection data collected within the preset time period are determined.

[0056] The process of determining the first filter data and the first filter error corresponding to the last first exhaust gas detection data collected during the first cleaning process can be found in steps A21-A22, and will not be repeated here.

[0057] The second filtering error is determined based on the second exhaust gas detection data and its corresponding second filtered data. The second filtering error can be the difference between the second exhaust gas detection data and the corresponding second filtered data.

[0058] Optionally, the second filtered data can be a second posterior estimate, and the second filtered error can be the second posterior error covariance. The second parameter values ​​may include the second state transition matrix, the second observation matrix, and the second prediction noise covariance matrix. Step B2 can be performed as follows: Steps B21-B22: Step B21: Based on the first posterior estimate and the first posterior error covariance corresponding to the last first exhaust gas detection data, as well as the second state transition matrix and the second prediction noise covariance matrix, determine the second prior estimate and the second prior error covariance of the first second exhaust gas detection data.

[0059] Specifically, the second prior estimate of the first second exhaust gas detection data can be determined based on the first posterior estimate and the second state transition matrix corresponding to the last first exhaust gas detection data. Furthermore, the second prior error covariance of the first second exhaust gas detection data can be determined based on the first posterior error covariance, the second state transition matrix, and the second prediction noise covariance matrix corresponding to the last first exhaust gas detection data.

[0060] Among them, the second prior estimate of the first second exhaust gas detection data Second prior error covariance The calculation formula can be found in formulas (1) and (2) above, and will not be repeated here. In the formula, This is the second state transition matrix. This is the first posterior estimate corresponding to the last first exhaust gas detection data. Let be the first posterior error covariance corresponding to the last first exhaust gas detection data. This is the second prediction noise covariance matrix.

[0061] Step B22: Based on the second prior estimate and second prior error covariance of the first second exhaust gas detection data, the second observation matrix, the second observation noise covariance matrix, and the first second exhaust gas detection data, determine the second posterior estimate and second posterior error covariance corresponding to the first second exhaust gas detection data.

[0062] Specifically, the Kalman gain corresponding to the first second exhaust gas detection data can be determined based on the second prior error covariance, the second observation matrix, and the second observation noise covariance matrix of the first second exhaust gas detection data. Therefore, the second posterior estimate corresponding to the first second exhaust gas detection data can be determined based on the Kalman gain, the second prior estimate, the second observation matrix, and the first second exhaust gas detection data. Furthermore, the second posterior error covariance corresponding to the first second exhaust gas detection data can be determined based on the Kalman gain, the second observation matrix, and the second prior error covariance of the first second exhaust gas detection data.

[0063] Among them, the Kalman gain corresponding to the first second exhaust gas detection data The second posterior estimate corresponding to the first second exhaust gas detection data Second posterior error covariance The calculation formula can be found in formulas (3) to (5) above, and will not be repeated here. In the formula, This is the second observation matrix. The second observation noise covariance matrix is... It is the second most influential factor. This is the first second exhaust gas detection data.

[0064] In this embodiment, by clearly defining the filtering process for the first second exhaust gas detection data collected within a preset time period, it is beneficial to ensure the accurate execution of the filtering process.

[0065] Step B3: For the kth second exhaust gas detection data collected within the preset time period, the second Kalman filter algorithm is used, and the second filter data and second filter error corresponding to the (k-1)th second exhaust gas detection data collected within the preset time period are used to determine the second filter data and second filter error corresponding to the kth second exhaust gas detection data collected within the preset time period.

[0066] Where k is an integer greater than or equal to 2. Step B3 can be executed as follows: Steps B31-B32: Step B31: For the kth second exhaust gas detection data collected within the preset time period, determine the second prior estimate and the second prior error covariance of the k-1th second exhaust gas detection data based on the second posterior estimate and the second posterior error covariance corresponding to the (k-1)th second exhaust gas detection data, as well as the second state transition matrix and the second prediction noise covariance matrix.

[0067] Specifically, the second prior estimate of the k-th second exhaust gas detection data can be determined based on the second posterior estimate and the second state transition matrix corresponding to the (k-1)-th second exhaust gas detection data. Furthermore, the second prior error covariance of the k-th second exhaust gas detection data can be determined based on the second posterior error covariance, the second state transition matrix, and the second prediction noise covariance matrix corresponding to the (k-1)-th second exhaust gas detection data.

[0068] Among them, the second prior estimate of the kth second exhaust gas detection data Second prior error covariance The calculation formula can be found in formulas (1) and (2) above, and will not be repeated here. In the formula, This is the second state transition matrix. It is the second posterior estimate corresponding to the (k-1)th second exhaust gas detection data. Let the second posterior error covariance be the covariance corresponding to the (k-1)th second exhaust gas detection data. This is the second prediction noise covariance matrix.

[0069] Step B32: Based on the second prior estimate and second prior error covariance of the kth second exhaust gas detection data, the second observation matrix, the second observation noise covariance matrix, and the kth second exhaust gas detection data, determine the second posterior estimate and second posterior error covariance corresponding to the kth second exhaust gas detection data.

[0070] Specifically, the Kalman gain corresponding to the k-th second exhaust gas detection data can be determined based on the second prior error covariance, the second observation matrix, and the second observation noise covariance matrix of the k-th second exhaust gas detection data. Therefore, the second posterior estimate corresponding to the k-th second exhaust gas detection data can be determined based on the Kalman gain, the second prior estimate, the second observation matrix, and the k-th second exhaust gas detection data. Furthermore, the second posterior error covariance corresponding to the k-th second exhaust gas detection data can be determined based on the Kalman gain, the second observation matrix, and the second prior error covariance of the k-th second exhaust gas detection data.

[0071] Wherein, the Kalman gain corresponding to the kth second exhaust gas detection data The second posterior estimate corresponding to the kth second exhaust gas detection data. Second posterior error covariance The calculation formula can be found in formulas (3) to (5) above, and will not be repeated here. In the formula, This is the second observation matrix. The second observation noise covariance matrix is... It is the second most influential factor. This is the kth second exhaust gas detection data.

[0072] In this embodiment, by clearly defining the filtering process for the second exhaust gas detection data collected within a preset time period of the second cleaning process, up to the last second exhaust gas detection data, it is beneficial to ensure the accurate execution of the filtering process.

[0073] In one embodiment, after a preset time period, the second exhaust gas detection data is subjected to a third filtering process to obtain the third filtered data (i.e., S110), which can be executed as follows: Steps C1-C3: Step C1: After a preset time period, update the filtering parameters of the Kalman filter algorithm using the first parameter value to obtain the first Kalman filter algorithm.

[0074] The first parameter value is the value of the three filtering parameters (state transition matrix, observation matrix, and prediction noise covariance matrix) of the Kalman filter algorithm mentioned above.

[0075] This step provides an algorithmic basis for filtering the second exhaust gas detection data collected after the preset time period of the second cleaning process, enabling the filtering process to be executed accurately, thereby achieving accurate filtering of the second exhaust gas detection data and making the filtering results more consistent with the actual process conditions.

[0076] Step C2: For the first second exhaust gas detection data collected after a preset time period, the first Kalman filter algorithm is used, and based on the second filter data and the second filter error corresponding to the last second exhaust gas detection data collected within the preset time period, the third filter data and the third filter error corresponding to the first second exhaust gas detection data collected after the preset time period are determined.

[0077] The determination process of the second filter data and the second filter error corresponding to the last second exhaust gas detection data collected within the preset time period of the second cleaning process can be found in steps B31-B32, and will not be repeated here.

[0078] The third filtering error is determined based on the second exhaust gas detection data and its corresponding third filtering data. The third filtering error can be the difference between the second exhaust gas detection data and the corresponding third filtering data.

[0079] Optionally, the third filtered data can be a third posterior estimate, and the third filtered error can be the third posterior error covariance. The first parameter value may include a first state transition matrix, a first observation matrix, and a first prediction noise covariance matrix. Step C2 can be performed as follows: Steps C21-C22: Step C21: Based on the second posterior estimate and second posterior error covariance corresponding to the last second exhaust gas detection data collected within the preset time period, as well as the first state transition matrix and the first prediction noise covariance matrix, determine the third prior estimate and third prior error covariance of the first second exhaust gas detection data collected after the preset time period.

[0080] Specifically, the third prior estimate of the first second exhaust gas detection data collected after the preset time period can be determined based on the second posterior estimate and the first state transition matrix corresponding to the last second exhaust gas detection data collected within the preset time period. Furthermore, the third prior error covariance of the first second exhaust gas detection data collected after the preset time period can be determined based on the second posterior error covariance, the first state transition matrix, and the first prediction noise covariance matrix corresponding to the last second exhaust gas detection data collected within the preset time period.

[0081] Among them, the third prior estimate of the first second exhaust gas detection data collected after a preset time period. and the third prior error covariance The calculation formula can be found in formulas (1) and (2) above, and will not be repeated here. In the formula, This is the first state transition matrix. This is the second posterior estimate corresponding to the last second exhaust gas detection data collected within the preset time period. This represents the second posterior error covariance corresponding to the last second exhaust gas detection data collected within a preset time period. Let be the first prediction noise covariance matrix.

[0082] Step C22: Based on the third prior estimate and third prior error covariance of the first second exhaust gas detection data collected after the preset time period, the first observation matrix, the first observation noise covariance matrix, and the first second exhaust gas detection data collected after the preset time period, determine the third posterior estimate and third posterior error covariance corresponding to the first second exhaust gas detection data collected after the preset time period.

[0083] Specifically, the Kalman gain corresponding to the first second exhaust gas detection data collected after a preset time period can be determined based on the third prior error covariance, the first observation matrix, and the first observation noise covariance matrix of the first second exhaust gas detection data collected after a preset time period. Therefore, based on the Kalman gain and the third prior estimate corresponding to the first second exhaust gas detection data collected after a preset time period, as well as the first observation matrix and the first second exhaust gas detection data collected after a preset time period, the third posterior estimate corresponding to the first second exhaust gas detection data collected after a preset time period can be determined. Furthermore, based on the Kalman gain, the first observation matrix, and the third prior error covariance corresponding to the first second exhaust gas detection data collected after a preset time period, the third posterior error covariance corresponding to the first second exhaust gas detection data collected after a preset time period can be determined.

[0084] Among them, the Kalman gain corresponding to the first second exhaust gas detection data collected after a preset time period. Third posterior estimate and third posterior error covariance The calculation formula can be found in formulas (3) to (5) above, and will not be repeated here. In the formula, This is the first observation matrix. The first observation noise covariance matrix is... As the number one impact factor, This refers to the first second exhaust gas detection data collected after a preset time period.

[0085] In this embodiment, by clearly defining the filtering process for the first second exhaust gas detection data collected after a preset time period, it is beneficial to ensure the accurate execution of the filtering process.

[0086] Step C3: For the kth second exhaust gas detection data collected after a preset time period, the first Kalman filter algorithm is used, and the third filter data and third filter error corresponding to the (k-1)th second exhaust gas detection data collected after the preset time period are used to determine the third filter data and third filter error corresponding to the kth second exhaust gas detection data collected after the preset time period.

[0087] Where k is an integer greater than or equal to 2. Step C3 can be executed as follows: Steps C31-C32: Step C31: For the kth second exhaust gas detection data collected after a preset time period, determine the third prior estimate and third prior error covariance of the kth second exhaust gas detection data based on the third posterior estimate and third posterior error covariance corresponding to the (k-1)th second exhaust gas detection data, as well as the first state transition matrix and the first prediction noise covariance matrix.

[0088] Specifically, the third prior estimate of the k-th second exhaust gas detection data can be determined based on the third posterior estimate corresponding to the (k-1)-th second exhaust gas detection data and the first state transition matrix. Furthermore, the third prior error covariance of the k-th second exhaust gas detection data can be determined based on the third posterior error covariance, the first state transition matrix, and the first prediction noise covariance matrix corresponding to the (k-1)-th second exhaust gas detection data.

[0089] Among them, the third prior estimate of the kth second exhaust gas detection data and the third prior error covariance The calculation formula can be found in formulas (1) and (2) above, and will not be repeated here. In the formula, This is the first state transition matrix. This is the third posterior estimate corresponding to the (k-1)th second exhaust gas detection data. Let the third posterior error covariance be the value corresponding to the (k-1)th second exhaust gas detection data. Let be the first prediction noise covariance matrix.

[0090] Step C32: Based on the third prior estimate and third prior error covariance of the kth second exhaust gas detection data, the first observation matrix, the first observation noise covariance matrix, and the kth second exhaust gas detection data, determine the third posterior estimate and third posterior error covariance corresponding to the kth second exhaust gas detection data.

[0091] Specifically, the Kalman gain corresponding to the k-th second exhaust gas detection data can be determined based on the third prior error covariance, the first observation matrix, and the first observation noise covariance matrix of the k-th second exhaust gas detection data. Therefore, the third posterior estimate corresponding to the k-th second exhaust gas detection data can be determined based on the Kalman gain, the third prior estimate, the first observation matrix, and the k-th second exhaust gas detection data. Furthermore, the third posterior error covariance corresponding to the k-th second exhaust gas detection data can be determined based on the Kalman gain, the first observation matrix, and the third prior error covariance of the k-th second exhaust gas detection data.

[0092] Wherein, the Kalman gain corresponding to the kth second exhaust gas detection data The third posterior estimate corresponding to the kth second exhaust gas detection data and third posterior error covariance The calculation formula can be found in formulas (3) to (5) above, and will not be repeated here. In the formula, This is the first observation matrix. The first observation noise covariance matrix is... As the number one impact factor, This is the kth second exhaust gas detection data.

[0093] In this embodiment, by clearly defining the filtering process for the second exhaust gas detection data collected after the preset time period of the second cleaning process up to the last second exhaust gas detection data, it is beneficial to the accurate execution of the filtering process.

[0094] Furthermore, if the filtering parameters used in the first filtering process and the third filtering process are different, then after a preset time period, the second exhaust gas detection data will undergo the third filtering process to obtain the third filtered data (i.e., S110), which can be executed as follows: Steps D1-D3: Step D1: After a preset time period, update the filtering parameters of the Kalman filter algorithm using the third parameter value to obtain the third Kalman filter algorithm.

[0095] The third parameter value is the value of the three filtering parameters (state transition matrix, observation matrix, and prediction noise covariance matrix) of the Kalman filter algorithm described above. Optionally, the third parameter value may include the third state transition matrix, the third observation matrix, and the third prediction noise covariance matrix.

[0096] This step provides an algorithmic basis for filtering the second exhaust gas detection data collected after the preset time period of the second cleaning process, enabling the filtering process to be executed accurately, thereby achieving accurate filtering of the second exhaust gas detection data and making the filtering results more consistent with the actual process conditions.

[0097] Step D2: For the first second exhaust gas detection data collected after the preset time period, the third Kalman filter algorithm is used, and based on the second filter data and the second filter error corresponding to the last second exhaust gas detection data collected within the preset time period, the third filter data and the third filter error corresponding to the first second exhaust gas detection data collected after the preset time period are determined.

[0098] The third filtering error is determined based on the second exhaust gas detection data and its corresponding third filtered data. The third filtering error can be the difference between the second exhaust gas detection data and the corresponding third filtered data. Optionally, the third filtered data can be a third posterior estimate, and the third filtering error can be the third posterior error covariance.

[0099] Step D3: For the kth second exhaust gas detection data collected after a preset time period, the third Kalman filter algorithm is used, and the third filter data and third filter error corresponding to the (k-1)th second exhaust gas detection data collected after the preset time period are used to determine the third filter data and third filter error corresponding to the kth second exhaust gas detection data collected after the preset time period.

[0100] It is understood that, compared with steps C1-C3 above, the only difference in this embodiment is the filtering parameters of the filtering algorithm. Therefore, apart from the use of the third parameter value for the filtering parameters, the other processing procedures can refer to steps C1-C3 above, and will not be repeated here.

[0101] It should be noted that before performing the cleaning process, the values ​​of the first and second parameters, or even more parameters (such as the third parameter), need to be determined. During the determination of these parameters, considering that filtering is essentially an estimation, and according to the theory of Fourier transform, a wave is a superposition of many waves of different frequencies. Therefore, the exhaust gas detection data obtained from the endpoint detection device is a sample of the superposition of many waves of different frequencies. The detection results from the endpoint detection device are similar, including both low-frequency and high-frequency signals. Therefore, it is necessary to statistically analyze the factors that may affect the detection results of the endpoint detection device. These influencing factors are generally sources of change during the cleaning process, such as gas, water, and electricity. By examining the data and pressure data of the MFC (Mass Flow Controller), it can be found that if the semiconductor chamber operates stably and the airflow changes slowly, a low-frequency signal is generated. The water flow is constant, so its influence is negligible. Electricity, however, will generate a high-frequency signal because it is 50Hz (Hertz).

[0102] Since the endpoint detection device measures the concentration of gaseous products, and gas conditions change slowly, so too does the concentration change slowly; therefore, the detected signal should be low-frequency. Based on this, a low-pass filter can be used to allow low-frequency signals to pass while filtering out high-frequency signals. The Kalman filter model is one type of low-pass filter, characterized by low computational complexity. By simplifying the formula according to the application scenario, fewer settings can be required. This means that the computational parts in the code only need to use constants for storage, eliminating the need for a stack and reducing design complexity. After determining the selected low-pass filter, the cleaning process can be executed multiple times, and the filtering results observed. If the filtering results meet the expected results, the values ​​of the filtering parameters are recorded to obtain the first parameter value, the second parameter value, etc.

[0103] The filtering expectations may include: obtaining the required frequency, removing the unwanted frequency, being able to observe the switching of the cleaning process, and making it easy to determine the process endpoint, etc.

[0104] In one embodiment, determining the cleaning status of the semiconductor chamber based on the first filtered data, the second filtered data, and the third filtered data (i.e., S112) can be performed as follows: if the first filtered data and the second filtered data meet preset conditions, determine that the cleaning process switching of the semiconductor chamber has been successful; or, determine the process endpoint of the second cleaning process based on the steepness of the third filtered data.

[0105] The preset conditions may include: the smoothness of the first filtered data and the second filtered data are different, and / or, the second filtered data has spikes.

[0106] In this embodiment, since the filtered data retains the burr details when the cleaning process is switched and is relatively close to the actual process situation, the switching status of the cleaning process and the cleaning completion status of the semiconductor chamber can be determined relatively accurately through the filtered data.

[0107] For example, Figure 2 The figure shows the curves obtained by filtering exhaust gas detection data using a filtering algorithm with only one filtering parameter. Figure 3 The curves obtained by filtering exhaust gas detection data using filtering algorithms with different filtering parameters are shown. (Comparison) Figure 2 and Figure 3 It can be seen that if only one filtering algorithm with one filtering parameter is used to filter the exhaust gas detection data, the exhaust gas detection data when the cleaning process is switched will be filtered out as noise. This makes it difficult to monitor the switching of the cleaning process, and thus it is difficult to determine the completion status of the cleaning process based on the exhaust gas detection data.

[0108] Figure 4 This is a schematic flowchart illustrating a semiconductor chamber cleaning method according to another embodiment of this application. In this embodiment, taking the cleaning of the semiconductor chamber using three cleaning processes as an example, the processing method for exhaust gas detection data is described in detail. The filtering algorithms used in each filtering process are the same. To simplify the control logic of the filtering processes, the filtering parameters used in the first and second filtering processes are different, the filtering parameters used in the second and third filtering processes are different, the filtering parameters used in the third and fourth filtering processes are different, the filtering parameters used in the fourth and fifth filtering processes are different, the filtering parameters used in the first, third, and fifth filtering processes are the same, and the filtering parameters used in the second and fourth filtering processes are the same. Figure 4 As shown, the method may include: S401 uses the first cleaning process to clean the semiconductor chamber.

[0109] S402, the first exhaust gas detection data collected during the first cleaning process is subjected to a first filtering process to obtain the first filtered data.

[0110] S403, if the first filtered data meets the preset process switching conditions, switch to the second cleaning process to clean the semiconductor chamber.

[0111] S404, within a preset time period after switching to the second cleaning process, the second exhaust gas detection data collected during the execution of the second cleaning process is subjected to a second filtering process to obtain the second filtered data.

[0112] S405, after a preset time period, performs a third filtering process on the second exhaust gas detection data to obtain the third filtered data.

[0113] The second filtering process removes burrs to a lesser extent than the first filtering process, and also to a lesser extent than the third filtering process.

[0114] S406, if the third filtered data meets the preset process switching conditions, switch to the third cleaning process to clean the semiconductor chamber.

[0115] S407, within a preset time period after switching to the third cleaning process, the third exhaust gas detection data collected during the execution of the third cleaning process is subjected to a fourth filtering process to obtain the fourth filtered data.

[0116] S408, after a preset time period, performs a fifth filtering process on the third exhaust gas detection data to obtain the fifth filtered data.

[0117] The fourth filtering process removes burrs to a lesser extent than the third filtering process, and also to a lesser extent than the fifth filtering process.

[0118] S409, Based on the first filtered data, the second filtered data, the third filtered data, the fourth filtered data, and the fifth filtered data, determine the cleaning status of the semiconductor chamber.

[0119] The specific implementation methods of S401-S409 have been described in detail in the above embodiments and will not be repeated here.

[0120] The technical solution adopted in this application has two advantages. First, when switching from one cleaning process to the next, different filtering processes are used before and after the switch. The filtering after the switch removes fewer burrs than before, preserving the burr details in the exhaust gas detection data during the cleaning process switch. This makes the obtained filtered data more closely reflect the actual situation during the process switch, facilitating the monitoring of the cleaning process switch based on the filtered data. Second, different filtering processes can be applied to the exhaust gas detection data collected in two time periods within a cleaning process. Since the variation range of the exhaust gas detection data collected immediately after entering the cleaning process differs from that collected after a certain duration of the cleaning process, different filtering processes can more effectively filter the exhaust gas detection data, making the obtained filtered data more closely reflect the actual process situation, avoiding over-filtering, and facilitating the determination of the semiconductor chamber cleaning completion status based on the filtered data.

[0121] In summary, specific embodiments of this subject matter have been described. Other embodiments are within the scope of the appended claims. In some cases, the actions recited in the claims can be performed in a different order and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing can be advantageous.

[0122] The above describes a method for cleaning a semiconductor chamber according to an embodiment of this application. Based on the same idea, this application also provides a semiconductor device.

[0123] Figure 5 This is a schematic block diagram of a semiconductor device according to an embodiment of this application, such as... Figure 5 As shown, the semiconductor device includes a semiconductor chamber 510, an endpoint detection device 520, and a control unit 530 connected in sequence.

[0124] Semiconductor chamber 510 is used to perform a first cleaning process and a second cleaning process.

[0125] The endpoint detection device 520 is used to detect the exhaust gas in the semiconductor chamber during the execution of the first cleaning process to obtain first exhaust gas detection data; and to detect the exhaust gas in the semiconductor chamber during the execution of the second cleaning process to obtain second exhaust gas detection data.

[0126] Optionally, the endpoint detection device can use methods such as infrared spectroscopy, connection to a residual gas analyzer, optical emission spectroscopy, or infrared spectroscopy to detect the concentration of gaseous products. The endpoint detection device can be connected in parallel with the vacuum line of the semiconductor chamber. When no cleaning process is performed, the valve between the endpoint detection device and the vacuum line is closed to extend the lifespan of the endpoint detection device by reducing usage time. The exhaust gas detection data output by the endpoint detection device can be a voltage signal.

[0127] The control unit 530 includes a memory and a processor. The memory stores a computer program, and the processor is used to call and execute the computer program from the memory to implement the above-described semiconductor chamber cleaning method.

[0128] The technical solution of this application embodiment involves cleaning the semiconductor chamber using a first cleaning process. First exhaust gas detection data collected during the first cleaning process undergoes a first filtering process to obtain first filtered data. If the first filtered data meets preset process switching conditions, the process switches to a second cleaning process to clean the semiconductor chamber. During a preset time period after switching to the second cleaning process, second exhaust gas detection data collected during the second cleaning process undergoes a second filtering process to obtain second filtered data. After the preset time period, the second exhaust gas detection data undergoes a third filtering process to obtain third filtered data. The degree to which the second filtering process removes burrs is less than that of the first filtering process and less than that of the third filtering process. Therefore, the cleaning status of the semiconductor chamber is determined based on the first filtered data, the second filtered data, and the third filtered data. It is evident that, on the one hand, this technical solution employs different filtering processes before and after switching to the next cleaning process, with the post-switching filtering removing fewer burrs than the pre-switching filtering. This preserves the burr details in the exhaust gas detection data during the cleaning process switch, resulting in filtered data that more closely reflects the actual situation during the process switch. This facilitates monitoring the cleaning process switch based on the filtered data. On the other hand, this technical solution can apply different filtering processes to exhaust gas detection data collected at two different time points within a cleaning process. Since the variation range of exhaust gas detection data collected immediately upon entering the cleaning process differs from that collected after a certain duration of the cleaning process, different filtering processes can more effectively filter the exhaust gas detection data. This results in filtered data that more closely reflects the actual process conditions, avoids over-filtering, and is beneficial for determining the completion status of the semiconductor chamber cleaning based on the filtered data.

[0129] In one embodiment, such as Figure 6 As shown, the control unit 530 includes a sampling subunit 610, a filtering subunit 620, and a computer storage subunit 630 connected in sequence. The sampling subunit 610 is connected to the endpoint detection device 520. It should be noted that... Figure 6 Is Figure 5 The schematic block diagram obtained after refining the control unit 530 based on this is therefore... Figure 6 Zhongyu Figure 5 Please refer to the connection relationship between components with the same number. Figure 5 This will not be elaborated upon here.

[0130] The sampling subunit 610 is used to sample and process the exhaust gas detection data during the execution of each cleaning process to obtain the first exhaust gas detection data and the second exhaust gas detection data.

[0131] When the exhaust gas detection data output by the endpoint detection device is a voltage signal, the sampling subunit can sample the voltage signal, convert the analog quantity into a digital quantity, obtain discrete values ​​after sampling, and use the discrete values ​​to input the filtering subunit for filtering processing.

[0132] The filtering subunit 620 is used to perform a first filtering process on the first exhaust gas detection data collected during the execution of the first cleaning process to obtain first filtered data; within a preset time period after switching to the second cleaning process, it performs a second filtering process on the second exhaust gas detection data collected during the execution of the second cleaning process to obtain second filtered data; after the preset time period, it performs a third filtering process on the second exhaust gas detection data to obtain third filtered data; wherein, the degree of burr removal by the second filtering process is less than the degree of burr removal by the first filtering process and less than the degree of burr removal by the third filtering process.

[0133] The computer storage subunit 630 is used to receive and store first filtered data, second filtered data and third filtered data; and to determine the cleaning status of the semiconductor chamber based on the first filtered data, second filtered data and third filtered data.

[0134] In this embodiment, by setting up a filtering subunit, the filtering operation is removed from the computer, thereby reducing the computer's computational burden.

[0135] In one embodiment, such as Figure 7 As shown, the control unit 530 includes a sampling subunit 710 and a computer system 720 connected in sequence. The sampling subunit 710 is connected to the endpoint detection device 520. The computer system 720 includes a filtering subsystem 7210 and a computer storage subsystem 7220 connected to each other. The filtering subsystem 7210 is connected to the sampling subunit 710. It should be noted that... Figure 7 Is Figure 5 The schematic block diagram obtained after refining the control unit 530 based on this is therefore... Figure 7 Zhongyu Figure 5 Please refer to the connection relationship between components with the same number. Figure 5 This will not be elaborated upon here.

[0136] The sampling subunit 710 is used to sample and process the exhaust gas detection data during the execution of each cleaning process to obtain the first exhaust gas detection data and the second exhaust gas detection data.

[0137] The filtering subsystem 7210 is used to perform a first filtering process on the first exhaust gas detection data collected during the execution of the first cleaning process to obtain first filtered data; within a preset time period after switching to the second cleaning process, it performs a second filtering process on the second exhaust gas detection data collected during the execution of the second cleaning process to obtain second filtered data; after the preset time period, it performs a third filtering process on the second exhaust gas detection data to obtain third filtered data; wherein, the degree of burr removal by the second filtering process is less than the degree of burr removal by the first filtering process and less than the degree of burr removal by the third filtering process.

[0138] The computer storage subsystem 7220 is used to receive and store first filtered data, second filtered data and third filtered data; and to determine the cleaning status of the semiconductor chamber based on the first filtered data, second filtered data and third filtered data.

[0139] In this embodiment, the filtering subsystem can be implemented in software via computer programming. This saves on the hardware cost of filtering.

[0140] In one embodiment, such as Figure 8 As shown, the semiconductor device may include a semiconductor chamber 810, an endpoint detection system 820, and a computer storage 830 connected in sequence. The endpoint detection system 820 includes an endpoint detection device 8210, a sampling device 8220, and a filtering device 8230 connected in sequence. The endpoint detection device 8210 is connected to the semiconductor chamber 810, and the filtering device 8230 is connected to the computer storage 830.

[0141] Semiconductor chamber 810 is used to perform a first cleaning process and a second cleaning process.

[0142] The endpoint detection device 8210 is used to detect the exhaust gas in the semiconductor chamber during the execution of the first cleaning process to obtain first exhaust gas detection data; and to detect the exhaust gas in the semiconductor chamber during the execution of the second cleaning process to obtain second exhaust gas detection data.

[0143] The sampling device 8220 samples and processes the exhaust gas detection data during the execution of each cleaning process to obtain the first exhaust gas detection data and the second exhaust gas detection data.

[0144] The filtering device 8230 is used to perform a first filtering process on the first exhaust gas detection data collected during the execution of the first cleaning process to obtain first filtered data; within a preset time period after switching to the second cleaning process, it performs a second filtering process on the second exhaust gas detection data collected during the execution of the second cleaning process to obtain second filtered data; after the preset time period, it performs a third filtering process on the second exhaust gas detection data to obtain third filtered data; wherein, the degree of burr removal by the second filtering process is less than the degree of burr removal by the first filtering process and less than the degree of burr removal by the third filtering process.

[0145] Computer storage 830 is used to receive and store first filtered data, second filtered data and third filtered data; and to determine the cleaning status of the semiconductor chamber based on the first filtered data, second filtered data and third filtered data.

[0146] In this embodiment, the filtering device can be implemented in hardware using an FPGA (Field Programmable Gate Array). The FPGA can be integrated with the endpoint detection device and the sampling device to obtain an optimized endpoint detection system. This endpoint detection system can directly output the processed digital signal to the computer, thereby removing the computation from the computer and reducing the computational burden on the computer.

[0147] Those skilled in the art will understand that Figures 5 to 8 The semiconductor equipment described above can be used to implement the cleaning method for the semiconductor chamber described above. The details of the method description should be similar to those in the previous section. To avoid being too complicated, they will not be repeated here.

[0148] Based on the same idea, this application also provides a cleaning device for semiconductor chambers, such as... Figure 9 As shown. The semiconductor chamber cleaning apparatus can vary considerably depending on its configuration or performance, and may include one or more processors 901 and a memory 902. The memory 902 may store one or more application programs or data. The memory 902 may be temporary or persistent storage. The application programs stored in the memory 902 may include one or more modules (not shown), each module may include a series of computer-executable instructions for the semiconductor chamber cleaning apparatus. Furthermore, the processor 901 may be configured to communicate with the memory 902 and execute the series of computer-executable instructions in the memory 902 on the semiconductor chamber cleaning apparatus. The semiconductor chamber cleaning apparatus may also include one or more power supplies 903, one or more wired or wireless network interfaces 904, one or more input / output interfaces 905, and one or more keyboards 906.

[0149] Specifically, in this embodiment, the semiconductor chamber cleaning apparatus includes a memory and one or more application programs, wherein one or more application programs are stored in the memory, and the one or more application programs may include one or more modules, and each module may include a series of computer-executable instructions for the semiconductor chamber cleaning apparatus, and is configured to be executed by one or more processors. The one or more application programs include computer-executable instructions for performing the following: The semiconductor chamber is cleaned using the first cleaning process; The first exhaust gas detection data collected during the first cleaning process is subjected to a first filtering process to obtain the first filtered data. If the first filtered data meets the preset process switching conditions, the process is switched to the second cleaning process to clean the semiconductor chamber. Within a preset time period after switching to the second cleaning process, the second exhaust gas detection data collected during the execution of the second cleaning process is subjected to a second filtering process to obtain the second filtered data. After a preset time period, the second exhaust gas detection data is subjected to a third filtering process to obtain the third filtered data; wherein, the degree of burr removal by the second filtering process is less than the degree of burr removal by the first filtering process, and less than the degree of burr removal by the third filtering process. The cleaning status of the semiconductor chamber is determined based on the first, second, and third filtered data.

[0150] The technical solution of this application embodiment involves cleaning the semiconductor chamber using a first cleaning process. First exhaust gas detection data collected during the first cleaning process undergoes a first filtering process to obtain first filtered data. If the first filtered data meets preset process switching conditions, the process switches to a second cleaning process to clean the semiconductor chamber. During a preset time period after switching to the second cleaning process, second exhaust gas detection data collected during the second cleaning process undergoes a second filtering process to obtain second filtered data. After the preset time period, the second exhaust gas detection data undergoes a third filtering process to obtain third filtered data. The degree to which the second filtering process removes burrs is less than that of the first filtering process and less than that of the third filtering process. Therefore, the cleaning status of the semiconductor chamber is determined based on the first filtered data, the second filtered data, and the third filtered data. It is evident that, on the one hand, this technical solution employs different filtering processes before and after switching to the next cleaning process, with the post-switching filtering removing fewer burrs than the pre-switching filtering. This preserves the burr details in the exhaust gas detection data during the cleaning process switch, resulting in filtered data that more closely reflects the actual situation during the process switch. This facilitates monitoring the cleaning process switch based on the filtered data. On the other hand, this technical solution can apply different filtering processes to exhaust gas detection data collected at two different time points within a cleaning process. Since the variation range of exhaust gas detection data collected immediately upon entering the cleaning process differs from that collected after a certain duration of the cleaning process, different filtering processes can more effectively filter the exhaust gas detection data. This results in filtered data that more closely reflects the actual process conditions, avoids over-filtering, and is beneficial for determining the completion status of the semiconductor chamber cleaning based on the filtered data.

[0151] This application also proposes a storage medium that stores one or more computer programs, which include computer-executable instructions. When executed by an electronic device including multiple applications, the computer-executable instructions enable the electronic device to perform the various processes of the above-described semiconductor chamber cleaning method embodiments and achieve the same technical effect. To avoid repetition, these will not be described again here.

[0152] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0153] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0154] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0155] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0156] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0157] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0158] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0159] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0160] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0161] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0162] This application can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0163] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0164] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for cleaning a semiconductor chamber, characterized in that, include: The semiconductor chamber is cleaned using the first cleaning process; The first exhaust gas detection data collected during the first cleaning process is subjected to a first filtering process to obtain the first filtered data; If the first filtered data meets the preset process switching conditions, the process is switched to the second cleaning process to clean the semiconductor chamber. Within a preset time period after switching to the second cleaning process, the second exhaust gas detection data collected during the execution of the second cleaning process is subjected to a second filtering process to obtain the second filtered data. After the preset time period, the second exhaust gas detection data is subjected to a third filtering process to obtain third filtered data; wherein, the degree to which the second filtering process removes burrs is less than the degree to which the first filtering process removes burrs, and less than the degree to which the third filtering process removes burrs. The cleaning status of the semiconductor chamber is determined based on the first filtered data, the second filtered data, and the third filtered data.

2. The method according to claim 1, characterized in that, The filtering algorithms for the first filtering process, the second filtering process, and the third filtering process are the same; The first filtering process and the second filtering process use different filtering parameters; The second filtering process uses different filtering parameters than the third filtering process.

3. The method according to claim 2, characterized in that, The first filtering process and the third filtering process use the same filtering parameters.

4. The method according to claim 3, characterized in that, When the filtering algorithm is the Kalman filter algorithm, The first filtering process is performed on the first exhaust gas detection data collected during the execution of the first cleaning process to obtain the first filtered data, including: The filtering parameters of the Kalman filter algorithm are updated using the first parameter value to obtain the first Kalman filter algorithm; For the kth first exhaust gas detection data collected during the execution of the first cleaning process, the first Kalman filter algorithm is used, and the first filter data and first filter error corresponding to the (k-1)th first exhaust gas detection data are used to determine the first filter data and first filter error corresponding to the kth first exhaust gas detection data. Where k is an integer greater than or equal to 2; the first filtering error is determined based on the first exhaust gas detection data and the corresponding first filtering data.

5. The method according to claim 4, characterized in that, Within a preset time period after switching to the second cleaning process, the second exhaust gas detection data collected during the execution of the second cleaning process is subjected to a second filtering process to obtain second filtered data, including: Within a preset time period after switching to the second cleaning process, the filtering parameters of the Kalman filter algorithm are updated using the second parameter value to obtain the second Kalman filter algorithm; For the first second exhaust gas detection data collected within the preset time period, the second Kalman filter algorithm is used, and based on the first filter data and the first filter error corresponding to the last first exhaust gas detection data collected during the execution of the first cleaning process, the second filter data and the second filter error corresponding to the first second exhaust gas detection data collected within the preset time period are determined. For the kth second exhaust gas detection data collected within the preset time period, the second Kalman filter algorithm is used, and the second filtered data and second filtered error corresponding to the (k-1)th second exhaust gas detection data collected within the preset time period are used to determine the second filtered data and second filtered error corresponding to the kth second exhaust gas detection data collected within the preset time period; the second filtered error is determined based on the second exhaust gas detection data and its corresponding second filtered data.

6. The method according to claim 5, characterized in that, After the preset time period, the second exhaust gas detection data undergoes a third filtering process to obtain third filtered data, including: After the preset time period, the filtering parameters of the Kalman filter algorithm are updated using the first parameter value to obtain the first Kalman filter algorithm; For the first second exhaust gas detection data collected after the preset time period, the first Kalman filter algorithm is used, and based on the second filter data and the second filter error corresponding to the last second exhaust gas detection data collected within the preset time period, the third filter data and the third filter error corresponding to the first second exhaust gas detection data collected after the preset time period are determined. For the kth second exhaust gas detection data collected after the preset time period, the first Kalman filter algorithm is used, and the third filter data and third filter error corresponding to the (k-1)th second exhaust gas detection data collected after the preset time period are used to determine the third filter data and third filter error corresponding to the kth second exhaust gas detection data collected after the preset time period; the third filter error is determined based on the second exhaust gas detection data and its corresponding third filter data.

7. The method according to claim 1, characterized in that, Determining the cleaning status of the semiconductor chamber based on the first filtered data, the second filtered data, and the third filtered data includes: If the first filtered data and the second filtered data meet preset conditions, the cleaning process switching of the semiconductor chamber is determined to be successful; the preset conditions include: the smoothness of the first filtered data and the second filtered data is different, and / or, the second filtered data has burrs; or The endpoint of the second cleaning process is determined based on the steepness of the third filtered data.

8. A semiconductor device, characterized in that, It includes a semiconductor chamber, an endpoint detection device, and a control unit; among which, The semiconductor chamber is used to perform the first cleaning process and the second cleaning process; The endpoint detection device is used to detect the exhaust gas in the semiconductor chamber during the execution of the first cleaning process to obtain first exhaust gas detection data; and to detect the exhaust gas in the semiconductor chamber during the execution of the second cleaning process to obtain second exhaust gas detection data. The control unit includes a memory and a processor. The memory stores a computer program, and the processor is used to call and execute the computer program from the memory to implement the semiconductor chamber cleaning method as described in any one of claims 1-7.

9. A cleaning device for a semiconductor chamber, characterized in that, The device includes a processor and a memory electrically connected to the processor, the memory storing a computer program, and the processor calling and executing the computer program from the memory to implement the semiconductor chamber cleaning method as described in any one of claims 1-7.

10. A storage medium, characterized in that, The storage medium is used to store a computer program, which is executed by a processor to implement the method for cleaning a semiconductor chamber as described in any one of claims 1-7.

Images