State monitoring method, method for manufacturing semiconductor device, program, and substrate processing device
The method of comparing sensor data with reference values in the predictive detection system addresses the challenge of detecting abnormalities in substrate processing apparatus components, enabling timely maintenance and ensuring apparatus reliability.
Patent Information
- Application Number
- PCT/JP2025/037441
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2025-10-24
- Publication Date
- 2026-11-12
AI Technical Summary
Existing substrate processing apparatuses face challenges in determining the optimal timing for replacing deteriorating components due to the lack of effective methods for detecting abnormalities in members before failure occurs.
A method involving the comparison of measured peak frequencies from sensors with reference values to predict abnormalities in exhaust devices by monitoring variations in frequency, utilizing a predictive detection system to identify potential failures and facilitate timely maintenance.
Enables proactive maintenance by accurately predicting component failures in exhaust systems, ensuring continuous and reliable operation of the substrate processing apparatus.
Smart Images

Figure JP2025037441_12112026_PF_FP_ABST
Abstract
Description
State monitoring method, semiconductor device manufacturing method, program, and substrate processing apparatus
[0001] The present disclosure relates to a state monitoring method, a semiconductor device manufacturing method, a program, and a substrate processing apparatus.
[0002] A substrate processing apparatus is composed of various members such as a processing mechanism for processing a substrate and a transport mechanism for transporting the substrate. Some members gradually deteriorate with use and fail, making it difficult to determine the timing for replacement with new members. For example, attempts to detect abnormalities in members as shown in Patent Document 1 to Patent Document 3 have been proposed.
[0003] International Publication No. 2020-188747, International Publication No. 2021-059333, International Publication No. 2022-064814
[0004] The present disclosure provides a technique that monitors the state of a target member and enables maintenance of the member before failure occurs.
[0005] According to one aspect of the present disclosure, a step of comparing an actually measured value of a peak frequency of a primary frequency of the exhaust device calculated based on data acquired from a sensor provided in the exhaust device with a reference value of the peak frequency of the primary frequency of the exhaust device, and confirming a variation in a frequency corresponding to a preset rated rotation speed of the exhaust device in a specific period, and a first step of identifying an abnormality predicted based on the variation in the frequency are provided.
[0006] According to the present disclosure, it is possible to monitor the state of a target member and perform maintenance on the member before a failure occurs.
[0007] This is a diagram showing an example configuration of a processing furnace for a substrate processing apparatus preferably used in one aspect of this disclosure. This is a block diagram showing an example configuration of a control module for a substrate processing apparatus preferably used in one aspect of this disclosure. This is a diagram showing an example configuration of an anomaly prediction detection system including a substrate processing apparatus preferably used in one aspect of this disclosure. This is a diagram showing an example configuration of an exhaust device preferably used in one aspect of this disclosure. This is a diagram illustrating a sequence for detecting an anomaly prediction performed in an anomaly prediction detection system preferably used in one aspect of this disclosure. This is an example of a diagram showing processed data detected by an acceleration sensor preferably used in one aspect of this disclosure. This is a diagram showing an example of a list used to detect an anomaly prediction performed in an anomaly prediction detection system preferably used in one aspect of this disclosure. This is a diagram showing an example of processed data detected by an acceleration sensor preferably used in one aspect of this disclosure. This is a diagram showing an example of exhaust device-related data output from an exhaust device preferably used in one aspect of this disclosure. This is a diagram showing an example of a list used to detect an anomaly prediction performed in an anomaly prediction detection system preferably used in one aspect of this disclosure. This is a diagram showing an example of exhaust device-related data output from an exhaust device preferably used in one aspect of this disclosure.
[0008] An aspect of this disclosure will be described below with reference to the drawings. Note that the drawings used in the following description are schematic, and the dimensional relationships and proportions of the elements shown in the drawings do not necessarily correspond to reality. Furthermore, the dimensional relationships and proportions of the elements do not necessarily correspond between multiple drawings.
[0009] (Processing Furnace) Figure 1 illustrates a processing furnace 42 of a substrate processing apparatus preferably used in this embodiment. In this embodiment, the substrate processing apparatus is configured as a vertical substrate processing apparatus (hereinafter referred to as the processing apparatus) that performs a substrate processing process such as heat treatment as one step in the manufacturing process of a semiconductor device (device) manufacturing method.
[0010] The processing furnace 42 is equipped with a reaction tube 41. The reaction tube 41 is made of a heat-resistant nonmetallic material such as quartz (SiO2) or silicon carbide (SiC), and has a cylindrical shape with a closed upper end and an open lower end.
[0011] A processing chamber 34 is formed inside the cylinder of the reaction tube 41. A boat 30, which serves as a substrate holder, is inserted into the processing chamber 34 from below, and the boat 30 is configured to house wafers 14 that are held in a horizontal position and arranged in multiple stages in the vertical direction. The boat 30 housed in the processing chamber 34 is configured to be rotatable while maintaining the airtightness of the processing chamber 34, by rotating its rotation axis 44 using a rotation mechanism 43, while carrying multiple wafers 14 as substrates.
[0012] Below the reaction tube 41, a manifold 45 is arranged concentrically with the reaction tube 41. The manifold 45 is made of a metal material such as stainless steel and has a cylindrical shape with an open upper and lower end. The reaction tube 41 is supported vertically from the lower end by this manifold 45. In other words, the reaction tube 41 that forms the processing chamber 34 is supported vertically via the manifold 45 to form the processing furnace 42. The lower end of the manifold 45 is configured to be airtightly sealed by a seal cap 46 when a boat elevator (not shown) rises. Between the lower end of the manifold 45 and the seal cap 46, a sealing member 46a such as an O-ring is provided to airtightly seal the processing chamber 34.
[0013] Furthermore, the manifold 45 is connected to a gas inlet pipe 47 for introducing raw material gas and purge gas into the processing chamber 34, and an exhaust pipe 48 for exhausting the gas from the processing chamber 34. The exhaust pipe 48, which serves as the exhaust line, is equipped with, in order from the upstream side, a pressure sensor 37 for detecting the pressure in the processing chamber 34, an APC valve 35 as an adjustment valve for adjusting the pressure in the processing chamber 34, a pump 36 as an exhaust device for exhausting the atmosphere of the processing chamber 34, and a pressure sensor 38 for detecting the pressure on the secondary side of the pump. The pressure sensor 37, adjustment valve 35, and pressure sensor 38 are configured as an exhaust system, and the pump 36 may also be included in the exhaust system. With this exhaust system configuration, the pressure in the processing chamber 34 can be adjusted by adjusting the valve opening of the adjustment valve 35 based on the pressure information detected by the pressure sensor 37 while the pump 36 is operating.
[0014] A heater unit 49, which serves as a heating mechanism, is arranged concentrically with the reaction tube 41 on its outer circumference. The heater unit 49 is configured to heat the processing chamber 34 so that the processing chamber 34 has a uniform or predetermined temperature distribution throughout.
[0015] Figure 2 shows an example configuration of a control system for a substrate processing apparatus, including an input / output device 31 and a control unit 32 as operating units. The controller 260, which serves as the main control unit, is configured to include at least the input / output device 31 and the control unit 32.
[0016] The control unit 32 is configured as a computer equipped with a CPU (Central Processing Unit) 260a, RAM (Random Access Memory) 260b, storage device 260c, and I / O port 260d. The RAM 260b, storage device 260c, and I / O port 260d are configured to exchange data with the CPU 260a via an internal bus 260e.
[0017] The CPU 260a, acting as the arithmetic unit, is configured to read and execute control programs from the storage device 260c, and to read files from the storage device 260c in response to input operation commands from the input / output device 31. It is also configured to compare and calculate calculation data with setting values input from the receiving unit 285 and files and control data stored in the storage device 260c. The RAM 260b is configured as a memory area (work area) where programs, calculation data, processing data, etc., read by the CPU 260a are temporarily held.
[0018] The storage device 260c is composed of, for example, flash memory, HDD (Hard Disk Drive), etc. The storage device 260c contains various files including control programs that control the operation of the substrate processing device, screen files such as editing screens for recipe editing and setting screens for setup, a list of predicted abnormalities for the substrate processing device and the components that make up the substrate processing device, a list of error handling settings according to the identified abnormalities, process recipes that describe the procedures and conditions for substrate processing, recipes for transporting the substrate, and data detected from each sensor. It also contains readable calculation data and processing data generated during the process of setting the process recipe used for processing the wafer 14.
[0019] The operation screen of the input / output device 31 displays, for example, an editing screen for creating a data acquisition sequence under the same conditions as those for processing the wafer 14, and a selection screen for selecting the target component. Furthermore, the input / output device 31, which may be configured as a touch panel or the like, is configured to allow connection to an external storage device 262. In addition, the controller 260 is configured to allow connection to the network 263 via a receiving unit 285. This means that the controller 260 can also connect to higher-level devices such as a host computer located on the network 263. Therefore, if the input / output device 31 is located on the network 263, it can connect to the controller 260. In other words, the input / output device 31 is not limited to the above embodiment and may be located in a location separate from the substrate processing apparatus.
[0020] Hereinafter, the files including this process recipe and control program may be collectively referred to simply as a program (program product). In this specification, the term "program" may refer to a process recipe alone, a control program alone, or both.
[0021] The controller 260 is not limited to being configured as a dedicated computer; it may also be configured as a general-purpose computer. For example, the controller 260 according to this embodiment can be configured by preparing an external storage device 262 containing the above-mentioned program (e.g., magnetic tape, magnetic disks such as flexible disks or hard disks, optical disks such as CDs or DVDs, semiconductor memory such as USB memory or memory cards), and installing the program on a general-purpose computer using such an external storage device 262.
[0022] The means of supplying a program to a computer are not limited to supplying it via an external storage device 262. For example, a communication means such as a network 263 (the internet or a dedicated line) may be used to supply the program without going through the external storage device 262.
[0023] Furthermore, the storage device 260c and the external storage device 262 are configured as computer-readable recording media. Hereinafter, these will be collectively referred to simply as recording media. In this specification, when the term recording media is used, it may include only the storage device 260c, only the external storage device 262, or both.
[0024] <Control System in This Embodiment> Next, a control system for detecting abnormal signs in each component of the substrate processing apparatus will be described with reference to Figure 3.
[0025] As shown in Figure 3, the control system comprises a board processing device controller 260, a predictive detection controller 82 as a predictive detection unit, various sensors 124, a data acquisition unit (Data Collection Unit, hereinafter abbreviated as DCU) 126, and an edge controller (Edge Controller, hereinafter abbreviated as EC) 128 as a data processing unit. These components constituting the control system are connected by wired or wireless connections. Although described separately in this embodiment, it goes without saying that the functions of the predictive detection controller 82 may also be integrated into the board processing device controller 260.
[0026] The predictive detection controller 82 includes a data acquisition unit that acquires sensor data from sensors of various components constituting the substrate processing apparatus, and a monitoring unit that monitors the status of the substrate processing apparatus and components. The predictive detection controller 82 is configured to detect signs of abnormalities (i.e., predicted abnormalities) that may occur in the target component by comparing data related to the target component from the data of various sensors 124 with pre-acquired reference data (reference data) corresponding to the data related to the component, thereby confirming the status of the component. In this embodiment, data related to the pump 36, which is an exhaust device and one of the components, is acquired, and signs of abnormalities (i.e., predicted abnormalities) are detected. The predictive detection controller 82 also incorporates a predictive detection program that predicts the occurrence of abnormal signs, which will be described later, based on the movement of sensor data.
[0027] Furthermore, the predictive detection controller 82 has two systems: one that is directly connected to the board processing device controller 260, and another that is connected to the board processing device controller 260 via the DCU 126. Therefore, when the predictive detection controller 82 detects an abnormality, it sends a signal directly to the board processing device controller 260 without going through the DCU 126 to generate an alarm, and also displays the sensor data information from the sensor installed on the component where the abnormality was detected on the screen of the input / output device 31.
[0028] The various sensors 124 are sensors (for example, pressure sensors and temperature sensors) provided on various components constituting the substrate processing apparatus, and detect data such as flow rate, concentration, temperature, humidity (dew point), pressure, current, voltage, torque, vibration, position, and rotational speed of each component. In this embodiment, in addition to the pressure sensors 37 and 38, the sensors 124 also include components that measure temperature and current provided on an exhaust device 36 (not shown).
[0029] The DCU 126 collects and stores data from various sensors 124 during the execution of the process recipe. The EC 128 also takes in the sensor data as needed depending on the type of sensor, applies processing such as a Fast Fourier Transform (FFT) to the raw data, and then transmits it to the predictive detection controller 82.
[0030] Furthermore, the various sensors 124 are divided into a first sensor system 124A and a second sensor system 124B, which have different transmission paths. The first sensor system 124A is a system that acquires data in real time in 0.1-second intervals, and data is transmitted in real time from the first sensor system 124A to the predictive detection controller 82 via the substrate processing device controller 260 and the DCU 126. This first sensor system 124A includes data detected by sensors such as temperature sensors, pressure sensors, and gas flow sensors. In this embodiment, at least data related to the exhaust system 36 is transmitted in real time to the predictive detection controller 82. The data related to the exhaust system 36 will be described later.
[0031] On the other hand, the second sensor system 124B is a system in which data is processed by EC128, such as by performing FFT, to extract only the parts necessary for analysis, and the processed data is transmitted in a modified file format. The processed data is transmitted from the second sensor system 124B to the predictive detection controller 82 via EC128. This second sensor system 124B includes data detected by sensors such as acceleration sensors. In this embodiment, the measured value of the peak frequency of the primary frequency of the exhaust system 36, calculated based on data acquired from at least the acceleration sensor provided in the exhaust system 36, is obtained. Specifically, data (which constitutes) a graph with acceleration on the vertical axis and frequency on the horizontal axis, as described later, is transmitted to the predictive detection controller 82 via EC128.
[0032] If the sensor is an acceleration sensor, vibration data is accumulated in milliseconds, resulting in a massive amount of data. If this data is sent directly to the predictive detection controller 82, it will consume a large amount of memory capacity in the predictive detection controller 82. Since this vibration sensor data is ultimately processed using methods such as FFT for analysis, performing this processing in advance with the EC128 reduces the amount of information and allows it to be sent to the predictive detection controller 82 in a format that is easy to analyze.
[0033] Next, the exhaust device 36, on which the acceleration sensor 101 is located, will be explained using Figure 4.
[0034] The exhaust device 36 is, for example, a rotary pump, with the rotor located within the pump module 102. A flange 36a, located in the negative direction of the Y-axis from the pump module 102, is connected to an exhaust pipe 48. The upper pump module 102a and the lower pump module 102b house a drive system for driving the rotor. The exhaust device 36 may also include a housing that is divided into an upper housing and a lower housing (divided configuration).
[0035] Four vibration-damping rubbers 103 are provided between the pump module 102 and the upper housing as vibration-damping members to reduce the effects of vibration caused by the operation of the rotor within the pump module 102. The vibration-damping rubbers 103 are provided in the Y-axis direction to suppress vibrations in the Y-axis direction, which has the greatest acceleration among the X, Y, and Z axes. The rubbers are also provided between the upper housing and the pump module 102 because this is the direction in which the center of gravity is applied (Z-axis direction), and therefore is considered to be more susceptible to vibrations.
[0036] The acceleration sensor 101 is positioned on top of the upper pump module 102a. Specifically, as can be easily inferred from the arrangement of the vibration-damping member 103 located in the exhaust device 36 shown in Figure 4, the acceleration sensor 101 is installed on the pump module 102a in a location where the influence of the vibration-damping member 103 is minimal. This allows for accurate acquisition of vibration data detected by the acceleration sensor 101 shown in Figure 4. The acceleration sensor 101 can acquire vibration data in three axes (X-axis, Y-axis, Z-axis), but vibration data in a single axis direction is also acceptable.
[0037] On the other hand, while a flexible bellows pipe is used to connect the flange 36a to the exhaust pipe 48 to suppress vibrations, there are concerns about the effects of vibration. The equipment has been designed to prevent vibrations from affecting the boat 30 (or wafer 14) located in the processing chamber 34 at the rated rotational speed frequency range, and the boat 30 itself has been designed with its natural frequency in mind so as not to resonate with the rated frequency of the exhaust device 36. However, if there is a frequency shift, there is a concern that the above measures will be ineffective or that a new resonance point will emerge on the boat 30.
[0038] As mentioned above, since the effects of vibration are a concern, a system was developed to monitor the peak frequency of the primary frequency, which has the greatest vibration power, in order to detect any signs of abnormality in the exhaust system 36.
[0039] First, the operation procedure for processing a wafer 14 as one step in semiconductor device manufacturing using the substrate processing apparatus according to this embodiment will be explained with reference to Figures 1 and 2. Here, the timing for acquiring reference data (hereinafter sometimes referred to as normal data) for vibration data detected by the acceleration sensor 101 will be explained. Subsequently, the procedure for acquiring data related to the exhaust device 36 will be explained.
[0040] (Wafer charging process) A substrate transfer machine (not shown) performs a wafer charging operation in the transfer chamber, loading unprocessed wafers 14 into the boat 30 before it is transported to the processing chamber 34. As a result, the boat 30 holds multiple wafers 14 in a stacked state with spacing between them in the vertical direction.
[0041] (Loading Process) After the wafer charging operation, the boat 30 holding multiple unprocessed wafers 14 is loaded into the processing room 34 by the raising and lowering operation of a boat elevator (not shown). In other words, the boat elevator is operated to load the boat 30 holding the unprocessed wafers 14 from the transfer room to the processing room 34. As a result, the seal cap 46 seals the lower end of the manifold 45 via the sealing member 46a.
[0042] During the wafer charging and loading processes, the APC valve 35 is in a fully closed state, the status of the exhaust device 36 is "standby," and normal data related to the exhaust device 36 is acquired during these processes.
[0043] (Film deposition preparation process) After boat loading, specifically, exhaust is performed using the exhaust pipe 48 to bring the processing chamber 34 to the desired pressure (vacuum level) (depressurization process). Then, the processing chamber 34 is heated using the heater unit 49, and the rotation mechanism 43 is operated to rotate the boat 30, and the wafer 14 rotates accordingly. This sets the processing chamber 34 to the desired temperature (processing temperature).
[0044] (Processing Step) When the film formation preparation is completed, a predetermined process (film formation process) is performed on the unprocessed wafers 14 held by the boat 30 carried into the processing chamber 34. Further, source gas, purge gas and the like are supplied into the processing chamber 34 through the gas introduction pipe 47. Thereby, a thin film is formed on the surface of the unprocessed wafer 14 held by the boat 30 by utilizing, for example, a thermal decomposition reaction, a chemical reaction, or the like.
[0045] In this processing step, the APC valve 35 is in a pressure control state, the status of the exhaust device 36 is "film formation in progress", and normal data related to the exhaust device 36 is acquired.
[0046] After the film formation on the surface of the wafer 14 is completed, heating by the heater unit 49 is stopped to lower the temperature of the processed wafer 14 to a predetermined temperature. Then, when a preset time has elapsed, the gas supply to the processing chamber 34 is stopped, and the supply of inert gas to the processing chamber 34 is started. As a result, the inside of the processing chamber 34 is replaced with the inert gas, and the pressure in the processing chamber 34 is returned to normal pressure. Further, during or after the return of the processing chamber 34 to normal pressure, the rotation of the boat 30 by the rotation mechanism 43 is also stopped.
[0047] (Unloading Step) Thereafter, the seal cap 46 is lowered to open the lower end of the manifold 45, and the boat 30 holding the processed wafers 14 is unloaded (boat unloading) from the lower end of the manifold 45 to the outside of the processing chamber 34.
[0048] (Wafer Discharge Step) After the wafers 14 on the waiting boat 30 have cooled to a predetermined temperature (for example, about room temperature), a substrate transfer device disposed in the transfer chamber unloads the wafers 14 from the boat 30. Then, a wafer discharge operation is performed, in which the processed wafers 14 unloaded from the boat 30 are transferred and stored in an empty pod placed on a pod opener. In this way, a series of processing operations of the substrate processing steps by the substrate processing apparatus is completed.
[0049] During the wafer discharge and unloading processes, the APC valve 35 is in a fully closed state, the status of the exhaust device 36 is "standby," and normal data related to the exhaust device 36 is acquired during these processes.
[0050] Thus, the vibration data detected by the acceleration sensor 101 is not acquired for all stages of the substrate processing process. As described above, if vibration data is accumulated in milliseconds, the amount of data becomes enormous. Therefore, data is acquired only for specific periods (events), and normal data is acquired for each event. If the amount of data is not a concern, it is preferable to acquire data for all events (all stages).
[0051] (Acquisition of Normal Data) First, the data related to the exhaust device 36 will be described. In this specification, the data related to the exhaust device 36 includes not only vibration data detected by the acceleration sensor 101, but also the ultimate pressure of the processing chamber 34 detected by the pressure sensor 37, the secondary pressure of the exhaust device 36 detected by the pressure sensor 38, the rotational speed of the exhaust device 36, the current value, and the temperature of the casing (pump module as a container enclosing the exhaust device 36) 102, as detected by sensors provided in the exhaust device 36 (not shown).
[0052] In other words, unlike vibration data, there is no data volume issue with data related to the exhaust device 36, excluding vibration data detected by the acceleration sensor 101, even if the sensor is turned on at each event in the substrate processing process. However, the target pressure of the processing chamber 34 detected by the pressure sensor 37 is obtained when the processing chamber 34 is in a reduced pressure state, for example, in the above substrate processing process, by detecting the pressure of the processing chamber 34 from the end of the (reduced pressure process) to the (processing process). It goes without saying that the pressure detected by the pressure sensor 37 may also be obtained by turning on the sensor at each event in the substrate processing process.
[0053] It is preferable to acquire normal data related to these exhaust devices 36 when the same processing as the substrate processing process described above is performed on the wafer 14 for the first time. For example, it is preferable to acquire normal data during the film deposition process in the setup operation of the device. Alternatively, normal data may be acquired in real time while processing the substrate in the first batch of continuous batch processing, and the previously acquired normal data may be updated. Alternatively, normal data may be updated by accumulating and learning new normal data without fixing the normal data. The normal data related to these exhaust devices 36 is stored on a recording medium, for example, in a storage device 260c.
[0054] (Anticipation) Figure 5 shows an example of a sequence executed by the predictive detection controller 82. It is executed in the substrate processing step of the substrate processing process, that is, the film deposition step.
[0055] When the predictive detection controller 82 starts executing this sequence, it acquires the necessary data and files (S101). Here, it acquires actual measurement data related to the exhaust system 36, reference data for the data related to the exhaust system 36 that is stored in the storage device 260c beforehand (hereinafter sometimes referred to as normal data), and files that will be used in the first and second steps, which will be described later.
[0056] Next, the measured data is compared with the normal data (S102). Here, the measured value of the peak frequency of the primary frequency of the exhaust system, which is calculated based on vibration data detected by the acceleration sensor 101 installed in the exhaust system 36, is compared with the reference data for the peak frequency of the primary frequency of the exhaust system 36. If the comparison shows no shift to a lower frequency in the peak frequency of the primary frequency, this predictive detection sequence is terminated.
[0057] Here, the presence or absence of a shift in the peak frequency of the primary frequency towards the lower frequency side is checked, and the process moves to the next step (first step). Alternatively, the presence or absence of a shift in the peak frequency of the primary frequency towards the lower frequency side may be determined by a threshold (first threshold). However, in this case, even if the value is above the threshold, it is only an alert (warning) level. This is because even if there is a slight fluctuation in frequency, it cannot be said that the exhaust capacity of the exhaust device 26 has decreased.
[0058] For example, if the exhaust capacity decreases due to a reduction in the rated rotational speed of the exhaust device 36, the ultimate pressure may increase. Also, if the exhaust capacity decreases, it may take longer to stabilize at the predetermined processing pressure. On the other hand, this does not prevent the wafer 14 from being processed, so it is at an alert level.
[0059] As will be described later, many abnormalities are caused by a decrease in the rated rotational speed. However, a decrease in the rated rotational speed alone is not sufficient to identify abnormalities that require maintenance (e.g., replacement of components). On the other hand, as mentioned above, fluctuations in the peak frequency of the primary frequency can affect the wafer 14, so in this embodiment, fluctuations in the peak frequency of the primary frequency are used as the starting timing for abnormality prediction. In other words, according to this embodiment, it is possible to identify abnormalities that are expected to occur by using a shift of the peak frequency of the primary frequency to the lower frequency side (or a decrease in rotational speed) as a trigger.
[0060] In other words, according to this embodiment, since many abnormalities are caused by a decrease in the rated rotational speed of the exhaust device 36, it becomes possible to identify the predicted abnormality by triggering a shift in the peak frequency of the primary frequency to the lower frequency side (or a decrease in rotational speed) and perform maintenance before replacing the components.
[0061] In the example relationship between acceleration and frequency shown in Figure 6, the peak frequency is 117 Hz at 7000 rpm. For example, if the pump rotation speed decreases by 5%, the peak frequency of the primary frequency shifts by approximately 6 Hz in the direction of the arrow. The amplitude of the peak frequency (the width of the dotted line shown in Figure 6) is ±10 Hz around the peak frequency (the frequency at which the acceleration is at its maximum). Note that Figure 6 graphs the Y-axis, where the acceleration is greatest at the peak frequency. Although the focus is on the peak frequency of the primary frequency, this does not negate the existence of higher-order frequencies other than the primary frequency.
[0062] If a shift in the peak frequency of the primary frequency of the exhaust system 36 is confirmed in S102, the process proceeds to the first step (S103). In the first step, abnormalities related to the exhaust system 36 are predicted based on the first table, which shows the relationship between abnormality items indicating predicted abnormalities shown in Figure 7 and the data necessary to identify the factors causing those items, and data related to the exhaust system 36. In this first step, if there is a shift in the peak frequency of the primary frequency to the lower frequency side (or a decrease in rotational speed), the abnormalities shown in Figure 7 that may occur in the future are identified. Specifically, predicted abnormalities are identified using a list that summarizes the abnormalities related to the exhaust system 36 shown in Figure 7 as the first table and the data related to the exhaust system 36 necessary to identify those abnormalities. Figure 7 is a list of abnormality items related to the exhaust system 36 and the detection item data that identifies those abnormalities.
[0063] The data related to the exhaust system 36 (detection item data in Figure 7) necessary to identify abnormalities related to the exhaust system 36 will be explained. The detection item data includes data related to the exhaust system 36, such as the rotational speed of the exhaust system 36, the current value of the exhaust system 36, the casing temperature of the exhaust system 36 (casing temperature), the ultimate pressure value detected by the pressure sensor 37, the pressure value of the pressure sensor 38 (pump secondary side pressure), and acceleration based on vibration data detected by the acceleration sensor 101. Specifically, for the acceleration shown in Figure 7, the frequency and acceleration are calculated for the X, Y, and Z axes, respectively, as shown in Figure 8.
[0064] The detection item data described above is acquired in S101. When there is a change in the peak frequency of the primary frequency of the exhaust device 36, that is, when there is a change in rotational speed, the abnormal items shown in Figure 7 are checked, and which abnormal items are potentially likely to occur in the future are identified. Note that the abnormal items may be identified sequentially from No. 1, or all at once.
[0065] Anomaly item A is a sudden gas load. For example, as shown in Figure 9, if the rotational speed remains almost constant at the rated speed (e.g., 7000 rpm) and there is a sudden decrease in rotational speed, the measured value of the secondary side pressure is compared with the normal secondary side pressure data. If a difference is found as a result of the comparison, anomaly item A is identified. In other words, a sudden gas load is identified as a candidate for a predicted anomaly. Alternatively, a threshold can be used to determine whether or not there is a difference.
[0066] Anomaly item B is a decrease in the temperature of the casing surrounding the rotor. Specifically, the temperature detected by a temperature sensor (not shown) is compared with normal temperature data. If a difference is found as a result of the comparison, anomaly item B is identified. Alternatively, a threshold can be used to determine whether or not there is a difference. In this way, a gas load due to a decrease in casing temperature is identified as a candidate for a predicted anomaly.
[0067] Anomaly item C is a minute leak in the exhaust piping of the exhaust device 36, and looseness of the O-ring fastenings at the pump inlet (outlet) and the pump. For anomaly item C, the difference between the normal data and the target pressure and the secondary pressure of the pump is compared. If a difference is found as a result of the comparison, anomaly item C is identified. Alternatively, a threshold value may be used to determine whether or not there is a difference. When a minute leak is introduced, the load increases and the rotational speed decreases. Therefore, a gas load due to the introduction of minute leak gas is identified as a predicted candidate for anomaly.
[0068] Anomaly item D is an anomaly caused by clogging of by-products. This includes not only overload due to by-product accumulation, but also insufficient cleaning and inadequate deactivation treatment. For anomaly item D, the difference between the ultimate pressure, the pump's secondary pressure, and the casing temperature is compared with normal data. If a difference is found in the comparison of each data, anomaly item D is identified. Alternatively, a threshold can be used to determine whether or not a difference exists. When by-products clog, the conductance deteriorates, resulting in a decrease in ultimate pressure. Also, when by-products accumulate, the load increases and the rotational speed decreases. Therefore, a gas load caused by clogging of by-products is identified as a predicted anomaly candidate.
[0069] Anomaly item E is an anomaly caused by deterioration of the vibration-damping rubber or loosening of the fastening parts. For anomaly item E, the difference in acceleration of the exhaust device 36 with normal data is compared for each axis. If a difference is found as a result of comparing the data for each axis, anomaly item D is identified. Alternatively, a threshold can be used to determine whether or not there is a difference. Because the exhaust device 36 rotates at an angle, the rotation speed of the Y axis decreases. However, an increasing acceleration should be observed for the other axes. Therefore, a decrease in rotation speed due to deterioration of the vibration-damping rubber is identified as a predicted candidate for anomaly.
[0070] Anomaly item F is an anomaly caused by the current value supplied to the drive unit that attempts to rotate the rotor. As the rotational speed of the exhaust device 30 decreases, the current value increases in an attempt to rotate the motor. For anomaly item F, the difference between the measured value of the exhaust device 36 current value and the normal data is compared. If a difference is found as a result of the comparison, anomaly item D is identified. If the current value becomes too high, it can lead to motor damage or component failure, so if there is a fluctuation in the current value, it is identified as an anomaly indicating motor overload as a predicted anomaly candidate.
[0071] Once the first step (S103) is completed, the process moves to the second step (S104). In the second step, the abnormal items identified as predicted abnormality candidates in the first step are judged to be abnormal and a decision is made on what action to take. Here, the abnormality is compared with a threshold, but the threshold here does not indicate a failure, but rather whether the part is in a state where it should be replaced before it fails. For example, it is preferable that the threshold is not simply a recommended value from the part manufacturer, but is calculated or set from past performance data for the same part.
[0072] Figure 8 is the second table, which is a list summarizing the judgment criteria and processing (error handling) for each abnormality item. In the second step, which determines the error handling according to the abnormality identified in the first step, the first processing (continuous monitoring), the second processing (stopping the next batch), or the third processing (immediate stop) is determined based on the second table which identifies the error handling. Here, the first processing (continuous monitoring) is configured to notify the input / output device 32 of an alert (warning), and the second processing (stopping the next batch) and the third processing (immediate stop) are configured to notify the input / output device of an alarm (abnormality). In the second processing (stopping the next batch), an alarm (abnormality) is notified to prompt replacement before proceeding to the next batch. This makes it possible to detect signs of abnormality. Each abnormality item will be explained below.
[0073] If abnormality item A is identified in the first process, it is determined whether the rotation speed continues to decrease beyond a predetermined threshold (duration). If the duration exceeds the threshold, the second process (stopping the next batch) is decided; if the duration is less than the threshold, the first process (continued monitoring) is decided.
[0074] When abnormal item B is identified in the first step, it is determined whether it is lower than a pre-set threshold (casing temperature). Specifically, if the difference between the measured value and the normal data is lower than the threshold, the second process (stopping the next batch) is decided; if it is higher than the threshold, the first process (continuing monitoring) is decided.
[0075] When abnormality item C is identified in the first step, it is determined whether the rotational speed continues to decrease beyond a preset threshold (duration) and whether the secondary pressure of the exhaust device 36 is higher than the threshold. Specifically, if both the duration of the rotational speed decrease and the secondary pressure exceed the threshold, the second process (stopping the next batch) is decided; otherwise, the first process (continued monitoring) is decided.
[0076] When abnormality item D is identified in the first process, it is determined whether the rotational speed continues to decrease beyond a preset threshold (duration), and whether both the secondary pressure of the exhaust device 36 and the ultimate pressure of the processing chamber are higher than the threshold or lower than a preset threshold (casing temperature). Specifically, if all measured values exceed the threshold, the second process (stop the next batch) is decided; otherwise, the first process (continue monitoring) is decided.
[0077] When an abnormality item E is identified in the first step, it is determined whether the rotational speed is fluctuating above a preset threshold (acceleration). Specifically, as shown in Figure 8, it is determined for each axis whether the preset threshold for each axis has increased or decreased. If at least one axis is outside the threshold, the second process (stop the next batch) is decided; if all axes are within the threshold range, the first process (continue monitoring) is decided.
[0078] When abnormal item F is identified in the first step, it is determined whether the condition exceeding a preset threshold (e.g., 1% or less) continues for a preset threshold (duration). Specifically, as shown in Figure 11, it is determined whether the current value has continuously exceeded the threshold for 3 seconds. If the measured current value continues to exceed the current value (threshold) as a result of the determination, it is determined to be a motor failure and the third process (immediate stop) is performed. If this is not the case, the second process (stop the next batch) is performed. The threshold is set so that it is not an abnormality that immediately results in exhaust device failure, but for example, the current value of the exhaust device 36 is related to wire breakage or motor lifespan, and is therefore directly related to the failure of the motor as a component of the exhaust device 36, so the abnormality (error handling) is made strict for safety reasons.
[0079] In the second process, the corrective actions are prioritized in the following order: third process (immediate stop), second process (stop next batch), and first process (continue processing). Specifically, if at least one corrective action determined for each abnormal item identified in the first process is immediate stop, then immediate stop is performed. If all corrective actions for each abnormal item are first process (continue processing), then continue processing is performed. Otherwise, the second process (stop next batch) is determined. In other words, if immediate stop is not among the corrective actions, then the second process (stop next batch) is determined for each abnormal item.
[0080] Then, the predictive detection controller 82 transmits the process determined in the second step to the controller 260, and the sequence shown in Figure 5 ends. The controller 260 then executes the process determined in the second step. Specifically, in the third step, which executes the process determined in the second step as error handling, if the second step determines it to be the first step, the substrate processing is allowed to continue. If the second step determines it to be the second step, the first step is not allowed to be executed after the substrate processing in the processing chamber is completed.
[0081] Here, the threshold values in the first step and the threshold values in the second step for the pressure detected by the pressure sensors 37 and 38 may be the same.
[0082] Furthermore, in S102, measured values and normal data related to the exhaust system 36 other than the primary frequency may be acquired. For example, measured values and normal data of the data included in the verification item data shown in Figure 7 may be acquired. Alternatively, the first table shown in Figure 7 and the second table shown in Figure 10 may be combined into a single table. This allows the first step (S103) and the second step (S104) to be combined into a single step.
[0083] According to this disclosure, one or more of the following multiple effects (a) to (h) are achieved.
[0084] (a) By comparing the measured peak frequency of the primary frequency of the exhaust system, calculated based on data acquired from sensors installed in the exhaust system, with normal data, fluctuations in the frequency corresponding to the rotational speed of the exhaust system, which is set in advance, can be confirmed over a specific period, thereby enabling the detection of potential abnormalities related to the exhaust system that may occur in the future.
[0085] (b) By capturing fluctuations in the peak frequency of the primary frequency of the exhaust system, fluctuations in the rotational speed of the exhaust system can be captured. This allows for comprehensive detection of any abnormalities related to the exhaust system caused by rotational speed.
[0086] (c) By using the point in time when a fluctuation occurs in the peak frequency of the primary frequency of the exhaust system as a trigger to check for abnormalities related to the exhaust system, it is possible to predict abnormalities related to the exhaust system.
[0087] (d) In order to identify predicted abnormalities, the measured values of data related to the exhaust system are compared with pre-acquired normal data (reference values), so that all predicted abnormalities can be identified without fail.
[0088] (e) In order to identify predicted abnormalities, the system compares the difference between the measured values of data related to the exhaust system and pre-acquired normal data (reference values) with a threshold, thereby enabling accurate identification of predicted abnormalities.
[0089] (f) By using a list that compiles abnormal items related to the exhaust system and the detection item data necessary to identify those abnormalities, it is possible to accurately identify all predicted abnormalities without omission.
[0090] (g) By using a list that summarizes abnormal items related to the exhaust system, the criteria for determining an abnormality, and the abnormality handling procedures, it is possible to prohibit the execution of the next process after the circuit board processing, thereby encouraging replacement of the exhaust system before it fails.
[0091] (h) The current value of the exhaust system has been added as an abnormality item related to the exhaust system. This includes abnormality items related to wire breaks and component lifespan, so it is possible to encourage replacement before abnormalities unrelated to component failure occur.
[0092] (Other Embodiments) Although embodiments of the present disclosure have been described in detail above, the present disclosure is not limited to the embodiments described above and can be modified in various ways without departing from its essence.
[0093] Furthermore, the processing performed by the substrate processing apparatus may include, for example, film deposition, oxide film formation, nitride film formation, and metal-containing film formation. Moreover, the specific content of the substrate processing is not limited, and it can be suitably applied not only to the above-mentioned processing such as film deposition, but also to other substrate processing such as annealing, oxidation, nitriding, diffusion, and lithography.
[0094] Furthermore, this disclosure can be suitably applied to other substrate processing equipment, such as annealing equipment, oxidation equipment, nitriding equipment, exposure equipment, coating equipment, drying equipment, heating equipment, and plasma-based processing equipment. Moreover, this disclosure may apply to a mixture of these equipment.
[0095] Furthermore, although this embodiment describes a semiconductor manufacturing process, this disclosure is not limited thereto. For example, this disclosure can also be applied to substrate processing such as the manufacturing process of liquid crystal devices, solar cells, light-emitting devices, glass substrate processing, ceramic substrate processing, and conductive substrate processing.
[0096] Furthermore, it is possible to replace parts of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add configurations from other embodiments to the configuration of one embodiment. In addition, it is possible to add, delete, or replace parts of the configuration of each embodiment with other configurations.
[0097] For example, this technology is applied to monitoring the state of components that make up a substrate processing apparatus for forming a film on a substrate (e.g., a semiconductor wafer).
[0098] 36 Exhaust system (pump) 82 Predictive detection controller 260 Controller
Claims
1. A condition monitoring method comprising: a step of comparing a measured value of the peak frequency of the primary frequency of the exhaust system, calculated based on data acquired from a sensor installed in the exhaust system, with a reference value of the peak frequency of the primary frequency of the exhaust system, and confirming a fluctuation in the frequency corresponding to the rated rotational speed of the exhaust system, which is set in advance for a specific period; and a first step of identifying an abnormality that is predicted based on the fluctuation in the frequency.
2. The state monitoring method according to claim 1, wherein the first step is performed if the frequency fluctuation is greater than or equal to a first threshold.
3. The condition monitoring method according to claim 2, wherein the frequency fluctuation is a shift of the peak frequency of the primary frequency of the exhaust device toward the lower frequency side.
4. The condition monitoring method according to claim 1, further comprising the step of acquiring data related to the exhaust system, including data acquired from a sensor provided in the exhaust system.
5. The condition monitoring method according to claim 4, further comprising a first table showing the relationship between items indicating predicted abnormalities and data necessary for identifying the factors causing those items, wherein in the first step, the measured values of the data related to the exhaust system are compared with the normal data based on the normal data of the data related to the exhaust system and the first table, and an abnormality related to the exhaust system is predicted.
6. The status monitoring method according to claim 5, further comprising a second step of determining error processing in accordance with the abnormality identified in the first step, wherein in the second step, one of the first process, second process and third process is determined based on a second table for identifying the error processing.
7. The condition monitoring method according to claim 5, wherein in the first step, if there is a decrease in the rotational speed of the exhaust device, an abnormal item indicating gas load is identified.
8. The state monitoring method according to claim 7, wherein in the second step, if the decrease in rotational speed continues beyond a predetermined duration in relation to an abnormal item indicating gas load, it is determined to be a second process, and if the decrease in rotational speed is less than the predetermined duration, it is determined to be a first process.
9. The condition monitoring method according to claim 5, wherein in the first step, an abnormal item indicating a gas load due to the inclusion of minute leak gas is identified.
10. The state monitoring method according to claim 9, wherein in the second step, if the decrease in the rotational speed of the exhaust device and the increase in the secondary pressure of the exhaust device exceed their respective thresholds, the second process is determined to be performed.
11. The condition monitoring method according to claim 5, wherein in the first step, if there are fluctuations in the ultimate pressure of the processing chamber, the secondary pressure of the exhaust device, and the temperature of the casing, abnormal items caused by clogging of by-products are identified.
12. The state monitoring method according to claim 11, wherein in the second step, if there is a fluctuation of more than a preset threshold for each of the ultimate pressure and the secondary pressure, and a decrease in the rotational speed of more than a preset threshold, the second process is determined to be performed.
13. The condition monitoring method according to claim 5, wherein in the first step, the acceleration of each axis is compared with a reference acceleration, and if the acceleration deviates from the reference acceleration, an abnormal item indicating deterioration of the vibration-damping rubber is identified.
14. The state monitoring method according to claim 13, wherein in the second step, if the decrease in rotational speed exceeds a predetermined threshold for each axis, the second process is determined to be performed.
15. The state monitoring method according to claim 6, wherein, if it is determined in the first step that there is a fluctuation in the current value of the exhaust device, and the state of exceeding the threshold of the exhaust device continues, the second step determines that there is a motor failure of the exhaust device and proceeds to the third process.
16. The status monitoring method according to claim 6, further comprising a third step of executing the error processing, wherein in the third step, if the second step is determined to be the first process, the substrate processing is continued, and if the second process is determined to be the second process, the first step is not executed after the substrate processing in the processing chamber is completed.
17. The condition monitoring method according to claim 1, further comprising a step of acquiring standard normal data, wherein the normal data includes data related to the exhaust system as target data for detection items that identify abnormalities.
18. A method for manufacturing a semiconductor device, comprising the steps of: monitoring the state of a component using the state monitoring method described in claim 1; and processing a substrate.
19. A program that causes a computer to perform the following steps: compare the measured peak frequency of the primary frequency of the exhaust system, calculated based on data acquired from a sensor installed in the exhaust system, with a reference value for the peak frequency of the primary frequency of the exhaust system, and to confirm fluctuations in the rated frequency corresponding to a preset rotational speed of the exhaust system over a specific period; and to identify an abnormality predicted based on the fluctuations in the frequency.
20. A substrate processing apparatus comprising: a sensor provided in an exhaust device for exhausting the atmosphere of a processing chamber for detecting vibrations of the exhaust device; a data processing unit for processing data detected by the sensor; a data acquisition unit for acquiring data related to the exhaust, including the data detected by the sensor; a storage unit for storing normal data that serves as a reference for data related to the exhaust device; and a control unit configured to perform a process of comparing a measured value of the peak frequency of the primary frequency of the exhaust device calculated based on the data acquired from the sensor with a reference value of the peak frequency of the primary frequency of the exhaust device, and a process of identifying an abnormality predicted based on fluctuations in the frequency corresponding to the rated rotational speed of the exhaust device.
