Cavity temperature rise monitoring method and semiconductor process equipment

Abstract

The invention relates to the technical field of semiconductor manufacturing, and provides a cavity temperature rise monitoring method and semiconductor process equipment. The method comprises the steps that real-time data of related variables influencing the heating process of the process chamber are obtained, the real-time temperature actual value of a heating part in the process chamber is obtained, the related variables comprise continuous variables and digital variables, and the digital variables comprise the water passing state and the temperature control mode of the process chamber; the real-time data of the related variables are input into a temperature prediction model obtained through pre-training, the real-time temperature prediction value of the heating component is output through the temperature prediction model, and the temperature prediction model is obtained through training based on the historical temperature actual value of the heating component and the historical data of the related variables; by comparing the real-time actual temperature value of the heating component with the real-time predicted temperature value of the heating component, the real-time monitoring result of the temperature rising process of the process chamber is obtained, so that the accuracy of the monitoring result of the temperature rising process of the process chamber is improved.

Description

Technical Field

[0001] This specification relates to the field of semiconductor manufacturing technology, and particularly to a method for monitoring chamber temperature rise and a semiconductor process equipment. Background Art

[0002] In the field of semiconductor manufacturing technology, when processing semiconductors through a process chamber, for example, when processing wafers through a process chamber, it is often necessary to maintain the process chamber at a relatively high temperature. Correspondingly, it is necessary to monitor the temperature rise process of the process chamber. Currently, the conventional monitoring of the process chamber temperature rise process mainly relies on monitoring the critical set values of heating parameters such as the power and current of the heating component. The critical set values of the heating parameters are mainly set manually based on experience and experimental data, and different critical value settings have different effects on the chamber state monitoring. The monitoring results of the process chamber temperature rise process relying on the critical set values are inaccurate, which in turn affects the normal operation of the process chamber and even leads to component damage and other situations. Summary of the Invention

[0003] To solve the above technical problems, this application provides a method for monitoring chamber temperature rise and a semiconductor process equipment to improve the accuracy of the monitoring results of the process chamber temperature rise process.

[0004] To achieve the above technical purpose, the embodiments of this application provide the following technical solutions:

[0005] In a first aspect, an embodiment of this specification provides a method for monitoring chamber temperature rise, which is applied to a process chamber, and the process chamber includes a heating component; the method includes: obtaining real-time data of relevant variables affecting the process chamber temperature rise process, and obtaining the actual real-time temperature value of the heating component in the process chamber, where the relevant variables include continuous variables and digital variables, and the digital variables include the water passing state and temperature control mode of the process chamber; inputting the real-time data of the relevant variables into a pre-trained temperature prediction model, and outputting the predicted real-time temperature value of the heating component through the temperature prediction model, where the temperature prediction model is trained based on the historical actual temperature value of the heating component and the historical data of the relevant variables; obtaining the real-time monitoring result of the process chamber temperature rise process by comparing the actual real-time temperature value of the heating component with the predicted real-time temperature value of the heating component.

[0006] According to the chamber temperature rise monitoring method provided by the first aspect, after obtaining the real-time monitoring result of the process chamber temperature rise process by comparing the actual real-time temperature value of the heating component with the predicted real-time temperature value of the heating component, it further includes: if the real-time monitoring result indicates that the process chamber temperature rise is normal, updating the real-time data of the relevant variables to the historical data of the relevant variables, and updating the actual real-time temperature value of the heating component to the historical actual temperature value of the heating component; wherein, the updated historical actual temperature value of the heating component and the updated historical data of the relevant variables are used to retrain the temperature prediction model.

[0007] According to the chamber temperature rise monitoring method provided by the first aspect, before training the temperature prediction model, the historical data of the relevant variables and the historical actual temperature value are divided into a training set and a validation set according to a preset ratio; the training process of the temperature prediction model is as follows: inputting the historical data of the relevant variables in the training set into a preset original prediction model, and outputting the predicted training temperature value of the heating component through the original prediction model, wherein the original prediction model is pre-configured based on a long short-term memory neural network; comparing whether the predicted training temperature value of the heating component is less than a training threshold with the historical actual temperature value of the heating component; if so, after testing and optimizing the parameters in the original prediction model through the historical data of the relevant variables in the validation set, determining the original prediction model as the temperature prediction model; if not, adjusting the internal parameters of the original prediction model and retraining the original prediction model until the temperature prediction model is obtained.

[0008] According to the chamber temperature rise monitoring method provided by the first aspect, the temperature prediction model includes an input layer, a first operation layer, a second operation layer, a fully connected layer, a random dropout layer, and an output layer; the input layer, the first operation layer, the second operation layer, the fully connected layer, the random dropout layer, and the output layer are connected in sequence.

[0009] According to the chamber temperature rise monitoring method provided in the first aspect, the step of inputting the real-time data of the relevant variables into a pre-trained temperature prediction model and outputting the real-time temperature prediction value of the heating component by the temperature prediction model includes: inputting the real-time data of the relevant variables into the input layer; inputting the intermediate data output by the input layer into the first operation layer, and the first operation layer performs the first data processing transformation; inputting the intermediate data output by the first operation layer into the second operation layer, and the second operation layer performs the second data processing transformation; inputting the intermediate data output by the second operation layer into the fully connected layer, and the fully connected layer performs regularization processing; inputting the intermediate data output by the fully connected layer into the dropout layer, and neurons are randomly lost according to a preset ratio during the operation of the dropout layer; inputting the intermediate data output by the dropout layer into the output layer, and the output layer performs regularization processing and outputs the real-time temperature prediction value of the heating component.

[0010] According to the chamber temperature rise monitoring method provided in the first aspect, the step of acquiring the real-time data of the relevant variables during the temperature rise of the process chamber includes: collecting the real-time data of the relevant variables during the temperature rise of the process chamber; performing standardization processing on the real-time data of the continuous variables and one-hot encoding on the real-time data of the digital variables to obtain the standardized data of the relevant variables; the step of inputting the real-time data into a pre-trained temperature prediction model and outputting the real-time temperature prediction value of the heating component by the temperature prediction model includes: inputting the standardized data of the relevant variables into the pre-trained temperature prediction model, and outputting the real-time temperature prediction value of the heating component by the temperature prediction model.

[0011] According to the chamber temperature rise monitoring method provided in the first aspect, the step of obtaining the real-time monitoring result of the temperature rise process of the process chamber by comparing the actual real-time temperature of the heating component with the predicted real-time temperature of the heating component includes: calculating the difference between the actual real-time temperature of the heating component and the predicted real-time temperature of the heating component; if the difference is less than or equal to a preset temperature threshold, generating a real-time monitoring result that the temperature rise of the process chamber is normal; if the difference is greater than the temperature threshold, generating a real-time monitoring result that the temperature rise of the process chamber is abnormal.

[0012] According to the chamber temperature rise monitoring method provided in the first aspect, the historical data of the relevant variables is historical time-series data of the relevant variables, and the historical time-series data includes the historical data of the relevant variables corresponding to multiple consecutive historical moments during the historical normal temperature rise process of the process chamber.

[0013] According to the chamber temperature rise monitoring method provided by the first aspect, the continuous variables include one or more of the inner heating wire power, the outer heating wire power, the inner heating wire current, and the outer heating wire current; the digital variables include the chamber water flow state and / or the chamber temperature control mode.

[0014] In a second aspect, an embodiment of the present specification provides a semiconductor processing apparatus, including: a process chamber, the process chamber including a heating component; an upper electrode assembly located above the process chamber; a lower electrode assembly located in the process chamber for carrying a wafer and applying a bias voltage to the wafer; a controller including at least one memory and at least one processor, the memory being used for storing a computer program; the processor being used for implementing the chamber temperature rise monitoring method as described in any one of the above by running the computer program stored in the memory.

[0015] According to the semiconductor processing apparatus provided by the second aspect, the process chamber is a physical vapor deposition process chamber, and the heating component includes an electrostatic chuck; the semiconductor processing apparatus further includes a power controller, the controller being used for determining a temperature control mode and transmitting power control parameters in the temperature control mode to the power controller; the power controller being used for adjusting the real-time temperature of a heater based on the power control parameters, and the temperature of the heater being conducted to the electrostatic chuck.

[0016] In a third aspect, an embodiment of the present specification provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is run by a processor, the chamber temperature rise monitoring method as described above is implemented.

[0017] In a fourth aspect, an embodiment of the present specification provides a computer program product or a computer program, the computer program product including a computer program, the computer program being stored in a computer-readable storage medium; a processor of the computer device reads the computer program from the computer-readable storage medium, and when the processor executes the computer program, the chamber temperature rise monitoring method as described above is implemented.

[0018] As can be seen from the above technical solutions, the embodiments of the present application provide a method for monitoring the temperature rise of a chamber and a semiconductor process equipment. Among them, in the method for monitoring the temperature rise of the chamber, the real-time data of relevant variables in the process of heating the process chamber is input into a pre-trained temperature prediction model to obtain the real-time temperature prediction value of the heating component output by the temperature prediction model. Then, by comparing the actual real-time temperature of the heating component with the real-time temperature prediction value of the heating component, the real-time monitoring result of the process of heating the process chamber is obtained. The temperature prediction model is trained based on the actual historical temperature of the heating component and the historical data of relevant variables, realizing the intelligent monitoring process of the chamber temperature rise, avoiding the uncertainty caused by manually setting the critical value, and improving the accuracy of the monitoring result of the process of heating the process chamber. At the same time, the relevant variables in the process of heating the process chamber of the present application include not only continuous variables used in traditional methods, but also digital variables such as the water supply state and temperature control mode of the process chamber. Digital variables are more stable and less susceptible to interference, improving the stability of the chamber temperature rise monitoring process. At the same time, the water supply state and temperature control mode can characterize the actual state of the process chamber itself, making the monitoring process of the process chamber more comprehensive and further improving the accuracy of the monitoring result of the temperature rise process. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0020] Figure 1 It is a schematic diagram of the structure principle of a PVD Al chamber provided by the prior art;

[0021] Figure 2 It is one of the schematic diagrams of the changes of relevant variables when the temperature rise of the chamber is abnormal provided by the prior art;

[0022] Figure 3 It is the second of the schematic diagrams of the changes of relevant variables when the temperature rise of the chamber is abnormal provided by the prior art;

[0023] Figure 4 It is a schematic diagram of the flow steps of the method for monitoring the temperature rise of the chamber provided by an embodiment of the present specification;

[0024] Figure 5 It is a schematic diagram of the structural connection of the temperature prediction model provided by an embodiment of the present specification;

[0025] Figure 6 It is a schematic diagram of the process from the model training stage to the model use stage of the method for monitoring the temperature rise of the chamber provided by an embodiment of the present specification;

[0026] Figure 7 A structural schematic diagram of a semiconductor process equipment provided for an embodiment of this specification. Specific embodiments

[0027] Unless otherwise defined, the technical terms or scientific terms used in the embodiments of this specification should have the ordinary meanings understood by those of ordinary skill in the field to which this specification belongs. The "first", "second" and similar terms used in the embodiments of this specification do not denote any order, quantity or importance, but are only used to avoid confusion of components.

[0028] Unless otherwise required by the context, throughout the specification, "a plurality of" means "at least two", and "including" is interpreted as an open and inclusive meaning, that is, "including, but not limited to". In the description of the specification, the terms "one embodiment", "some embodiments", "exemplary embodiments", "examples", "specific examples" or "some examples", etc., are intended to indicate that the specific features, structures, materials or characteristics related to the embodiment or example are included in at least one embodiment or example of this specification. The schematic representations of the above terms do not necessarily refer to the same embodiment or example.

[0029] Next, the technical solutions in the embodiments of this specification will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all the embodiments. Based on the embodiments in this specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope protected by this specification.

[0030] Overview

[0031] As described in the background art, during the preparation of semiconductor devices, when preparing semiconductors through a process chamber, it is often necessary to maintain the process chamber at a relatively high temperature. In particular, Physical Vapor Deposition (PVD) technology mainly refers to Magnetron Sputtering technology. This technology mainly deposits thin films by bombarding solid targets with plasma, and is mainly used for the deposition of metal thin films. It is a widely used thin film manufacturing process in the semiconductor industry. In the PVD process flow, the deposition of metal thin films is often accompanied by a certain high-temperature environment. The chamber controls the output of the heater through a temperature controller to achieve the temperature control target.

[0032] In the PVD process flow, the chamber heating process mainly experiences two temperature control modes: the Ramp mode and the Servo mode. In the Ramp control mode, when the deviation between the actual temperature value and the set value of the heating component is large in the early stage of heating, the temperature controller (referred to as the temperature controller) controls the temperature of the heating component to increase at a fixed heating rate (for example: 3°C / min). The Servo control mode means that during the heating process, when the actual temperature value is about to reach the set value (for example, the difference between the actual value and the set value is 7°C), at this time, it switches from the Ramp mode to the Servo mode, and the temperature controller controls the actual temperature to slowly approach the set value. In the early stage of heating, when the difference between the temperature set value and the actual value is large, the temperature controller is in the Ramp temperature control mode. At this time, the temperature of the electrostatic chuck (ESC), which is a heating component in the process chamber, is controlled at a fixed heating rate. When the actual value of the ESC temperature gradually approaches the set value (that is, when the temperature difference is less than a certain fixed value), the temperature controller is in the Servo mode, causing the ESC temperature to rise slowly.

[0033] Exemplarily, such as Figure 1 Taking the PVD Al chamber shown as an example, the temperature controller determines the control parameters for the power controller in different temperature control modes according to the temperature set value and the water flow state of the chamber, as well as the temperature difference between the actual temperature value and the set value of the electrostatic chuck. It acts on the inner and outer heating wires of the heater (Heater), making the inner and outer heating wires heat up at the set power, and conducts the temperature to the heating component through the back-blowing method. The heating component includes a ceramic disc ESC, and then conducts the temperature to the silicon wafer (also known as the wafer, Wafer) through the heating component, achieving the heating target of increasing the chamber temperature by a fixed value per minute. In addition, a water pipe is pre-laid in the heater and communicates with the outside. Cooling water is stored in the water pipe, and the flow of the cooling water in the water pipe is controlled by a valve. According to the process requirements, when it is necessary to reduce the temperature of the heater and the electrostatic chuck, the flow of the cooling water is controlled by the valve, thereby reducing the temperature of the heater. The high-temperature electrostatic chuck conducts heat to the relatively low-temperature heater, realizing the cooling of the electrostatic chuck, and then meeting the cooling requirements during wafer processing. The water flow state of the water pipe in the chamber can be obtained by monitoring the state of the control valve of the water pipe. When the valve is in the open state, it means that the water flow state of the chamber is through water. When the valve is in the closed state, it means that the water flow state of the chamber is non-through water.

[0034] During the heating process, since the ESC is made of ceramic and has a high cost, when a hardware failure occurs in the heating circuit, it is easy to cause abnormal power output of the heating wire, resulting in a sudden rise or fall in the chamber temperature, which may cause the ceramic ESC disc to crack. Therefore, it is very necessary to monitor the state of the chamber heating process and timely detect abnormal states during the heating process.

[0035] In the prior art, the state monitoring during the chamber heating process is mainly carried out by monitoring the mutual relationship between several variables. Taking the Figure 1 shown PVD Al chamber as an example, during the heating process, the chamber mainly monitors variables such as the actual temperature value of the ESC, the current of the inner heating wire, the power of the inner heating wire, the current of the outer heating wire, and the power of the outer heating wire. The temperature controller compares the actual temperature of the ESC with the set temperature value and determines the set value of the power controller in the corresponding temperature control mode according to the water passing state of the chamber, so as to realize the control of the heating wire. When a fault occurs in the heating circuit, the output power and current of the inner and outer circles of the heater will deviate from the normal values and show abnormal fluctuations.

[0036] As Figure 2 shown, when a certain type of fault occurs, the power of the inner heating wire and the power of the outer heating wire shown in the upper coordinate system both fluctuate between 40% and 60%, while the current of the outer heating wire shown in the lower coordinate system suddenly drops from about 2.4 A to about 0.2 A. After the fault occurs, the actual temperature of the ESC decreases abnormally. Based on this, the existing solution determines this type of abnormal heating by monitoring whether a small current (less than 0.5 A) appears when the inner heating wire or the outer heating wire has a high power output (greater than or equal to 20%). 20% and 0.5 A are the critical values set manually in advance.

[0037] As Figure 3 shown, when another type of fault occurs, the current of the inner heating wire and the current of the outer heating wire shown in the lower coordinate system first mutate, and the current frequently jumps between 0 A and 5.6 A in a short time. At this time, the power of the inner heating wire and the power of the outer heating wire shown in the upper coordinate system have not started to change, but after a period of time, the power of the inner heating wire and the power of the outer heating wire decrease rapidly. Correspondingly, the actual temperature of the ESC decreases abnormally. Therefore, by monitoring whether a small current (less than 0.5 A) appears when the inner heating wire or the outer heating wire has a high power output (greater than or equal to 20%), it can be determined whether the chamber heating is abnormal. 20% and 0.5 A are the critical values set manually in advance.

[0038] Based on the above, the conventional state monitoring of the chamber heating process is determined by the mutual relationship between the power and current of the inner heating wire and the outer heating wire. When the power is maintained at a certain value, if the current is lower than a specific value, it is determined that the chamber heating is abnormal during this period, the machine alarms, and emergency cooling measures are taken.

[0039] However, for the state monitoring of the conventional chamber heating process, the determination of the chamber heating circuit failure depends on the critical setting values of the power and current of the inner and outer heating wires (such as 20% and 0.5 A). The values of the critical setting values mainly come from experience and experimental data, and different values of the critical setting values have different effects on the chamber state monitoring. If the value of the critical setting value is too small, the effect of the state monitoring of the chamber heating process is too sensitive, which is prone to false alarms and seriously affects the chamber process efficiency; if the value of the critical setting value is too large, when the chamber heating is abnormal, the machine response is not timely, which is prone to faults such as the fragmentation of the ESC ceramic disc. In addition, the conventional state monitoring technology for the chamber heating process monitors the current, power, etc. When the current and power signals are invalid or interfered, it is impossible to sense the abnormal temperature change of the ESC when a fault occurs, and the error tolerance rate is relatively low.

[0040] Through the above analysis, the present application provides a chamber heating monitoring method and a semiconductor process equipment. The chamber heating monitoring method is based on a pre-trained temperature prediction model to realize the real-time monitoring of the state of the chamber heating process. The temperature prediction model can be trained through the historical actual temperature values of the heating components and the historical data of relevant variables, and the real-time temperature prediction value of a heating component can be obtained according to the real-time data of the relevant variables. Then, by comparing the real-time actual temperature value of the heating component with the real-time temperature prediction value of the heating component, the real-time monitoring result of the chamber heating process is obtained. If the deviation between the real-time actual temperature value of the heating component and the real-time temperature prediction value of the heating component is large, the real-time monitoring result indicates that the chamber heating is abnormal, indicating that the temperature of the heating component seriously deviates from the original temperature trend. At this time, there may be a hardware fault, and the industrial control computer throws an alarm and performs a protection action; if the deviation between the real-time actual temperature value of the heating component and the real-time temperature prediction value of the heating component is small, the real-time monitoring result indicates that the chamber heating is normal, indicating that the real-time temperature value of the heating component is normal, and the chamber can continue the next operation.

[0041] Based on the above inventive concept, the chamber heating monitoring method provided by the embodiments of the present specification is described exemplarily below.

[0042] Exemplary Method

[0043] The embodiment of this specification provides a method for monitoring the temperature rise of a chamber. This method is applied to a process chamber to monitor the temperature rise process of the process chamber. The process chamber includes a heating component, which can be an electrostatic chuck in the process chamber. The temperature controller configured for the process chamber can adjust the temperature control mode. The temperature controller adjusts the heating power of the inner and outer heating wires in the heater based on different temperature control models. The temperature of the heater changes accordingly, and the temperature conducted to the electrostatic chuck also changes, thereby adjusting the temperature of the entire process chamber. The chamber temperature rise monitoring method can be implemented in the temperature controller to achieve real-time monitoring of the process chamber temperature rise process.

[0044] As Figure 4 shown, the chamber temperature rise monitoring method includes:

[0045] Step 401, obtain the real-time data of relevant variables affecting the process chamber temperature rise process, and obtain the actual real-time temperature value of the heating component in the process chamber. Among them, the relevant variables include continuous variables and digital variables, and the digital variables include the water passing state and temperature control mode of the process chamber.

[0046] In this embodiment, when any variable changes during the process chamber temperature rise process and affects the process chamber temperature rise process, then this variable is a relevant variable affecting the process chamber temperature rise process. For example, the power of the inner heating wire, the current of the outer heating wire, etc. One or more relevant variables can be selected according to the actual situation and needs. When there are multiple relevant variables, each relevant variable is collected for real-time data in a suitable manner. For example, the current is collected through a current sampling circuit, and the power is collected using a power meter.

[0047] In this embodiment, the heating component in the process chamber refers to the main component that can increase the temperature inside the process chamber. For example, in the PVD process flow, the heating component can be the Figure 1 electrostatic chuck described above. The actual real-time temperature value of the heating component in the process chamber can also be collected in a suitable manner according to the actual situation and needs. For example, it can be collected through a temperature sensor. Collect the real-time data of relevant variables affecting the process chamber temperature rise process at present, and the actual real-time temperature value of the heating component in the process chamber for subsequent processing.

[0048] In this embodiment, the process of collecting the real-time data of relevant variables and the actual real-time temperature value of the heating component can be continuously collected throughout the process chamber temperature rise process according to a preset cycle. Each time it is collected, the following processing process is completed to obtain a real-time monitoring result, thereby realizing continuous monitoring of the process chamber temperature rise process.

[0049] In this embodiment, in order to ensure a more comprehensive characterization of the process chamber heating process, the relevant variables include as many different types of variables as possible. The relevant variables include continuous variables and digital variables; continuous variables refer to variables that change continuously over time, such as current, and digital variables refer to variables that do not change continuously over time, such as mode. The relevant variables in the process chamber heating process include not only continuous variables such as current and power used in traditional methods, but also digital variables such as the water flow state and temperature control mode of the process chamber. Digital variables are more stable and less susceptible to interference, improving the stability of the chamber heating monitoring process. At the same time, the water flow state and temperature control mode can characterize the actual state of the process chamber itself. Multiple variables such as continuous variables and digital variables increase the diversity of the basic data, can more comprehensively characterize the heating state of the process chamber, and make the monitoring results of the heating process more accurate.

[0050] Step 402: Input the real-time data of the relevant variables into the pre-trained temperature prediction model, and output the real-time temperature prediction value of the heating component through the temperature prediction model, where the temperature prediction model is trained based on the historical actual temperature value of the heating component and the historical data of the relevant variables.

[0051] In this embodiment, the real-time data of the relevant variables is input into the pre-trained temperature prediction model, which is used to predict the temperature value of the heating component, that is, the temperature prediction model outputs the real-time temperature prediction value of the heating component. The temperature prediction model is trained based on the historical actual temperature value of the heating component and the historical data of the relevant variables. Using the historical actual temperature value of the heating component and the historical data of the relevant variables as training data can make the prediction process of the temperature prediction model more conform to the actual temperature change process of the heating component, improve the accuracy of the real-time temperature prediction value of the heating component output by the temperature prediction model, and thus improve the accuracy of the chamber heating monitoring. In addition, through the temperature prediction model, the process of manually setting the critical set value is replaced, improving the intelligent level of the chamber heating monitoring process, avoiding increasing labor costs and time costs, and improving the processing efficiency of the chamber heating monitoring process.

[0052] Step 403: Obtain the real-time monitoring result of the process chamber heating process by comparing the real-time actual temperature value of the heating component with the real-time temperature prediction value of the heating component.

[0053] In this embodiment, after obtaining the real-time temperature prediction value of the heating component output by the temperature prediction model, the real-time actual temperature value of the heating component collected in real time is compared with the real-time temperature prediction value of the heating component to determine whether the temperature rise process of the process chamber is normal at the current moment, so as to obtain the real-time monitoring result of the temperature rise process of the process chamber. Since the temperature prediction model is trained based on the historical actual temperature values of the heating component and the historical data of related variables, the intelligent monitoring process of the chamber temperature rise is realized, the uncertainty caused by manual setting of the critical value is avoided, and the accuracy of the monitoring result of the temperature rise process of the process chamber is improved.

[0054] In one embodiment, after obtaining the real-time monitoring result of the temperature rise process of the process chamber by comparing the real-time actual temperature value of the heating component with the real-time temperature prediction value of the heating component, if the real-time monitoring result indicates that the temperature rise of the process chamber is normal, the real-time data of the related variables is updated to the historical data of the related variables, and the real-time actual temperature value of the heating component is updated to the historical actual temperature value of the heating component; wherein, the updated historical actual temperature value of the heating component and the updated historical data of the related variables are used to retrain the temperature prediction model.

[0055] In this embodiment, in order to further improve the monitoring accuracy, if it is determined that the temperature rise of the process chamber is normal based on the real-time data of the related variables and the real-time actual temperature value of the heating component collected this time, the real-time data of the related variables can be updated to the historical data of the related variables, and the real-time actual temperature value of the heating component can be updated to the historical actual temperature value of the heating component. The updated historical actual temperature value of the heating component and the updated historical data of the related variables are used to retrain the temperature prediction model, so that the temperature prediction model can continuously learn with the temperature rise process of the process chamber, be more in line with the actual situation of the current temperature rise process of the process chamber, further improve the accuracy of the real-time temperature prediction value of the heating component output by the temperature prediction model, and thus improve the accuracy of the chamber temperature rise monitoring.

[0056] In this embodiment, if the real-time monitoring result indicates that the temperature rise of the process chamber is normal, the real-time data of the relevant variables just obtained and the actual real-time temperature of the heating component can be used to update the training data of the temperature prediction model. The historical actual temperature of the updated heating component and the historical data of the updated relevant variables are used to retrain the temperature prediction model, making the internal logic of the temperature prediction model more suitable for the current temperature rise process of the process chamber, greatly improving the accuracy of the real-time temperature prediction value of the heating component output by the temperature prediction model, and further improving the accuracy of the real-time monitoring result during the temperature rise process of the process chamber. Compared with the method of determining the chamber state by monitoring the relationship between the inner and outer circuits and the power, the temperature prediction model trained through continuous learning can quickly determine whether an abnormality occurs during the temperature rise process of the process chamber, has strong anti-interference ability, and is not prone to false alarms.

[0057] In one embodiment, to facilitate the efficient operation of the temperature prediction model, preprocessing is first performed on different types of variables to achieve data standardization. Specifically, the real-time data of the relevant variables during the temperature rise process of the process chamber is obtained as follows: collect the real-time data of the relevant variables during the temperature rise process of the process chamber; perform standardization processing on the real-time data of continuous variables and one-hot encoding on the real-time data of digital variables to obtain the standardized data of the relevant variables. When the real-time data is input into the pre-trained temperature prediction model and the real-time temperature prediction value of the heating component is output through the temperature prediction model, the standardized data of the relevant variables is input into the pre-trained temperature prediction model, and the real-time temperature prediction value of the heating component is output through the temperature prediction model.

[0058] In one embodiment, in the PVD process flow, preferably, the continuous variables include one or more of the inner ring power of the heating wire, the outer ring power of the heating wire, the inner ring current of the heating wire, and the outer ring current of the heating wire; the digital variables include the water supply state of the chamber and / or the chamber temperature control mode. Of course, according to the different operating environments of the specific process chamber, more relevant variables suitable for the actual operating environment can be selected for the continuous variables and the digital variables.

[0059] In an exemplary embodiment, taking the PVD process flow as an example, the relevant variables include the inner ring power of the heating wire, the outer ring power of the heating wire, the inner ring current of the heating wire, the outer ring current of the heating wire, the water supply state of the chamber, and the chamber temperature control mode. Among them, the inner ring current of the heating wire (unit: A), the outer ring current of the heating wire (unit: A), the inner ring power of the heating wire (unit: W), and the outer ring power of the heating wire (unit: %) are continuous variables, while the water supply state of the chamber (0 indicates no water supply, 1 indicates water supply) and the chamber temperature control mode (0 indicates Servo mode, 1 indicates Ramp mode) are digital variables.

[0060] In this embodiment, during the heating process of the process chamber, after collecting the real-time data of the above relevant variables, different preprocessing is performed on the real-time data according to the different data types.

[0061] For any continuous variable, to avoid overfitting of the model, mean-standard deviation (mean-std) normalization is performed on it. Specifically, based on the historical data of the relevant variables used for model training, the minimum (min) and maximum (max) values in the historical data are used to convert the real-time data into values within the range of 0-1. The calculation formula is as follows:

[0062]

[0063] where x min represents the minimum value in the historical data of the relevant variable, x max represents the maximum value in the historical data of the relevant variable, x represents the real-time data of the relevant variable before normalization, and X represents the normalized data of the relevant variable after normalization.

[0064] For digital variables (chamber water supply status and chamber temperature control mode), one-hot encoding is performed on them according to the combination method. One-hot encoding can solve the problem that the model is not good at processing discrete data and can also play a role in data augmentation to a certain extent. The encoding rules are shown in Table 1 below:

[0065] Table 1 One-Hot Encoding Table for Water Supply Status and Temperature Control Mode

[0066]

[0067] According to the real-time data (i.e., real-time status) of the chamber water supply status and chamber temperature control mode, select one of the above required one-hot encodings.

[0068] Through data preprocessing, the data processing difficulty of the subsequent temperature prediction model is reduced, and the data processing efficiency of the temperature prediction model is improved, thereby improving the processing efficiency of the entire chamber heating monitoring process.

[0069] In one embodiment, the temperature prediction model is pre-trained. Before training the temperature prediction model, the historical data of the relevant variables and the actual historical temperature values are divided into a training set and a validation set according to a preset ratio; the training set is used to train the temperature prediction model, and the test set is used to test the generalization ability of the temperature prediction model and is used to test and optimize the trained temperature prediction model.

[0070] The training process of the temperature prediction model is as follows: Input the historical data of relevant variables in the training set into a preset original prediction model, and output the predicted training temperature value of the heating component through the original prediction model, where the original prediction model is pre-configured based on a long short-term memory neural network; Compare whether the predicted training temperature value of the heating component is less than the training threshold with the actual historical temperature value of the heating component; If so, after testing and optimizing the parameters in the original prediction model through the historical data of relevant variables in the validation set, determine the original prediction model as the temperature prediction model; If not, adjust the internal parameters of the original prediction model and retrain the original prediction model until the temperature prediction model is obtained.

[0071] In this embodiment, the preset ratio for dividing into the training set and the validation set is set according to the actual situation and needs. Preferably, the actual historical temperature value of the heating component and the historical data of relevant variables are divided into the training set and the test set according to a ratio of 8:2.

[0072] In this embodiment, since the temperature rise process of the process chamber is a process that changes with time and is strongly correlated with time changes. Then the original prediction model is pre-configured based on a long short-term memory neural network (Long-Short Term Memory, LSTM). LSTM is a variant of the recurrent neural network (Recurrent Neural Network, RNN). RNN is a type of neural network used to process sequential data, drawing on the thinking process of the human brain's connection before and after in the reasoning and operation process, emphasizing the influence of memory on the input and output of the neural network. And LSTM, on the basis of RNN, introduces a gating mechanism, which is suitable for processing time series data samples and is more suitable for predicting the real-time temperature during the temperature rise process of the process chamber.

[0073] In one embodiment, in order to better train the temperature prediction model, the historical data of relevant variables is the historical time series data of relevant variables, and the historical time series data includes the historical data of relevant variables corresponding to multiple consecutive historical moments during the normal temperature rise process of the process chamber.

[0074] In this embodiment, the historical data of relevant variables for training the temperature prediction model is the historical time series data of relevant variables. Each relevant variable stores the data in chronological order (i.e., the order of multiple consecutive historical moments). The actual historical temperature value of the heating component is also stored in chronological order corresponding to the historical data of relevant variables, that is, the historical time series data of relevant variables collected at a certain historical moment corresponds to the actual historical temperature value of the heating component at that historical moment.

[0075] In this embodiment, for the convenience of training, the historical data of relevant variables can be preprocessed. The preprocessing process is similar to the real-time data preprocessing process of the relevant variables provided in the above embodiment. It should be noted that, in order to further improve the training efficiency, the historical actual temperature values of the heating components can be preprocessed according to continuous variables.

[0076] In an exemplary embodiment, taking the PVD process flow as an example, the relevant variables include the inner-ring power of the heating wire, the outer-ring power of the heating wire, the inner-ring current of the heating wire, the outer-ring current of the heating wire, the water passing state of the chamber, and the chamber temperature control mode. Among them, the inner-ring current of the heating wire (unit: A), the outer-ring current of the heating wire (unit: A), the inner-ring power of the heating wire (unit: %), and the outer-ring power of the heating wire (unit: %) are continuous variables, while the water passing state of the chamber (0 means no water passing, 1 means water passing) and the chamber temperature control mode (0 means Servo mode, 1 means Ramp mode) are digital variables.

[0077] In this embodiment, according to the different data types, different preprocessing is performed on the historical time-series data of relevant variables.

[0078] For any continuous variable or the historical actual temperature value of the heating component, to avoid overfitting of the model, mean (mean-) variance (std) normalization is performed on it. Specifically, based on the historical data of relevant variables used for model training, using the minimum (min) value and the maximum (max) value in the historical data, the real-time data is converted into a value within the range of 0-1. The calculation formula is as follows:

[0079]

[0080] where x min represents the minimum value in the historical data of this relevant variable (or the historical actual temperature value of the heating component), x max represents the maximum value in the historical data of the relevant variable (or the historical actual temperature value of the heating component), x represents the historical data of this relevant variable (or the historical actual temperature value of the heating component) before normalization, and X represents the historical data of this relevant variable (or the historical actual temperature value of the heating component) after normalization.

[0081] For digital variables (the water passing state of the chamber and the chamber temperature control mode), one-hot encoding is performed on them according to the combination method. One-hot encoding can solve the problem that the model is not good at processing discrete data, and can also play a role in data augmentation features to a certain extent. The encoding rule is the same as Table 1 above.

[0082] In addition, to exclude abnormal data according to the trend, it may also be necessary to randomly discard a certain proportion (such as 10%) of the historical data.

[0083] In one embodiment, in order to further improve the prediction accuracy of the temperature prediction model, when constructing the basic structure of the temperature prediction model, targeted adjustments are made based on the temperature rise process of the process chamber. Specifically, as Figure 5 shown, the temperature prediction model includes an input layer (Input Layer), a first operation layer (LSTM), a second operation layer (LSTM), a fully connected layer (Dense), a dropout layer (Dropout), and an output layer; the input layer, the first operation layer, the second operation layer, the fully connected layer, the dropout layer, and the output layer are connected in sequence. Among them, the output layer also adopts the basic structure of the fully connected layer.

[0084] In one embodiment, the real-time data of relevant variables is input into the pre-trained temperature prediction model, and the real-time temperature prediction value of the heating component is output through the temperature prediction model. The specific processing process is as follows: the real-time data of relevant variables is input into the input layer; the intermediate data output by the input layer is input into the first operation layer, and the first operation layer performs the first data processing transformation; the intermediate data output by the first operation layer is input into the second operation layer, and the second operation layer performs the second data processing transformation; the intermediate data output by the second operation layer is input into the fully connected layer, and the fully connected layer performs regularization processing; the intermediate data output by the fully connected layer is input into the dropout layer, and the dropout layer randomly discards neurons according to a preset ratio during operation; the intermediate data output by the dropout layer is input into the output layer, and the output layer performs regularization processing and outputs the real-time temperature prediction value of the heating component.

[0085] In this embodiment, the construction and training process of the temperature prediction model can utilize the Keras API. Keras is a simple and easy-to-use deep learning framework that provides some high-level APIs to conveniently construct and train neural network models. A temperature prediction model containing a double-layer LSTM network is constructed, and the historical actual temperature values of the input heating component and the historical data of relevant variables are introduced to train and optimize the model. The structural distribution of the temperature prediction model is shown in Table 2 below:

[0086] Table 2 Structural Distribution of Temperature Prediction Model

[0087]

[0088]

[0089] As can be seen from Table 2 above, a total of six layers are used to construct the temperature prediction model. Each layer can achieve the coupling transformation between the relevant variables and data of the input. Among them, the number of layers is only the structure of the neural network used in the temperature prediction model. After the relevant variables and data provided in this application are preprocessed, they are sent into the neural network through the input layer, and then after two LSTM layers, one Dense layer and Dropout layer, corresponding data processing and transformation are performed, and the output is sent out through the Dense layer as the output layer.

[0090] Specifically, in the training stage, the historical data of the relevant variables after preprocessing enters the input layer, and then is sent into the first operation layer and the second operation layer in sequence. After being deformed and operated by two LSTM layers, it reaches the fully connected layer (Dense). In the fully connected layer, L 1 -L 2 mix regularization is used to prevent the model from overfitting. Then it is sent into the Dropout layer. The Dropout layer can randomly discard a certain proportion of neurons in the hidden layer during each training process, which effectively suppresses the mutual dependence between neurons, helps to learn more powerful features, and can also prevent the model from overfitting. Finally, it is output through a fully connected layer (i.e., the output layer).

[0091] In this embodiment, in machine learning, regularization is used to prevent the model from overfitting and improve the generalization ability of the model. L 1 regularization adds the sum of the absolute values of the model coefficients to the loss function, while L 2 regularization is to add the sum of the squares of the model coefficients to the loss function, so as to prevent the model from falling into a local optimum during training and improve the generalization ability of the model. L 1 -L 2 regularization combines the above two regularization methods and is a more flexible regularization method. In the fully connected layer, L 1 -L 2 mix regularization, that is, when performing optimization training in the fully connected layer, by adding an L 1 -L 2 regularization term to the optimization objective function, adding a regular deviation term to the training objective to prevent the model from falling into a local optimal solution during the optimization process.

[0092] In this embodiment, when constructing the above temperature prediction model, the activation functions selected for each layer are as follows:

[0093] The gating function of the LSTM layer selects the sigmoid function, and the activation function selects the hyperbolic tangent tanh function; the fully connected layer selects the rectified linear unit ReLU as the activation function. Specifically as follows:

[0094]

[0095]

[0096] ReLU; f(x) = max(x, 0)

[0097] Among them, x represents the data input by the function, f(x) represents the result output by the function, e is a constant, and max represents the maximum value.

[0098] In this embodiment, after the original prediction model is constructed based on the above, the Hyperopt toolkit can be used to optimize the hyperparameters of the above model. Hyperopt is a Python library mainly used to intelligently search for the best parameters of an algorithm model. By optimizing the loss function of the overall original prediction model to be minimized, the optimal parameters of each layer are obtained, including the number of neurons, the learning rate, and the selection of the activation function, and the optimizer is used during the training process to minimize the mean square error between the model output and the actual output, thereby determining the parameters.

[0099] In this embodiment, when training the original prediction model, the BackPropagation Through Time (BPTT) algorithm is adopted. This algorithm is based on the chain rule of differentiation, unfolds the LSTM network according to time steps, and uses the forward propagation algorithm to train it, optimizing the weights and biases between adjacent neurons in the neural network, thereby reducing errors. In addition, during the model training process, the mean square error (MSE) is selected as the optimization loss function, and the mean absolute percentage error (MAPE) and the mean absolute error (MAE) are selected as the indicators to evaluate the training effect. The specific calculation formulas are as follows:

[0100]

[0101]

[0102]

[0103] Among them, T act represents the actual historical temperature value, T pre represents the predicted training temperature value, and N represents the number of actual historical temperature values in the test set.

[0104] Among them, the optimized loss function can be understood as the objective function. The goal of model optimization training is to make the predicted value closer to the actual value. Therefore, in the optimization process, the MSE loss function is adopted, and the goal of the optimization process is to minimize the mean square error between the predicted value and the actual value. After the model training is completed, when evaluating the model prediction results, the MAPE and MAE methods are directly used to evaluate the training effect of the model based on the deviation between the predicted value and the actual value. Generally speaking, the smaller the values of these two indicators, the better the training result of the model and the higher the prediction accuracy. The result of MAPE is generally a value between 0 and 1, but it is required that the value of Tact cannot be 0. Therefore, comprehensively comparing the values of MAPE and MAE can better evaluate the training effect of the model.

[0105] In one embodiment, by comparing the actual value of the real-time temperature of the heating component with the predicted value of the real-time temperature of the heating component, the real-time monitoring result of the temperature rise process of the process chamber is obtained as follows: Calculate the difference between the actual value of the real-time temperature of the heating component and the predicted value of the real-time temperature of the heating component; If the difference is less than or equal to the preset temperature threshold, generate a real-time monitoring result indicating that the temperature rise of the process chamber is normal; If the difference is greater than the temperature threshold, generate a real-time monitoring result indicating that the temperature rise of the process chamber is abnormal.

[0106] In this embodiment, the temperature threshold is set according to the actual situation and needs. After the temperature prediction model outputs the predicted value of the real-time temperature of the heating component, calculate the difference between the actual value of the real-time temperature of the heating component and the predicted value of the real-time temperature of the heating component. If the difference is less than the set temperature threshold (preferably, the temperature threshold can be set to 5% of the actual value of the real-time temperature of the heating component), it is determined that the chamber temperature rise is normal at this time, and the real-time data of the relevant variables is updated to the historical data of the relevant variables, and the actual value of the real-time temperature of the heating component is updated to the historical actual temperature value of the heating component, and the temperature prediction model is iteratively trained, and then enter the next sampling period; If the difference is greater than the set temperature threshold, it is determined that the chamber temperature rise is abnormal, and an alarm is triggered, and the real-time data of the corresponding relevant variables and the actual value of the real-time temperature of the heating component are discarded.

[0107] In an exemplary embodiment, taking the PVD process flow as an example, as Figure 6 shown, the process of the chamber temperature rise monitoring method from the model training stage to the model usage stage is as follows:

[0108] In the model training stage:

[0109] Obtain the historical actual temperature value of the heating component and the historical data of the relevant variables;

[0110] Perform data preprocessing on the historical actual temperature values of the heating component and the historical data of related variables to reduce the difficulty of data processing. Among them, continuous variables such as current, power, and temperature are preprocessed by min-max standardization, and digital variables such as the water passing state of the chamber and the temperature control mode of the chamber are encoded by One-Hot;

[0111] Construct an original prediction model based on the LSTM architecture;

[0112] Use the historical actual temperature values of the heating component and the historical data of related variables after preprocessing to complete the training, verification, and optimization of the original prediction model, and compile the finally obtained temperature prediction model for subsequent use.

[0113] In the real-time usage stage of the model:

[0114] Obtain the real-time data of related variables and the real-time actual temperature value of the heating component during the real-time operation of the process chamber;

[0115] Perform data preprocessing on the real-time data of related variables and the real-time actual temperature value of the heating component, and the preprocessing method used is consistent with the pre-training method used in the model training stage;

[0116] Input the real-time data of related variables after preprocessing into the trained temperature prediction model;

[0117] Output the real-time temperature prediction value of the heating component by the temperature prediction model;

[0118] Calculate the difference between the real-time actual temperature value of the heating component and the real-time temperature prediction value of the heating component, and compare this difference with a preset temperature threshold;

[0119] If the difference is less than or equal to the preset temperature threshold, it is determined that the process chamber is heating up normally at this time, and the real-time data of related variables and the real-time actual temperature value of the heating component collected this time are updated to the training data of the temperature prediction model, and the temperature prediction model is iteratively trained, and then enter the next sampling period;

[0120] If the difference is greater than the temperature threshold, it is determined that the process chamber is heating up abnormally at this time, trigger an alarm, and discard the real-time data of related variables and the real-time actual temperature value of the heating component collected this time.

[0121] The chamber temperature rise monitoring method provided by the embodiments of the present application inputs the real-time data of relevant variables in the process of raising the temperature of the process chamber into a pre-trained temperature prediction model to obtain the real-time temperature prediction value of the heating component output by the temperature prediction model. Then, by comparing the actual real-time temperature of the heating component with the real-time temperature prediction value of the heating component, the real-time monitoring result of the process of raising the temperature of the process chamber is obtained. The temperature prediction model is trained based on the actual historical temperature of the heating component and the historical data of relevant variables, realizing the intelligent monitoring process of chamber temperature rise, avoiding the uncertainty caused by manual setting of critical values, and improving the accuracy of the monitoring result of the process of raising the temperature of the process chamber. At the same time, the relevant variables in the process of raising the temperature of the process chamber of the present application include not only continuous variables used in traditional methods, but also digital variables such as the water passing state and temperature control mode of the process chamber. Digital variables are more stable and less susceptible to interference, improving the stability of the chamber temperature rise monitoring process. At the same time, the water passing state and temperature control mode can characterize the actual state of the process chamber itself, making the monitoring process of the process chamber more comprehensive and further improving the accuracy of the monitoring result of the temperature rise process.

[0122] Exemplary Device

[0123] Reference Figure 7 , the embodiment of the present specification also provides a semiconductor process equipment 100, including: a process chamber 20, an intake assembly 20A, an exhaust assembly ( Figure 7 not shown in the figure), an upper electrode assembly 20B, a lower electrode assembly 20C, and a controller ( Figure 7 not shown in the figure), and the process chamber includes a heating component ( Figure 7 not shown in the figure). The controller includes at least one processor and at least one memory, and a computer program is stored in the memory. When the computer program is executed by the processor, it implements the chamber temperature rise monitoring method described in any one of the above embodiments.

[0124] Exemplarily, the controller can be a host computer or a slave computer. Among them, the controller can open the valve of the intake assembly 20A to introduce the corresponding process gas into the interior of the process chamber 20; the controller can also control the opening degree of the valve of the intake assembly 20A to control the flow rate of the process gas. The controller can also evacuate the interior of the process chamber 20 through the exhaust assembly to control the pressure inside the process chamber 20.

[0125] The upper electrode assembly 20B includes a radio frequency coil group 21, an upper radio frequency power supply 23, and an upper matcher 25. The controller is further configured to control the upper radio frequency power supply 23 to provide upper electrode power to the radio frequency coil group 21 through the upper matcher 25, so that the radio frequency coil group 21 excites the process gas inside the process chamber 20 to generate plasma.

[0126] The radio frequency coil group 21 may include an inner coil 21A and an outer coil 21B disposed around the inner coil 21A. In an optional embodiment, during the etching step or the deposition step, the current supplied to the inner coil 21A is less than the current supplied to the outer coil 21B. Optionally, the ratio of the current supplied to the outer coil 21B to the current supplied to the inner coil 21A is greater than 1 and less than or equal to 9.

[0127] The lower electrode assembly 20C includes a wafer carrier device 22, a lower radio frequency power supply 24, and a lower matcher 26. The controller is further configured to control the lower radio frequency power supply 24 to provide lower electrode power to the lower electrode of the wafer carrier device 22 through the lower matcher 26, so that the lower electrode of the wafer carrier device 22 provides a radio frequency bias voltage, enabling the plasma to have a bombardment ability.

[0128] The wafer carrier device 22 can be used to carry a wafer, and the wafer may include a tungsten-containing layer 10. The processor realizes the chamber temperature rise monitoring method as described in any of the above embodiments by running the computer program stored in the memory, so as to perform plasma etching on the tungsten-containing layer 10 in the wafer to form an opening with a target depth.

[0129] The semiconductor process equipment 200 according to the embodiments of the present application can be an inductively coupled plasma (ICP for short) etching equipment, or can also be a capacitively coupled plasma (CCP) etching equipment. The embodiments of the present application do not limit the type of the semiconductor process equipment 200.

[0130] In one embodiment, the process chamber is a physical vapor deposition process chamber, and the heating component includes an electrostatic chuck; the semiconductor process equipment further includes a power controller, and the controller is configured to determine a temperature control mode and transmit the power control parameters in the temperature control mode to the power controller; the power controller is configured to adjust the real-time temperature of the heater based on the power control parameters, and the temperature of the heater is conducted to the electrostatic chuck.

[0131] In this embodiment, the electrostatic chuck is used to conduct the heat provided by the heater to the wafer.

[0132] For the relevant specific limitations and beneficial effects of the chamber temperature rise monitoring method, reference can be made to the relevant descriptions in the above text, and the present specification will not elaborate here.

[0133] Exemplary Computer Program Product and Storage Medium

[0134] In addition to the above methods and devices, the chamber temperature rise monitoring method provided by the embodiments of this specification may also be a computer program product, which includes computer program instructions. When the computer program instructions are run by a processor, the processor is caused to execute the chamber temperature rise monitoring method according to various embodiments of this specification described in the "Exemplary Method" section above of this specification.

[0135] The computer program product can be written in any combination of one or more programming languages to write program code for performing the operations of the embodiments of this specification. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, executed as an independent software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0136] In addition, the embodiments of this specification also provide a computer-readable storage medium, on which a computer program is stored. The computer program is executed by a processor to perform the chamber temperature rise monitoring method according to various embodiments of this specification described in the "Exemplary Method" section above of this specification.

[0137] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it may include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the various embodiments provided by this specification may include non-volatile and / or volatile memories. Non-volatile memories may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0138] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0139] The above-described embodiments merely represent several implementation manners of this specification. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the solutions provided by the embodiments of this specification. It should be noted that for those of ordinary skill in the art, without departing from the concept of this specification, several deformations and improvements can still be made, and these all belong to the protection scope of this specification. Therefore, the protection scope of the patent of this specification should be subject to the appended claims.

Claims

1. A chamber heating monitoring method, characterized in that, Applied to a process chamber, the process chamber including a heating component; The method includes: Obtaining real-time data of relevant variables affecting the temperature rise process of the process chamber, and obtaining the actual real-time temperature value of the heating component in the process chamber. Among them, the relevant variables include continuous variables and digital variables, and the digital variables include the water flow state and temperature control mode of the process chamber; Inputting the real-time data of the relevant variables into a pre-trained temperature prediction model, and outputting the real-time temperature prediction value of the heating component through the temperature prediction model. Among them, the temperature prediction model is trained based on the historical actual temperature value of the heating component and the historical data of the relevant variables; By comparing the actual real-time temperature value of the heating component with the real-time temperature prediction value of the heating component, obtaining the real-time monitoring result of the temperature rise process of the process chamber.

2. The chamber heating monitoring method according to claim 1, characterized in that, After obtaining the real-time monitoring result of the temperature rise process of the process chamber by comparing the actual real-time temperature value of the heating component with the real-time temperature prediction value of the heating component, it further includes: If the real-time monitoring result indicates that the temperature rise of the process chamber is normal, updating the real-time data of the relevant variables to the historical data of the relevant variables, and updating the actual real-time temperature value of the heating component to the historical actual temperature value of the heating component; Among them, the updated historical actual temperature value of the heating component and the updated historical data of the relevant variables are used to retrain the temperature prediction model.

3. The chamber heating monitoring method according to claim 1, characterized in that, Before training the temperature prediction model, dividing the historical data of the relevant variables and the historical actual temperature value into a training set and a validation set according to a preset ratio; The training process of the temperature prediction model is as follows: Inputting the historical data of the relevant variables in the training set into a preset original prediction model, and outputting the training temperature prediction value of the heating component through the original prediction model. Among them, the original prediction model is pre-configured based on a long short-term memory neural network; Comparing whether the training temperature prediction value of the heating component is less than a training threshold with the historical actual temperature value of the heating component; If so, after testing and optimizing the parameters in the original prediction model through the historical data of the relevant variables in the validation set, determining the original prediction model as the temperature prediction model; If not, adjusting the internal parameters of the original prediction model and retraining the original prediction model until the temperature prediction model is obtained.

4. The chamber heating monitoring method according to claim 1, characterized in that, The temperature prediction model includes an input layer, a first operation layer, a second operation layer, a fully connected layer, a random dropout layer, and an output layer; The input layer, the first operation layer, the second operation layer, the fully connected layer, the random dropout layer, and the output layer are connected in sequence.

5. The chamber heating monitoring method according to claim 4, characterized in that, The step of inputting the real-time data of the relevant variables into a pre-trained temperature prediction model and outputting the real-time temperature prediction value of the heating component through the temperature prediction model includes: Inputting the real-time data of the relevant variables into the input layer; Inputting the intermediate data output by the input layer into the first operation layer, and the first operation layer performs the first data processing transformation; The intermediate data output by the first operation layer is input to the second operation layer, and the second operation layer performs a second data processing transformation; The intermediate data output by the second operation layer is input to the fully connected layer, and the fully connected layer performs a regularization process; The intermediate data output by the fully connected layer is input to the random dropout layer, and neurons are randomly lost according to a preset ratio during the operation of the random dropout layer; The intermediate data output by the random dropout layer is input to the output layer, and the output layer performs a regularization process and outputs the real-time temperature prediction value of the heating component.

6. The chamber heating monitoring method according to claim 1, characterized in that, Obtaining real-time data of relevant variables during the temperature rise process of the process chamber includes: Collecting real-time data of relevant variables during the temperature rise process of the process chamber; Performing standardization processing on the real-time data of the continuous variables and performing one-hot encoding on the real-time data of the digital variables to obtain standardized data of the relevant variables; Inputting the real-time data into a pre-trained temperature prediction model, and outputting the real-time temperature prediction value of the heating component through the temperature prediction model, including: Inputting the standardized data of the relevant variables into the pre-trained temperature prediction model, and outputting the real-time temperature prediction value of the heating component through the temperature prediction model.

7. The chamber heating monitoring method according to claim 1, characterized in that, Obtaining the real-time monitoring result of the temperature rise process of the process chamber by comparing the actual value of the real-time temperature of the heating component with the predicted value of the real-time temperature of the heating component, including: Calculating the difference between the actual value of the real-time temperature of the heating component and the predicted value of the real-time temperature of the heating component; If the difference is less than or equal to a preset temperature threshold, generating a real-time monitoring result that the temperature rise of the process chamber is normal; If the difference is greater than the temperature threshold, generating a real-time monitoring result that the temperature rise of the process chamber is abnormal.

8. The chamber temperature rise monitoring method according to claim 1, wherein The historical data of the relevant variables is historical time-series data of the relevant variables, and the historical time-series data includes the historical data of the relevant variables corresponding to multiple consecutive historical moments during the historical normal temperature rise process of the process chamber.

9. The chamber temperature rise monitoring method according to any one of claims 1-8, wherein The continuous variables include one or more of the inner heating wire power, the outer heating wire power, the inner heating wire current, and the outer heating wire current.

10. A semiconductor process equipment, wherein Including: A process chamber, the process chamber including a heating component; An upper electrode assembly located above the process chamber; A lower electrode assembly located inside the process chamber for carrying a wafer and applying a bias voltage to the wafer; A controller including at least one memory and at least one processor, the memory being used to store a computer program; The processor is configured to implement the chamber temperature rise monitoring method according to any one of claims 1 to 9 by running the computer program stored in the memory.

11. The semiconductor process equipment according to claim 10, wherein The process chamber is a physical vapor deposition process chamber, and the heating component includes an electrostatic chuck; The semiconductor process equipment further includes a power controller, the controller is configured to determine a temperature control mode and transmit the power control parameters in the temperature control mode to the power controller; the power controller is configured to adjust the real-time temperature of the heater based on the power control parameters, and the temperature of the heater is conducted to the electrostatic chuck.

Images