Substrate processing system and substrate abnormality detection method

Abstract

The substrate processing system includes a loading port, a processing chamber, a measuring portion, and a control portion. The loading port is configured to be connectable to a storage container for storing a substrate. The processing chamber is configured so as to be able to perform substrate processing on a substrate. The measuring part is arranged on a path for conveying the substrate between the storage container and the processing chamber, and is configured to be capable of measuring the spectral distribution in the surface of the substrate. The control unit is configured to generate an outlier detection model on the basis of the in-plane spectral distribution of the substrate measured by the measurement unit, calculate an in-plane score of the substrate from the measured in-plane spectral distribution of the substrate using the generated outlier detection model, and detect an abnormality of the substrate on the basis of the calculated score.

Description

Technical Field

[0001] The invention relates to a substrate processing system and a substrate abnormality detection method. Background Art

[0002] The following patent document 1 describes a substrate processing method for a substrate processing device, the substrate processing device comprising: a processing chamber, which processes a substrate to be processed based on a predetermined processing parameter; and a measuring chamber, which measures the surface profile of the substrate to be processed before and after the processing. The substrate processing method is characterized in that it comprises: a pre-processing measurement step, in which the surface profile of the substrate to be processed is measured in the measuring chamber before the processing; a first calculation step, before the substrate to be processed is transported to the processing chamber, a value of a processing parameter for achieving a target surface profile is calculated based on the measured value of the surface profile before the processing; and a judgment step, in which whether the calculated value of the processing parameter is within a second calculation step, in which, when it is determined in the determination step that the substrate to be processed is within the allowable range, the value of the processing parameter for achieving the target surface profile is recalculated based on the adjustment value and the measured value of the surface profile before processing during the period from when the substrate to be processed is conveyed to the processing chamber to when the substrate to be processed is conveyed to the processing chamber, wherein the adjustment value is an adjustment value based on the measured value of the surface profile after processing obtained from the substrate to be processed that has just been processed in the processing chamber; and a processing step, in which the substrate to be processed is conveyed to the processing chamber and the processing is performed based on the value of the processing parameter calculated in the second calculation step.

[0003] Prior art literature

[0004] Patent Literature

[0005] Patent Document 1: Japanese Patent Application Publication No. 2008-091816 Summary of the invention

[0006] Technical problem to be solved by the invention

[0007] The present invention provides a technology capable of detecting abnormality of a substrate even when a measured value of a surface profile of the substrate cannot be obtained.

[0008] Technical means for solving technical problems

[0009] A substrate processing system of one embodiment of the present invention includes a loading port, a processing chamber, a measuring unit, and a control unit. The loading port is configured to be connected to a storage container for storing substrates. The processing chamber is configured to perform substrate processing on the substrate. The measuring unit is arranged on a path for conveying the substrate between the storage container and the processing chamber, and is configured to measure the spectral distribution within the surface of the substrate. The control unit is configured to generate an outlier (also called an "abnormal value") detection model based on the spectral distribution within the surface of the substrate measured by the measuring unit, calculate the score within the surface of the substrate based on the measured spectral distribution within the surface of the substrate using the generated outlier detection model, and detect the abnormality of the substrate based on the calculated score.

[0010] Effects of the Invention

[0011] According to the present invention, even when the measured value of the surface profile of the substrate cannot be obtained, it is possible to detect abnormality of the substrate. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 1 is a system configuration diagram showing an example of the configuration of a substrate processing system according to an embodiment.

[0013] Figure 2 It is a diagram for explaining a structure for measuring the spectrum distribution within the plane of a substrate according to an embodiment.

[0014] Figure 3 A diagram for explaining an example of the operation of the substrate processing system according to the embodiment when performing substrate processing.

[0015] Figure 4A It is a diagram showing an example of the structure of a substrate as a test piece according to the embodiment.

[0016] Figure 4B It is a diagram showing an example of the structure of a substrate as a test piece according to the embodiment.

[0017] Figure 5 It is a diagram showing an example of the arrangement of measurement points on a substrate according to an embodiment.

[0018] Figure 6 FIG. 1 is a diagram showing an example of a spectrum at each measurement point of the substrate according to the embodiment.

[0019] Figure 7 It is a diagram showing an example of measurement points in the center portion and measurement points in the peripheral portion of the substrate according to the embodiment.

[0020] Fig. 8A This is a diagram showing an example of a spectrum of a measurement point at the center of the substrate according to the embodiment.

[0021] Figure 8B1 is a diagram showing an example of spectrum of a measurement point in the center and a measurement point in the peripheral portion of the substrate according to the embodiment.

[0022] Fig. 9 This is a diagram showing an example of the distribution of scores within the surface of the substrate according to the embodiment.

[0023] Fig.10 This is a diagram showing an example of the outline of a substrate according to an embodiment.

[0024] Fig.11 This is a diagram showing an example of an abnormality score for each substrate according to the embodiment.

[0025] Fig.12 This is a flowchart showing an example of a process of substrate processing including processing of the substrate abnormality detection method according to the embodiment. DETAILED DESCRIPTION

[0026] Hereinafter, embodiments of the substrate processing system and the substrate abnormality detection method will be described in detail based on the drawings. In addition, the substrate processing system and the substrate abnormality detection method of the present invention are not limited to the following embodiments.

[0027] In the prior art, the surface profile of the substrate is measured, and the pass or fail of the substrate is judged based on the surface profile. When judging based on such a surface profile, for example, characteristic values ​​such as CD (Critical Dimension) and Thickness of the pattern formed on the substrate are measured as surface profiles by ellipsometry (ellipsometry technology), scanning electron microscope (SEM), and transmission electron microscope (TEM). The pass or fail judgment is made based on whether these surface profiles are within the range of the judgment criterion. However, since the measurement of these surface profiles takes time, it is difficult to perform it on all substrates from the perspective of productivity. Therefore, in the case where the measured values ​​such as the surface profile of the substrate used to judge the pass or fail of the substrate cannot be obtained, the abnormality of the substrate cannot be detected.

[0028] Therefore, it is expected that abnormality of a substrate can be detected even when the measured value of the surface profile of the substrate cannot be obtained.

[0029] (Implementation Method)

[0030] [Structure of substrate processing system 1]

[0031] Figure 1 1 is a system configuration diagram showing an example of the configuration of the substrate processing system 1 according to the embodiment. Figure 1In the figure, for convenience of explanation, some internal components of the device are shown in perspective. The substrate processing system 1 includes a device body 10 and a control device 20 that controls the device body 10.

[0032] The apparatus body 10 includes a vacuum conveying module 11 , an alignment module 12 , a plurality of load lock modules 13 , an atmospheric conveying module 14 , and a plurality of processing modules 30 .

[0033] The vacuum transfer module 11 is formed in a rectangular shape in a plan view. The two side walls of the vacuum transfer module 11 on the long side are connected to a plurality of processing modules 30 via gate valves G2.

[0034] Each processing module 30 is configured to perform substrate processing on the substrate W. In the present embodiment, each processing module 30 is configured to perform etching on the substrate W using plasma. Figure 1 In the example, six processing modules 30 are connected to the vacuum transport module 11, but the number of processing modules 30 connected to the vacuum transport module 11 may be more than 6 or less than 6. In this embodiment, each processing module 30 corresponds to a processing chamber of the present invention.

[0035] The vacuum transfer module 11 is connected to a plurality of load lock modules 13 via gate valves G3 on one side wall of the short side. Figure 1 In the example, two load lock modules 13 are connected to the vacuum conveying module 11. The number of the load lock modules 13 connected to the vacuum conveying module 11 may be more than two or less than two.

[0036] A transfer robot (also referred to as a transfer robot) 110 is disposed in the vacuum transfer module 11. The transfer robot 110 transfers the substrate W between the load lock module 13 and the process module 30. The vacuum transfer module 11 is maintained at a pressure atmosphere lower than atmospheric pressure.

[0037] Each load lock module 13 is connected to the vacuum transfer module 11 via a gate valve G3. In addition, each load lock module 13 is connected to the atmospheric transfer module 14 via a gate valve G4 at the side wall opposite to the side wall connected to the vacuum transfer module 11. When the substrate W is delivered from the atmospheric transfer module 14 to the load lock module 13 via the gate valve G4, the gate valve G4 is closed, and the pressure in the load lock module 13 is reduced from the atmospheric pressure to a predetermined pressure. Then, the gate valve G3 is opened, and the substrate W in the load lock module 13 is delivered to the vacuum transfer module 11 by the transfer robot 110.

[0038] In addition, when the pressure in the load lock module 13 is lower than the atmospheric pressure, the substrate W is transferred from the vacuum transfer module 11 to the load lock module 13 via the gate valve G3 by the transfer robot 110, and then the gate valve G3 is closed. Then, the pressure in the load lock module 13 is raised to the atmospheric pressure. Then, the gate valve G4 is opened to transfer the substrate W in the load lock module 13 to the atmospheric transfer module 14.

[0039] The atmospheric transfer module 14 is provided with a plurality of loading ports 15 on the side wall opposite to the side wall connected to the load lock module 13. The atmospheric transfer module 14 is provided with an alignment module 12 on a side wall different from the side wall connected to the load lock module 13 and the side wall provided with the loading ports 15.

[0040] The alignment module 12 is configured to be able to perform alignment of the substrate W. For example, the alignment module 12 can be rotated by a motor, and has a stage on which the substrate W is mounted and a sensor for detecting the outer periphery of the substrate W mounted on the stage. The alignment module 12 detects a notch formed on the outer periphery of the substrate W by the sensor, and rotates the stage so that the direction of the notch becomes a predetermined direction, thereby aligning (adjusting) the rotation direction of the substrate W.

[0041] Furthermore, the alignment module 12 is configured to be able to measure the spectral distribution within the plane of the substrate W. Here, a configuration for measuring the spectral distribution within the plane of the substrate W in the alignment module 12 will be described. Figure 2 1 is a diagram for explaining a structure for measuring the spectrum distribution in the plane of the substrate W according to the embodiment. The alignment module 12 has a stage 50. The substrate W is placed on the stage 50. The stage 50 can be rotated by a motor. In addition, the stage 50 is configured to be able to slide and move in the lateral direction by a sliding mechanism not shown in the figure.

[0042] A light source 51 such as an LED (Light Emitting Diode) that outputs light is provided above the stage 50. The light source 51 is configured so that the output light is incident on the stage 50 at a predetermined incident angle. The light output from the light source 51 is reflected by the substrate W on the stage 50. The light reflected by the substrate W is incident on the detection unit 53. The detection unit 53 is composed of, for example, a hyperspectral camera, and is capable of measuring the spectral distribution of a hyperspectral spectrum obtained by splitting light according to wavelengths.

[0043] In the alignment module 12 of the substrate processing system 1, while moving the stage 50 with the substrate W placed thereon, light is output from the light source 51, and the spectral distribution of the light reflected by the substrate W is measured by the detection unit 53, thereby measuring the spectral distribution within the surface of the substrate W. In this way, the alignment module 12 can measure the spectral distribution within the surface of the substrate W in a short time. The detection unit 53 outputs the data of the measured spectral distribution within the surface of the substrate W to the control device 20. In this embodiment, the alignment module 12 corresponds to the measuring unit of the present invention. By providing a measuring unit capable of measuring in a short time in the path for conveying the substrate W in the substrate processing system 1, a full inspection of the substrate W can be performed.

[0044] return Figure 1 Each loading port 15 is connected to a storage container 16 such as a FOUP (Front Opening Unified Pod) capable of storing a plurality of substrates W. The pressure in the atmospheric transport module 14 is atmospheric pressure. The storage container 16 stores, for example, a plurality of substrates W of one batch.

[0045] A conveying robot 140 is provided in the atmospheric conveying module 14. The conveying robot 140 conveys the substrate W between the alignment module 12, the load lock module 13 and the storage container 16 connected to the loading port 15. An FFU (Fan Filter Unit) is provided at the upper part of the atmospheric conveying module 14, and air from which particles and the like are removed is supplied from the upper part to the atmospheric conveying module 14, so that a downward flow is formed in the atmospheric conveying module 14. In addition, in the present embodiment, the atmospheric conveying module 14 is an atmospheric pressure atmosphere, but as another mode, the pressure in the atmospheric conveying module 14 can also be controlled to be a positive pressure. Thus, it is possible to suppress particles and the like from invading the atmospheric conveying module 14 from the outside.

[0046] The control device 20 processes computer executable commands that cause the device body 10 to perform the various steps described in the present invention. The control device 20 can be configured to control the various elements of the device body 10 to perform the various steps described herein. In one embodiment, a part or all of the control device 20 may also be included in the device body 10. The control device 20 may also include a processing unit 20a1, a storage unit 20a2, and a communication interface 20a3. The control device 20 is implemented, for example, by a computer 20a. The processing unit 20a1 may be configured to perform various control actions by reading a program from the storage unit 20a2 and executing the read program. The program may be stored in the storage unit 20a2 in advance, or may be obtained via a medium when necessary. The obtained program is stored in the storage unit 20a2, and is read out and executed from the storage unit 20a2 by the processing unit 20a1. The medium may be various storage media readable by the computer 20a, or may be a communication line connected to the communication interface 20a3. In the storage unit 20a2, in addition to the program, the processing scheme and the data referenced when executing the processing scheme are also stored. The processing unit 20a1 can also be a CPU (Central Processing Unit). The storage unit 20a2 can also include RAM (Random Access Memory), ROM (Read Only Memory), HDD (Hard Disk Drive), SSD (Solid State Drive) or a combination thereof. The communication interface 20a3 can also communicate with the device body 10 via a communication line such as a LAN (Local Area Network). In the present embodiment, the control device 20 corresponds to the control unit of the present invention.

[0047] [Operation of the substrate processing system 1]

[0048] Next, an example of the operation of the substrate processing system 1 according to the embodiment, in which the substrate W is transported between the storage container 16 and the processing module 30 and etching is performed on the substrate W, will be described. Figure 3 1 is a diagram for explaining an example of the operation of the substrate processing system 1 according to the embodiment when performing substrate processing. Figure 3 . Here, the operation of the substrate processing system 1 in the case where the substrate W stored in the storage container 16 mounted on the rightmost loading port 15a is transported to the lower right processing module 30a for etching is described. Figure 3 In FIG. 1 , arrows are used to indicate the transport path of the substrate W. The solid arrows indicate the transport path of the normal substrate W. The dotted arrows indicate the transport path of the substrate W that is detected to be abnormal and excluded from the etching target.

[0049] The control device 20 controls the transport robot 140 to take out the substrate W before substrate processing from the storage container 16 of the load port 15 a , transport the taken-out substrate W to the alignment module 12 , and place it on the stage 50 .

[0050] The control device 20 controls the alignment module 12 to align (adjust) the position of the substrate W. The alignment module 12 detects a notch formed on the outer periphery of the substrate W by a sensor (not shown), and rotates the stage 50 to align the rotation direction of the substrate W so that the direction of the notch becomes a predetermined direction.

[0051] In addition, the control device 20 controls the alignment module 12 to measure the spectral distribution within the plane of the substrate W. The alignment module 12 outputs light from the light source 51 while moving the stage 50 with the substrate W placed thereon, and measures the spectral distribution of light reflected by the substrate W using the detection unit 53, thereby measuring the spectral distribution within the plane of the substrate W. The detection unit 53 outputs the data of the measured spectral distribution within the plane of the substrate W to the control device 20.

[0052] If the alignment and the measurement of the spectral distribution are completed, the control device 20 controls the transport robot 140 to take out the substrate W from the alignment module 12 .

[0053] Data of the spectral distribution within the surface of the substrate W is input from the detection unit 53 to the control device 20. Based on the data input from the detection unit 53, the control device 20 detects an abnormality of the substrate W according to the spectral distribution within the surface of the substrate W measured by the detection unit 53. For example, the control device 20 generates an outlier detection model based on the spectral distribution of a partial area within the surface of the substrate W in the spectral distribution within the surface of the substrate W measured. Alternatively, the control device 20 may also generate an outlier detection model based on all spectral distributions measured within the surface of the substrate W in the spectral distribution within the surface of the substrate W measured. The control device 20 uses the generated outlier detection model to calculate the score within the surface of the substrate W according to the spectral distribution within the surface of the substrate W measured. The control device 20 detects an abnormality of the substrate W based on the calculated score. The details of the abnormality detection for detecting the abnormality of the substrate W will be described later. In the abnormality detection, it is detected whether the substrate W has an abnormality, and if the substrate W has an abnormality, it is determined whether the abnormality can be corrected.

[0054] When the substrate W has an abnormality and the abnormality cannot be corrected, the control device 20 excludes the substrate W from the etching target and stores it in the storage container 16. For example, the control device 20 controls the conveying robot 140 to store the substrate W held by the conveying robot 140 in the storage container 16 of the leftmost loading port 15d as shown by the dotted arrow.

[0055] On the other hand, when there is no abnormality in the substrate W or when there is an abnormality in the substrate W but the abnormality can be corrected, the control device 20 transports the substrate to the processing module 30 a to perform etching.

[0056] For example, the control device 20 opens the gate valve G4 of any load lock module 13. The control device 20 controls the conveying robot 140 to convey and place the substrate W held by the conveying robot 140 into one load lock module 13. When the conveying robot 140 retreats from the load lock module 13, the control device 20 closes the gate valve G4 of one load lock module 13.

[0057] The control device 20 controls an exhaust device (not shown) of one load lock module 13 on which the substrate W is placed to exhaust the air in the chamber, thereby switching the one load lock module 13 from the air atmosphere to the vacuum atmosphere.

[0058] The control device 20 opens the gate valve G3 of one loading lock module 13. The control device 20 controls the conveying robot 110 to hold the substrate W placed on one loading lock module 13 and convey it to the vacuum conveying module 11. When the conveying robot 110 retreats from one loading lock module 13, the control device 20 closes the gate valve G3 of one loading lock module 13. The control device 20 opens the gate valve G2 of the processing module 30a. The control device 20 controls the conveying robot 110 to convey the substrate W held by the conveying robot 110 to the processing module 30a and place it. When the conveying robot 110 retreats from the processing module 30a, the control device 20 closes the gate valve G2 of the processing module 30a.

[0059] The control device 20 controls the processing module 30a to perform etching on the substrate W (substrate processing).

[0060] When etching of the substrate W is completed, the control device 20 opens the gate valve G2 of the processing module 30a. The control device 20 controls the transport robot 110 to take out the substrate W from the processing module 30a. When the transport robot 110 retreats from the processing module 30a, the control device 20 closes the gate valve G2 of the processing module 30a.

[0061] The control device 20 opens the gate valve G3 of any load lock module 13. The control device 20 controls the transport robot 110 to transport the processed substrate W held by the transport robot 110 to one load lock module 13 and place it. When the transport robot 110 retreats from one load lock module 13, the control device 20 closes the gate valve G3 of one load lock module 13.

[0062] The control device 20 controls a suction device (not shown) of one load lock module 13 on which the substrate W is placed to supply, for example, clean air into the chamber, thereby switching the one load lock module 13 from a vacuum atmosphere to an air atmosphere.

[0063] The control device 20 opens the gate valve G4 of one load lock module 13. The control device 20 controls the transport robot 140 to take out the substrate W from one load lock module 13 and store the taken out substrate W in the storage container 16 of the second load port 15b from the right.

[0064] In addition, the example in which the substrate processing system 1 transports the substrate W to the processing module 30a for etching is described, but the substrate W may be transported to other processing modules 30 for etching. In addition, the etched substrate W may be transported to, for example, other processing modules 30, and other substrate processing may be further performed on the substrate W in the other processing modules 30.

[0065] [Abnormality detection of substrate W]

[0066] Here, a specific example of detecting an abnormality of the substrate W is described. Figure 4A and Figure 4B The substrate W having the structure shown was used as a test piece to detect abnormalities. Figure 4A and Figure 4B 1 is a diagram showing an example of the structure of a substrate W as a test piece according to the embodiment. Figure 4A denoted a cross section of substrate W. Figure 4B : represents an enlarged upper surface of a substrate W. The substrate W is formed with an oxide layer 61, a nitride layer 62, a polysilicon layer 63 and a mask layer 64 in order from the bottom layer on a silicon substrate 60 such as bare silicon. The oxide layer 61 is, for example, a silicon oxide film. The nitride layer 62 is, for example, SiN. The mask layer 64 is, for example, TEOS (Tetraethoxysilane). Holes 65 are formed in the polysilicon layer 63 and the mask layer 64. The holes 65 are arranged on the entire upper surface of the substrate W.

[0067] The control device 20 generates an outlier detection model based on the measured spectral distribution within the surface of the substrate W. For example, the control device 20 generates an outlier detection model based on the spectral distribution of a partial area within the surface of the substrate W in the measured spectral distribution within the surface of the substrate W. For example, the control device 20 generates an outlier detection model based on the spectral distribution of the central part within the surface of the substrate W. For example, a plurality of measurement points are determined on the upper surface of the substrate W. The measurement points may be arranged at least in the central part and the peripheral part of the substrate W, and preferably arranged on the entire upper surface of the substrate W. Figure 5 FIG. 6 is a diagram showing an example of the arrangement of the measurement points 66 on the substrate W according to the embodiment. Figure 5In FIG. 7 , 70 measurement points 66 are arranged on the upper surface of the substrate W. When the upper surface of the substrate W is divided into regions to be semiconductor chips, the measurement points 66 are preferably arranged at the same positions in the regions to be semiconductor chips.

[0068] The control device 20 extracts the spectrum of each measurement point 66 on the upper surface of the substrate W based on the data input from the detection unit 53 . Figure 6 1 is a diagram showing an example of a spectrum at each measurement point 66 of the substrate W according to the embodiment. Figure 6 Overlapping Representation Figure 5 The spectra of 70 measurement points 66 are shown. Figure 6 As shown, differences are observed in the waveforms of the spectrum (ie, frequency spectrum) at the 70 measurement points 66. The width of the waveform of the spectrum at the 70 measurement points 66 indicates the difference in the profile (surface shape) within the plane of the substrate W.

[0069] Here, the spectrum of the measurement point 66 at the center of the surface of the substrate W and the spectrum of the measurement point 66 at the peripheral portion of the substrate W are compared. Figure 7 1 is a diagram showing an example of the measurement points 66 at the center and the measurement points 66 at the peripheral portion of the substrate W according to the embodiment. Figure 7 In FIG. 1 , four measurement points 66 a at the center of the substrate W are indicated by being surrounded by lines 67 , and measurement points 66 b at the peripheral portion are indicated by being surrounded by lines 68 .

[0070] Fig. 8A : is a diagram showing an example of a spectrum of a measurement point 66 at the center of a substrate W according to the embodiment. Fig. 8A In Figure 7 The spectrum of the four measurement points 66a surrounded by the line 67. The waveforms of the spectrum of the four measurement points 66a in the center are almost different. Therefore, it can be inferred that the contours of the four measurement points 66a in the center are almost the same.

[0071] Figure 8B 1 is a diagram showing an example of the spectrum of the measurement point 66 at the center and the measurement point 66 at the peripheral portion of the substrate W according to the embodiment. Figure 8B middle, Figure 7 The spectrum of the measurement point 66a1 surrounded by the line 69 among the four measurement points 66a in the center shown is represented by the line L11. Figure 8B In the example, line L12 is used to represent Figure 7 Spectra of measurement point 66b surrounded by line 68 in the peripheral portion of substrate W are shown. Differences in waveform are observed between the spectrum (line L11) of measurement point 66a1 in the center portion of substrate W and the spectrum (line L12) of measurement point 66b in the peripheral portion. It can be inferred that there are differences in the contours of the center portion and the peripheral portion of substrate W.

[0072] The control device 20 generates an outlier detection model based on the spectral distribution of the center portion within the plane of the substrate W. For example, the control device 20 generates an outlier detection model based on Figure 7 The outlier detection model is generated by using the spectra of the four measurement points 66a in the center of the substrate W surrounded by the line 67. As outlier detection models, for example, LOF (Local Outlier Factor), k-nearest neighbor method and One class SVM (Support Vector Machine) can be listed. LOF is an algorithm for finding outliers from a certain set of data. The outlier detection model uses the spectra of the four measurement points 66a in the center as the benchmark data and calculates the difference representing the degree of deviation from the benchmark data as a score.

[0073] The control device 20 uses the generated outlier detection model to calculate the score within the surface of the substrate W based on the measured spectral distribution within the surface of the substrate W. For example, the control device 20 inputs the spectra of all the measurement points 66 into the outlier detection model, and calculates the difference with the reference data as the score.

[0074] Fig. 9 This is a diagram showing an example of the distribution of scores within the surface of the substrate W according to the embodiment. Fig. 9 This is the result of calculating the score of each measurement point in the surface of the substrate W using the outlier detection model, and finding the distribution of the score in the surface of the substrate W by interpolation using the calculated score. Fig. 9 The score of the center of the substrate W shown is small, and the score increases toward the outer side. From this, it can be inferred that the profile of the substrate W is different between the center and the peripheral part.

[0075] Here, the profile of the cross section of the substrate W obtained by measuring the SEM image obtained by photographing the cross section of the substrate W as a test object using a scanning electron microscope (SEM) will be described. Fig.10 FIG. 1 is a diagram showing an example of the outline of a substrate W according to an embodiment. Fig.10 In the example, the line L21 indicates Figure 7 The contour at the measurement point 66a1 in the center is shown. Line L21 represents the average surface shape of the hole 65 using the average value of the CD of each depth of the plurality of holes 65 included in the measurement point 66a1. Fig.10 middle, Figure 7 The profile at the measurement point 66b of the peripheral portion of the substrate W shown is represented by line L22. Line L22 represents the average surface profile of the hole 65 using the average value of the CD of each depth of the plurality of holes 65 included in the measurement point 66ab. Fig.10, the range of the nitride layer 62, the polysilicon layer 63, and the mask layer 64 are shown. A difference is observed in the upper part between the profile of the measurement point 66a1 in the center of the substrate W (line L21) and the profile of the measurement point 66b in the peripheral portion (line L22), and in particular, the shape of the mask layer 64 is different.

[0076] This makes it possible to confirm that the score calculated by the outlier detection model is a value that indicates a change in the profile at each measurement point relative to the profile of the center portion of the substrate W serving as a reference.

[0077] The control device 20 detects an abnormality of the substrate W based on the calculated score. For example, the control device 20 obtains the difference between the score of the center portion of the substrate W and the score of the peripheral portion as an abnormality score indicating the degree of abnormality for each substrate W. For example, the control device 20 calculates (the average value of the score of the center portion of the substrate W) - (the average value of the score of the outermost peripheral portion of the substrate W) as the abnormality score for each substrate W.

[0078] The control device 20 detects a substrate W having an abnormality score of more than a threshold value as abnormal, and detects a substrate W having an abnormality score less than the threshold value as normal. The threshold value may be determined in advance. In addition, the threshold value may also be set dynamically. For example, the threshold value may also be set dynamically by comparing the cross-sectional analysis (cross-sectional analysis) and electrical characteristics of the processed substrate W with the abnormality score actually measured for the substrate W.

[0079] Fig.11 : is a diagram showing an example of an abnormality score for each substrate W in the embodiment. Fig.11 In the example, the wafers W are assigned numbers (Wafer Count) in sequence, indicating the anomaly score (Anomaly Score) of the wafers W with each number. Fig.11 In , the threshold is represented by line L31. Fig.11 In FIG. 1 , substrates W having abnormality scores with numbers above line L31 are judged to be abnormal.

[0080] The control device 20 controls to transport a normal substrate W in which no abnormality is detected to the processing module 30 , and to perform etching on the substrate W.

[0081] On the other hand, the control device 20 determines whether the substrate W in which the abnormality is detected can be corrected. For example, the control device 20 inputs the abnormality score of the substrate W in which the abnormality is detected into the control model, and determines whether the correction can be made through the control model. The control model is a model that learns whether the correction can be made and the correction amount of the processing condition of the processing module 30 for improving the uniformity of the profile within the surface of the substrate W if the correction can be made for each abnormality score.

[0082] The control device 20 inputs the abnormality score of the substrate W detected as abnormal into the control model. When the control device 20 obtains a result that cannot be corrected from the control model, the substrate W determined to be uncorrectable is excluded from the etching target and stored in the storage container 16. For example, Figure 3 As shown, the control device 20 stores the substrate W determined to be uncorrectable in the storage container 16 of the leftmost loading port 15d.

[0083] On the other hand, when the control device 20 obtains a correctable result from the control model, the control device 20 transports the correctable substrate W to the processing module 30. Then, the control device 20 controls so that the processing conditions are corrected by the correction amount obtained from the control model, and the substrate W is etched under the corrected processing conditions.

[0084] In this way, the substrate processing system 1 of the embodiment is configured to measure the spectral distribution within the surface of the substrate W by the alignment module 12. The control device 20 generates an outlier detection model based on the measured spectral distribution within the surface of the substrate W. For example, the substrate processing system 1 generates an outlier detection model based on the spectral distribution of a partial area within the surface of the substrate W in the measured spectral distribution within the surface of the substrate W. The substrate processing system 1 uses the generated outlier detection model to calculate the score within the surface of the substrate W according to the measured spectral distribution within the surface of the substrate W, and detects the abnormality of the substrate W based on the calculated score. Thus, the substrate processing system 1 can detect the abnormality of the substrate W even when the measured value of the surface profile of the substrate W, such as the information of the pattern of the substrate, cannot be obtained. Here, in the case of measuring the surface profile of the pattern formed on the substrate W, the measurement time becomes long, and it is impossible to perform a full inspection from the perspective of productivity. The substrate processing system 1 can inspect the abnormality of the substrate W in full even when the measured value of the surface profile of the substrate W cannot be obtained in this way. In addition, the substrate processing system 1 can monitor the in-plane variation of the substrates W before etching. In addition, the substrate processing system 1 can monitor the in-lot variation of the substrates W before etching by comparing the calculated scores of the respective substrates W.

[0085] [Substrate abnormality detection method]

[0086] Fig.12 The control device 20 reads the program from the storage unit 20a2, executes the read program, and controls each part of the device body 10 via the communication interface 20a3, thereby achieving Fig.12 The processing illustrated. Fig.12 The illustrated process is executed, for example, for each substrate W stored in the storage container 16 mounted on the load port 15 a.

[0087] The control device 20 controls the conveying robot 140 to convey the substrate W to the alignment module 12 (S10). The control device 20 controls the alignment module 12 to align the rotation position of the substrate W (S11). In addition, the control device 20 controls the alignment module 12 to measure the spectral distribution within the surface of the substrate W (S12). For example, the alignment module 12 moves the stage 50 while the substrate W is placed thereon, outputs light from the light source 51, and measures the spectral distribution of the light reflected by the substrate W using the detection unit 53, thereby measuring the spectral distribution within the surface of the substrate W. The detection unit 53 outputs the data of the measured spectral distribution within the surface of the substrate W to the control device 20.

[0088] The control device 20 generates an outlier detection model based on the spectral distribution in the plane of the substrate W based on the data input from the detection unit 53 (S13). For example, the control device 20 generates an outlier detection model based on the spectral distribution of a partial region in the plane of the substrate W among the spectral distribution in the plane of the substrate W based on the data input from the detection unit 53. For example, the control device 20 generates an outlier detection model based on the spectral distribution of the center portion in the plane of the substrate W.

[0089] The control device 20 calculates the score within the plane of the substrate W from the measured spectral distribution within the plane of the substrate W using the generated outlier detection model ( S14 ).

[0090] The control device 20 detects an abnormality of the substrate W based on the calculated score (S15). For example, the control device 20 obtains the difference between the score of the center portion and the score of the peripheral portion of the substrate W as the abnormality score. The control device 20 detects that the substrate W is abnormal when the abnormality score is greater than a threshold value.

[0091] The control device 20 determines whether an abnormality is detected in the substrate W (S16). If an abnormality is detected (S16: Yes), the control device 20 inputs the abnormality score of the substrate W in which the abnormality is detected into the control model, and determines whether correction can be performed through the control model (S17).

[0092] The control device 20 determines whether the determination result is that the correction is possible or not (S18). If the determination result is that the correction is not possible (S18: No), the control device 20 excludes the substrate W determined to be not correctable from the etching target and stores it in the storage container 16 (S19). Figure 3 As shown, the control device 20 stores the substrate W in the storage container 16 of the load port 15d.

[0093] On the other hand, when no abnormality is detected (S16: No) and when the judgment result is that the abnormality can be corrected (S18: Yes), the control device 20 controls the conveying robot arms 110, 140, gate valves G2 to G4, etc. to convey the substrate W to the processing module 30 (S20). The control device 20 controls the processing module 30a to perform etching on the substrate W (S21). Among them, when an abnormality is detected in the substrate W and it is judged that the abnormality can be corrected, the control device 20 controls so that the processing conditions are corrected by the correction amount obtained from the control model, and the substrate W is etched under the corrected processing conditions.

[0094] When the etching of the substrate W is completed, the control device 20 controls the conveying robot arms 110, 140, the gate valves G2 to G4, etc. to store the substrate W in the storage container 16 (S22), and the processing shown in this flowchart is terminated. Figure 3 As shown, the control device 20 stores the substrate W in the storage container 16 of the load port 15b.

[0095] In addition, in the above-mentioned embodiment, the case where the abnormality of the substrate W is detected based on the spectral distribution of the substrate W before etching is described as an example. However, it is not limited to this. The substrate processing system 1 can also measure the spectral distribution of the substrate W after etching, and detect the abnormality of the substrate W based on the spectral distribution of the substrate W after etching. For example, the control device 20 controls so that the substrate W that has been etched in the processing module 30 is transported to the alignment module 12, and after the spectral distribution of the substrate W is measured in the alignment module 12, the substrate W is transported to the storage container 16. The control device 20 can also detect the abnormality of the substrate W based on the spectral distribution of the substrate W after etching based on the data input from the detection unit 53. For example, the control device 20 generates an outlier detection model based on the spectral distribution within the surface of the substrate W after etching. For example, the control device 20 generates an outlier detection model based on the spectral distribution of a partial area (a part) within the surface of the substrate W in the spectral distribution within the surface of the substrate W after etching. For example, the control device 20 generates an outlier detection model based on the spectral distribution of the center portion of the surface of the substrate W. Then, the control device 20 can use the generated outlier detection model to calculate the score of the surface of the substrate W based on the measured spectral distribution of the surface of the substrate W after etching, and detect the abnormality of the substrate W after etching based on the calculated score. For example, the control device 20 detects the abnormality of the substrate W based on the difference between the score of the center portion of the substrate W and the score of the peripheral portion of the substrate W. Thus, the substrate processing system 1 can detect the abnormality of the substrate W after etching even if the measured value of the surface profile of the substrate W after etching is not obtained. In addition, the substrate processing system 1 can monitor the deviation in the surface of the substrate W after etching. In addition, the substrate processing system 1 can monitor the deviation within the batch of the substrate W after etching by comparing the calculated scores of each substrate W. In addition, the substrate processing system 1 can monitor the difference in etching rate between the center portion and the peripheral portion based on the difference between the score of the center portion of the substrate W and the score of the peripheral portion of the substrate W. The substrate W has a tilt or CD (Critical Dimension) value unevenness in the hole 65 at a portion where the etching rate difference is large. Therefore, the substrate processing system 1 can predict the occurrence of tilt or CD (Critical Dimension) value unevenness by monitoring the etching rate difference between the center and the peripheral portion of the substrate W.

[0096] In addition, the substrate processing system 1 may also measure the spectral distribution of the substrate W before and after etching, respectively, and detect abnormalities of the substrate W based on the spectral distribution of the substrate W before and after etching. For example, the substrate processing system 1 measures the spectral distribution of the substrate W before etching as in the above-mentioned embodiment. In addition, the control device 20 controls so that the substrate W that has been etched in the processing module 30 is transported to the alignment module 12, and after the spectral distribution of the substrate W is measured in the alignment module 12, the substrate W is transported to the storage container 16. The control device 20 may also detect abnormalities of the substrate W based on the spectral distribution of the substrate W before etching and the spectral distribution of the substrate W after etching based on the data input from the detection unit 53. For example, the control device 20 generates an outlier detection model based on the spectral distribution within the surface of the substrate W before etching and the spectral distribution within the surface of the substrate W after etching. For example, the control device 20 generates an outlier detection model based on the spectral distribution of the partial area (a portion) within the surface of the substrate W, respectively, of the spectral distribution within the surface of the substrate W before etching and the spectral distribution within the surface of the substrate W after etching. For example, the control device 20 calculates a difference spectral distribution (also called "differential spectral distribution") representing the difference (also called "difference") between the spectral distribution within the surface of the substrate W before etching and the spectral distribution within the surface of the substrate W after etching. For example, the control device 20 calculates the difference spectral distribution by obtaining the difference between the spectrum within the surface of the substrate W before etching and the spectrum within the surface of the substrate W after etching for each wavelength. Then, the control device 20 generates an outlier detection model based on the calculated difference spectral distribution. For example, the control device 20 generates an outlier detection model based on the spectral distribution of the partial area within the surface of the substrate W in the calculated difference spectral distribution. For example, the control device 20 generates an outlier detection model based on the spectral distribution of the central part within the surface of the substrate W. Then, the control device 20 uses the generated outlier detection model to calculate a score based on the spectral distribution within the surface of the substrate W before etching and the spectral distribution within the surface of the substrate W after etching. For example, the control device 20 uses the generated outlier detection model to calculate the score within the surface of the substrate W based on the difference spectral distribution. The control device 20 detects the abnormality of the substrate W based on the calculated score. Thus, the substrate processing system 1 can detect the abnormality of etching of the substrate W even when the measured value of the surface profile of the change of the film of the substrate W caused by etching cannot be obtained. In addition, the substrate processing system 1 can monitor the deviation within the surface of the substrate W before and after etching. In addition, the substrate processing system 1 can monitor the deviation within the batch of substrates W before and after etching by comparing the calculated scores of each substrate W. In addition, the spectrum within the surface of the substrate W before etching is related to the profile within the surface of the substrate W before etching. The spectrum within the surface of the substrate W after etching is related to the profile within the surface of the substrate W after etching. Therefore, the difference between the spectrum within the plane of the substrate W before etching and the spectrum within the plane of the substrate W after etching is related to the etching rate.Therefore, the substrate processing system 1 can more accurately monitor the difference in etching rate between the center and the peripheral portion of the substrate W by calculating the difference spectrum distribution, and can accurately predict the tilt and CD (Critical Dimension) value unevenness.

[0097] In addition, in the above-mentioned embodiment, the case where the alignment of the substrate W and the measurement of the spectral distribution are performed in the alignment module 12 is described as an example. However, it is not limited to this. The substrate processing system 1 may also measure the spectral distribution of the substrate W in a part other than the alignment module 12. For example, the substrate processing system 1 may also be provided with Figure 2 The substrate W is transported to the measurement module, and the spectral distribution of the substrate W is measured in the measurement module. In addition, the substrate processing system 1 can also be provided in the vacuum transport module 11, the load lock module 13, the atmospheric transport module 14, etc. Figure 2 The structure shown is capable of measuring the spectral distribution of the substrate W to measure the spectral distribution of the substrate W.

[0098] In addition, in the above-mentioned embodiment, the case where the spectral distribution within the surface of substrate W is measured for each substrate W, and the outlier detection model is generated according to the measured spectral distribution within the surface of substrate W for each substrate W is described as an example. However, it is not limited to this. The outlier detection model can measure the spectral distribution within the surface of a specific substrate W and be generated according to the spectral distribution within the surface of the specific substrate W. Substrates W other than the specific substrate W can also be scored using the outlier detection model generated using the specific substrate W. The specific substrate W can be, for example, the initial substrate W taken out from the storage container 16, or it can be a substrate W of a specified batch unit, or it can be the initial substrate W on which the substrate processing system 1 performs substrate processing after startup or maintenance.

[0099] In addition, in the above-mentioned embodiment, the case where each processing module 30 etches the substrate W as the substrate processing is described as an example. However, it is not limited to this. The substrate processing performed by each processing module 30 can be any processing performed on the substrate W such as film forming processing, ashing processing, etc.

[0100] The above is an explanation of the embodiment. As described above, the substrate processing system 1 of the embodiment has a loading port 15, a processing module 30 (processing chamber), an alignment module 12 (measuring unit) and a control device 20 (control unit). The loading port 15 is configured to be connected to a storage container 16 for storing substrates W. The processing module 30 is configured to be able to perform substrate processing on substrates W. The alignment module 12 is arranged on a path for conveying substrates W between the storage container 16 and the processing module 30, and is configured to be able to measure the spectral distribution within the surface of substrates W. The control device 20 is configured to generate an outlier detection model based on the spectral distribution within the surface of substrates W measured by the alignment module 12, calculate the score within the surface of substrates W based on the measured spectral distribution within the surface of substrates W using the generated outlier detection model, and detect the abnormality of substrates W based on the calculated score. Thus, the substrate processing system 1 can detect the abnormality of substrates W even when the measured value of the surface profile of substrates W is not obtained.

[0101] Generally, when the internal state of the processing module 30 (processing chamber) changes, the quality of the etched substrate W also changes. In addition, when the quality of the substrate W before substrate processing changes, the quality of the substrate W after substrate processing is affected.

[0102] In addition, the alignment module 12 is configured to be able to measure the spectral distribution within the surface of the substrate W before substrate processing that is transported from the storage container 16 to the processing module 30. The control device 20 is configured to be able to generate an outlier detection model based on the measured spectral distribution within the surface of the substrate W before substrate processing, calculate the score within the surface of the substrate W based on the measured spectral distribution within the surface of the substrate W before substrate processing using the generated outlier detection model, and detect abnormalities of the substrate W before substrate processing based on the calculated score. Thus, the substrate processing system 1 can detect abnormalities of the substrate W before substrate processing even when the measured value of the surface profile of the substrate W before substrate processing cannot be obtained.

[0103] In addition, the alignment module 12 is configured to be able to measure the spectrum distribution within the surface of the substrate W after the substrate processing that is conveyed from the processing module 30 to the storage container 16. The control device 20 is configured to be able to generate an outlier detection model based on the measured spectrum distribution within the surface of the substrate W after the substrate processing, calculate the score within the surface of the substrate W based on the measured spectrum distribution within the surface of the substrate W after the substrate processing using the generated outlier detection model, and detect abnormality of the substrate W after the substrate processing based on the calculated score. Thus, the substrate processing system 1 can detect abnormality of the substrate W after the substrate processing even when the measured value of the surface profile of the substrate W after the substrate processing cannot be obtained.

[0104] In addition, the alignment module 12 is configured to be able to measure the in-plane spectral distribution of the substrate W before substrate processing conveyed from the storage container 16 to the processing module 30 and the in-plane spectral distribution of the substrate W after substrate processing conveyed from the processing module 30 to the storage container 16. The control device 20 is configured to be able to generate an outlier detection model based on the in-plane spectral distribution of the substrate W before substrate processing and the in-plane spectral distribution of the substrate W after substrate processing measured by the alignment module 12, calculate the in-plane score of the substrate W based on the in-plane spectral distribution of the substrate W before substrate processing and the in-plane spectral distribution of the substrate W after substrate processing using the generated outlier detection model, and detect abnormality of the substrate W based on the calculated score. Thus, the substrate processing system 1 can detect abnormality of the substrate W even when the measured value of the surface profile of the substrate W before and after substrate processing cannot be obtained.

[0105] In addition, the control device 20 is configured to calculate a difference spectral distribution indicating the difference between the spectral distribution within the surface of the substrate W before substrate processing and the spectral distribution within the surface of the substrate W after substrate processing measured by the alignment module 12, generate an outlier detection model based on the calculated difference spectral distribution, calculate a score within the surface of the substrate W based on the difference spectral distribution using the generated outlier detection model, and detect an abnormality of the substrate W based on the calculated score. Thus, the substrate processing system 1 can detect an abnormality in substrate processing performed on the substrate W even when a measured value of a surface profile of the substrate W that is changed due to substrate processing cannot be obtained.

[0106] In addition, the control device 20 is configured to generate an outlier detection model based on the spectral distribution of the partial region within the surface of the substrate W. Thus, the substrate processing system 1 can calculate the degree of deviation of the spectral distribution based on the spectral distribution of the partial region within the surface of the substrate W as a score using the generated outlier detection model.

[0107] In addition, the control device 20 is configured to generate an outlier detection model based on the spectral distribution of the center portion within the surface of the substrate W. Thus, the substrate processing system 1 can calculate the degree of deviation of the spectral distribution based on the spectral distribution of the center portion within the surface of the substrate W as a score using the generated outlier detection model.

[0108] The control device 20 is configured to detect abnormality of the substrate W based on the difference between the score of the center portion of the substrate W and the score of the peripheral portion of the substrate W. Thus, the substrate processing system 1 can detect abnormality of the substrate W based on the difference in the contour of the center portion and the peripheral portion of the substrate W.

[0109] In addition, the control device 20 is configured to adjust the processing conditions of the substrate W in the processing module 30 according to the score of the substrate W in which the abnormality is detected. In this way, the substrate processing system 1 can improve the uniformity of the profile within the surface of the substrate W in which the abnormality is detected by performing substrate processing on the substrate W in which the abnormality is detected under the adjusted processing conditions.

[0110] In addition, it should be understood that the embodiments of the present invention are illustrative and non-restrictive in all aspects. In fact, the above-mentioned embodiments can be implemented in many ways. In addition, the above-mentioned embodiments can also be omitted, replaced, and changed in various ways without departing from the claims and their subject matter.

[0111] In addition, the following supplementary notes are also described with respect to the above-mentioned embodiment.

[0112] (Note 1)

[0113] A substrate processing system includes: a loading port configured to be connected to a storage container for storing substrates;

[0114] a processing chamber configured to perform substrate processing on the substrate;

[0115] a measuring unit provided in a path for conveying the substrate between the storage container and the processing chamber, and configured to measure a spectrum distribution within a plane of the substrate; and

[0116] A control unit is configured to generate an outlier detection model based on the spectral distribution within the surface of the substrate measured by the measuring unit, calculate the score within the surface of the substrate according to the measured spectral distribution within the surface of the substrate using the generated outlier detection model, and detect abnormalities of the substrate based on the calculated score.

[0117] (Note 2)

[0118] The substrate processing system according to Supplement 1, wherein:

[0119] The measuring unit is configured to measure a spectrum distribution within a surface of the substrate before being processed and transported from the storage container to the processing chamber.

[0120] The control unit is configured to generate an outlier detection model based on the measured spectral distribution within the surface of the substrate before substrate processing, use the generated outlier detection model to calculate the score within the surface of the substrate according to the measured spectral distribution within the surface of the substrate before substrate processing, and detect abnormalities of the substrate before substrate processing based on the calculated score.

[0121] (Note 3)

[0122] The substrate processing system according to Supplement 1, wherein:

[0123] The measuring unit is configured to measure a spectrum distribution within a surface of the substrate after the substrate is processed and transported from the processing chamber to the storage container.

[0124] The control unit is configured to generate an outlier detection model based on the measured spectral distribution within the surface of the substrate after substrate processing, use the generated outlier detection model to calculate the score within the surface of the substrate according to the measured spectral distribution within the surface of the substrate after substrate processing, and detect abnormalities of the substrate after substrate processing based on the calculated score.

[0125] (Note 4)

[0126] The substrate processing system according to Supplement 1, wherein:

[0127] The measuring unit is configured to measure the spectrum distribution within the surface of the substrate before being processed and transported from the storage container to the processing chamber, and the spectrum distribution within the surface of the substrate after being processed and transported from the processing chamber to the storage container.

[0128] The control unit is configured to generate an outlier detection model based on the spectral distribution within the surface of the substrate before substrate processing and the spectral distribution within the surface of the substrate after substrate processing measured by the measuring unit, respectively; use the generated outlier detection model to calculate the score within the surface of the substrate according to the spectral distribution within the surface of the substrate before substrate processing and the spectral distribution within the surface of the substrate after substrate processing; and detect abnormalities of the substrate based on the calculated score.

[0129] (Note 5)

[0130] The substrate processing system according to Supplement 4, wherein:

[0131] The control unit is configured to calculate a differential spectral distribution indicating the difference between the spectral distribution within the surface of the substrate before substrate processing and the spectral distribution within the surface of the substrate after substrate processing measured by the measuring unit, generate an outlier detection model based on the calculated differential spectral distribution, calculate a score within the surface of the substrate according to the differential spectral distribution using the generated outlier detection model, and detect abnormalities of the substrate based on the calculated score.

[0132] (Note 6)

[0133] A substrate processing system according to any one of Supplementary Notes 1 to 5, wherein:

[0134] The control unit is configured to generate an outlier detection model based on a spectral distribution of a partial region within a plane of the substrate.

[0135] (Note 7)

[0136] A substrate processing system according to any one of Supplementary Notes 1 to 5, wherein:

[0137] The control unit is configured to generate an outlier detection model based on a spectral distribution of a central portion within a plane of the substrate.

[0138] (Note 8)

[0139] A substrate processing system according to any one of Supplementary Notes 1 to 7, wherein:

[0140] The control unit is configured to be capable of detecting an abnormality of the substrate based on a difference between a score of a central portion of the substrate and a score of a peripheral portion of the substrate.

[0141] (Note 9)

[0142] A substrate processing system according to any one of Supplementary Notes 1 to 8, wherein:

[0143] The control unit is configured to adjust processing conditions of the substrate in the processing chamber according to the score of the substrate in which the abnormality is detected.

[0144] (Note 10)

[0145] A substrate processing system according to any one of Supplementary Notes 1 to 9, wherein:

[0146] The outlier detection model is any one of a LOF (Local Outlier Factor) model, a k-nearest neighbor method, and a One class SVM (Support Vector Machine).

[0147] (Note 11)

[0148] A substrate abnormality detection method, comprising:

[0149] The step of measuring the spectrum distribution in the plane of the substrate by using a measuring unit provided in a conveying path of the substrate, wherein the conveying path is a path for conveying the substrate between a storage container arranged at a loading port for storing the substrate and a processing chamber configured to perform substrate processing on the substrate;

[0150] The step of generating an outlier detection model based on the measured spectral distribution in the plane of the substrate;

[0151] The step of calculating a score within the plane of the substrate according to the measured spectral distribution within the plane of the substrate using the generated outlier detection model; and

[0152] The step of detecting an abnormality of the substrate based on the calculated score.

[0153] Description of Reference Numerals

[0154] G2~G4 Gate Valve

[0155] W substrate

[0156] 1. Substrate handling system

[0157] 10 Device body

[0158] 11 Vacuum conveying module

[0159] 110 conveyor robot

[0160] 12 Alignment module

[0161] 13 Load Lock Module

[0162] 14 Atmospheric transport module

[0163] 140 conveyor robot

[0164] 15, 15a, 15b, 15d loading ports

[0165] 20 Control device

[0166] 20a Computer

[0167] 20a1 Processing Department

[0168] 20a2 Storage unit

[0169] 20a3 Communication interface

[0170] 30, 30a processing module

[0171] 50 Stage

[0172] 51 Light Source

[0173] 53. Inspection Department

[0174] 60 Silicon substrate

[0175] 61 Oxide layer

[0176] 62 Nitride layer

[0177] 63 Polysilicon layer

[0178] 64 Mask layer

[0179] 65 holes

[0180] 66, 66a, 66a1, 66b measuring points.

Claims

1. A substrate processing system, characterized in that: include: A loading port configured to be connectable to a storage container for storing substrates; a processing chamber configured to perform substrate processing on the substrate; a measuring unit provided in a path for conveying the substrate between the storage container and the processing chamber, and configured to measure a spectrum distribution within a plane of the substrate; and A control unit is configured to generate an outlier detection model based on the spectral distribution within the surface of the substrate measured by the measuring unit, calculate the score within the surface of the substrate according to the measured spectral distribution within the surface of the substrate using the generated outlier detection model, and detect abnormalities of the substrate based on the calculated score.

2. The substrate processing system according to claim 1, wherein: The measuring unit is configured to measure a spectrum distribution within a surface of the substrate before being processed and transported from the storage container to the processing chamber. The control unit is configured to generate an outlier detection model based on the measured spectral distribution within the surface of the substrate before substrate processing, use the generated outlier detection model to calculate the score within the surface of the substrate according to the measured spectral distribution within the surface of the substrate before substrate processing, and detect abnormalities of the substrate before substrate processing based on the calculated score.

3. The substrate processing system according to claim 1, wherein: The measuring unit is configured to measure a spectrum distribution within a surface of the substrate after the substrate is processed and transported from the processing chamber to the storage container. The control unit is configured to generate an outlier detection model based on the measured spectral distribution within the surface of the substrate after substrate processing, use the generated outlier detection model to calculate the score within the surface of the substrate according to the measured spectral distribution within the surface of the substrate after substrate processing, and detect abnormalities of the substrate after substrate processing based on the calculated score.

4. The substrate processing system according to claim 1, wherein: The measuring unit is configured to measure the spectral distribution within the surface of the substrate before being processed and transported from the storage container to the processing chamber, and the spectral distribution within the surface of the substrate after being processed and transported from the processing chamber to the storage container. The control unit is configured to generate an outlier detection model based on the spectral distribution within the surface of the substrate before substrate processing and the spectral distribution within the surface of the substrate after substrate processing measured by the measuring unit, calculate the score within the surface of the substrate based on the spectral distribution within the surface of the substrate before substrate processing and the spectral distribution within the surface of the substrate after substrate processing using the generated outlier detection model, and detect abnormalities of the substrate based on the calculated score.

5. The substrate processing system according to claim 4, wherein: The control unit is configured to calculate a differential spectral distribution indicating the difference between the spectral distribution within the surface of the substrate before substrate processing measured by the measuring unit and the spectral distribution within the surface of the substrate after substrate processing, generate an outlier detection model based on the calculated differential spectral distribution, calculate a score within the surface of the substrate according to the differential spectral distribution using the generated outlier detection model, and detect abnormalities of the substrate based on the calculated score.

6. The substrate processing system according to claim 1, wherein: The control unit is configured to generate an outlier detection model based on a spectral distribution of a partial region within a plane of the substrate.

7. The substrate processing system according to claim 1, wherein: The control unit is configured to generate an outlier detection model based on a spectral distribution of a central portion within a plane of the substrate.

8. The substrate processing system according to claim 1, wherein: The control unit is configured to be capable of detecting an abnormality of the substrate based on a difference between a score of a central portion of the substrate and a score of a peripheral portion of the substrate.

9. The substrate processing system according to claim 1, wherein: The control unit is configured to adjust processing conditions of the substrate in the processing chamber according to the score of the substrate in which the abnormality is detected.

10. The substrate processing system according to claim 1, wherein: The outlier detection model is any one of a local anomaly factor model, a k-nearest neighbor method, and a first-class support vector machine.

11. A method for detecting substrate abnormality, characterized in that: include: The step of measuring the spectrum distribution in the plane of the substrate by using a measuring unit provided in a conveying path of the substrate, wherein the conveying path is a path for conveying the substrate between a storage container arranged at a loading port for storing the substrate and a processing chamber configured to perform substrate processing on the substrate; The step of generating an outlier detection model based on the measured spectral distribution in the plane of the substrate; The step of calculating a score within the plane of the substrate according to the measured spectral distribution within the plane of the substrate using the generated outlier detection model; and The step of detecting an abnormality of the substrate based on the calculated score.

Images