Substrate processing apparatus and substrate processing system

Abstract

The substrate processing method includes a first step to a sixth step. A substrate is held (first step). Shooting by a camera is started, and a captured image is generated (second step). The discharge of the processing liquid from the first nozzle to the substrate is started (third step). The discharge of the processing liquid from the first nozzle is stopped, and the discharge of the processing liquid from the second nozzle is started (fourth step). On the basis of image processing for the captured image, the time difference between the start time at which the processing liquid is ejected from the second nozzle and the stop time at which the processing liquid is ejected from the first nozzle is calculated (fifth step). When it is determined that the time difference is outside the predetermined range, at least one of the start time and the stop time is adjusted so that the time difference is within the predetermined range (sixth step).

Description

[0001] This application is a divisional application with a filing date of July 4, 2019, application number 2019800548799, and name “Substrate processing method and substrate processing device”. Technical Field

[0002] The invention relates to a substrate processing method, a substrate processing device and a substrate processing system. Background Art

[0003] A substrate processing device has been proposed for processing a substrate by sequentially supplying different processing liquids to the substrate. The substrate processing device includes a substrate holding portion for holding the substrate in a horizontal posture, a rotating mechanism for rotating the substrate holding portion to rotate the substrate in a horizontal plane, and a first spray nozzle and a second spray nozzle for spraying the processing liquid from above the substrate. When viewed from above, the first spray nozzle and the second spray nozzle spray the processing liquid from near the center of the substrate, respectively.

[0004] When the first treatment liquid is sprayed from the first spray nozzle toward the substrate, the first treatment liquid lands near the center of the substrate, is spread on the substrate by the centrifugal force accompanying the rotation of the substrate, and scatters from the periphery of the substrate. The first treatment liquid can use SC1 liquid (a mixture of ammonia water, hydrogen peroxide water, and water), SC2 liquid (a mixture of hydrochloric acid, hydrogen peroxide water, and water), DHF liquid (dilute hydrofluoric acid), etc. Thus, the substrate can be treated with the first treatment liquid.

[0005] Next, a treatment using a second treatment liquid is performed. That is, the nozzle for spraying the treatment liquid is switched from the first spray nozzle to the second spray nozzle. Specifically, the spraying of the first treatment liquid from the first spray nozzle is stopped, and the spraying of the second treatment liquid from the second spray nozzle is started. The second treatment liquid lands near the center of the substrate, and is spread on the substrate by the centrifugal force accompanying the rotation of the substrate, and scatters from the periphery of the substrate. As the second treatment liquid, pure water can be used, for example. In this way, the first treatment liquid can be rinsed from the substrate.

[0006] In addition, a technology for monitoring the discharge state of the processing liquid discharged from the discharge nozzle by using a camera has been proposed (e.g., Patent Document 1 and Patent Document 2). In Patent Document 1 and Patent Document 2, a camera is used to capture a captured area including the front end of the discharge nozzle, and based on the captured image of the camera, it is determined whether the processing liquid is being discharged from the discharge nozzle.

[0007] Prior art literature

[0008] Patent Literature

[0009] Patent Document 1: Japanese Patent Application Publication No. 2017-29883

[0010] Patent Document 2: Japanese Patent Application Publication No. 2015-173148 Summary of the invention

[0011] Problem that the invention aims to solve

[0012] When the nozzle ejecting the processing liquid is switched from the first ejection nozzle to the second ejection nozzle, the time difference between the timing of stopping ejection of the processing liquid from the first ejection nozzle and the timing of starting ejection of the processing liquid from the second ejection nozzle is important.

[0013] For example, the later the start time of spraying the processing liquid from the second spray nozzle is relative to the stop time of spraying from the first spray nozzle, the easier it is for the surface of the substrate to dry locally (liquid dries up). If the surface of the substrate is dry, it may cause problems (such as adhesion of particles).

[0014] In order to avoid the drying of the substrate, it is sufficient to start the ejection of the second ejection nozzle before the ejection of the first ejection nozzle stops. For example, during the period when the ejection of the first treatment liquid stops, the ejection amount of the first treatment liquid decreases over time and eventually becomes zero. By starting the ejection of the second treatment liquid during the period when the ejection of the first treatment liquid stops, the treatment liquid can be continuously supplied to the substrate, which can reduce the possibility of drying of the substrate.

[0015] However, if the second ejection nozzle starts ejecting too early, the second processing liquid will start ejecting while the first processing liquid is still ejecting in a sufficient amount. At this time, the total amount of processing liquid supplied to the substrate increases, and the processing liquid splashes back on the substrate. Such splashing is not preferred.

[0016] Therefore, it is preferable to adjust the time difference so as not to cause splashing of the liquid or drying up of the liquid.

[0017] In addition, the time difference may affect whether the substrate is processed well. In this case, it is preferable to adjust the time difference so that the processing result becomes good.

[0018] Therefore, an object of the present application is to provide a substrate processing method, a substrate processing apparatus, and a substrate processing system that can switch from a first nozzle to a second nozzle with a desired time difference.

[0019] Means used to solve problems

[0020] A first form of a substrate processing method is a substrate processing method comprising: a first step of holding a substrate; a second step of starting to use a camera to photograph a shooting area including a front end of a first nozzle and a front end of a second nozzle, thereby generating a captured image; a third step of starting to spray a processing liquid from the first nozzle to the substrate; a fourth step of stopping spraying the processing liquid from the first nozzle and starting to spray the processing liquid from the second nozzle; a fifth step of calculating, based on image processing of the captured image, a time difference between a start time of starting to spray the processing liquid from the second nozzle and a stop time of stopping spraying the processing liquid from the first nozzle in the fourth step; and a sixth step of determining whether the time difference is outside a specified range, and when determining that the time difference is outside the specified range, adjusting at least one of the start time and the stop time so that the time difference is within the specified range.

[0021] A second aspect of the substrate processing method is, in the substrate processing method of the first aspect, wherein the stop timing is adjusted without adjusting the start timing.

[0022] A third form of the substrate processing method is that, in the substrate processing method of the first form or the second form, in the fifth step, the stop time is determined based on the pixel value of the first ejection determination area extending from the front end of the first nozzle to the ejection direction of the first nozzle in each frame of the captured image, and the start time is determined based on the pixel value of the second ejection determination area extending from the front end of the second nozzle to the ejection direction of the second nozzle in each frame.

[0023] A fourth form of the substrate processing method is that, in the third form of the substrate processing method, the stop time is determined based on a frame in which the statistic of the pixel value of the first ejection determination area is greater than a threshold value and a frame in which the statistic of the first ejection determination area as the next frame of the frame is less than the threshold value, and the start time is determined based on a frame in which the statistic of the pixel value of the second ejection determination area is less than the threshold value and a frame in which the statistic of the first ejection determination area as the next frame of the frame is greater than the threshold value.

[0024] A fifth form of the substrate processing method is that in the fourth form of the substrate processing method, in the sixth step, a graph representing the time variation of the statistic with respect to the first nozzle and the second nozzle is displayed on a user interface, and when an input for an object time is made to the user interface, the object time is adjusted according to the input, and the object time is at least any one of the start time and the stop time.

[0025] A sixth form of the substrate processing method is that, in any one form of the substrate processing method from the first form to the third form, in the fifth step, a classifier based on machine learning is used to classify, for each nozzle of the first nozzle and the second nozzle, each frame contained in the captured image as ejection or stop of the processing liquid, and the time difference is calculated based on the classification result.

[0026] A seventh form of the substrate processing method is that, in the sixth form of the substrate processing method, the stop time is determined based on a frame in which the first nozzle is classified as ejecting and a frame in which the first nozzle is classified as stopping as the next frame of the frame, and the start time is determined based on a frame in which the second nozzle is classified as stopping and a frame in which the second nozzle is classified as ejecting as the next frame of the frame.

[0027] An eighth form of the substrate processing method is that, in the sixth form of the substrate processing method, the stop time is after the start time, and the time difference is calculated based on the number of frames classified as frames in which both the first nozzle and the second nozzle spray processing liquid and the time between the frames.

[0028] A ninth aspect of the substrate processing method is that, in any one of the first to eighth aspects, in the sixth step, when it is determined that the time difference is outside a prescribed range, the fact that the time difference is outside the prescribed range is notified to the operator.

[0029] The tenth form of the substrate processing method is that in any one form of the substrate processing method from the first form to the ninth form, the stop time is after the start time, and the substrate processing method also includes a seventh step, and the seventh step determines whether splashing of the processing liquid on the substrate is generated based on image processing of the captured image, and when it is determined that the splashing is generated, adjusts at least either the start time or the stop time in a manner that reduces the time difference between the start time and the stop time.

[0030] An eleventh aspect of the substrate processing method is, in the tenth aspect of the substrate processing method, wherein in the seventh step, a classifier obtained through machine learning is used to classify each frame of the captured image as having the splashed liquid or not having the splashed liquid.

[0031] The twelfth form of the substrate processing method is that in the eleventh form of the substrate processing method, in the seventh step, the splash judgment area near the first nozzle and the second nozzle in each frame of the captured image is cut out, and the classifier is used to classify the splash judgment area into whether there is splash or not or whether there is no splash.

[0032] The thirteenth form of the substrate processing method is, in any one of the sixth to eighth forms, the eleventh form, and the twelfth form, a classifier is selected from a plurality of machine-learned classifiers corresponding to at least any one of the type of the substrate, the type of the processing liquid, the positions of the first nozzle and the second nozzle, and the flow rate of the processing liquid, and each frame contained in the captured image is classified based on the selected classifier.

[0033] The fourteenth form of the substrate processing method is that in the thirteenth form of the substrate processing method, when at least any one of the type of the substrate, the type of the processing liquid, the positions of the first nozzle and the second nozzle, and the flow rate of the processing liquid is input into the input part, a classifier is selected from the plurality of classifiers based on the input to the input part.

[0034] A first form of a substrate processing device comprises: a substrate holding portion for holding a substrate, a processing liquid supply portion having a first nozzle for spraying processing liquid onto the substrate and a second nozzle for spraying processing liquid onto the substrate, a camera for photographing a photographing area including a front end of the first nozzle and a front end of the second nozzle to generate a photographed image, and a control portion; the control portion controls the processing liquid supply portion in such a manner that, after starting to spray processing liquid onto the substrate from the first nozzle, the processing liquid starts to spray from the second nozzle onto the substrate and stops spraying processing liquid onto the substrate from the first nozzle, and calculates a time difference between a start time of starting to spray processing liquid from the second nozzle and a stop time of stopping to spray processing liquid from the first nozzle based on image processing of the photographed image, and when it is determined that the time difference is outside a specified range, adjusts at least either the start time or the stop time in such a manner that the time difference is within the specified range.

[0035] A second form of the substrate processing device is that in the first form of the substrate processing device, the control unit uses a classifier based on machine learning to classify, for the first nozzle and the second nozzle, each frame contained in the captured image into a category indicating a state of ejection or stop of the processing liquid, and calculates the time difference based on the classification result.

[0036] A third form of the substrate processing device is that, in the substrate processing device of the second form, the control unit selects a classifier from a plurality of machine-learned classifiers corresponding to at least any one of the type of the substrate, the type of the processing liquid, the positions of the first nozzle and the second nozzle, and the flow rate of the processing liquid, and, based on the selected classifier, classifies each frame contained in the captured image.

[0037] A fourth form of the substrate processing device is that, in the third form of the substrate processing device, there is an input unit, wherein the input unit is used to input at least any one of the type of the substrate, the type of the processing liquid, the positions of the first nozzle and the second nozzle, and the flow rate of the processing liquid, and the control unit selects a classifier from the plurality of classifiers based on the input to the input unit.

[0038] The substrate processing system has a substrate processing device and a server communicating with the substrate processing device, wherein the substrate processing device has: a substrate holding portion for holding a substrate, a processing liquid supply portion having a first nozzle for spraying processing liquid onto the substrate and a second nozzle for spraying processing liquid onto the substrate, a camera for photographing a photographing area including a front end of the first nozzle and a front end of the second nozzle to generate a photographed image, and a control portion for starting to spray processing liquid from the second nozzle onto the substrate after starting to spray processing liquid from the first nozzle onto the substrate, and stopping spraying processing liquid from the first nozzle onto the substrate The processing liquid supply unit is controlled in a manner; the substrate processing device and the server use a classifier based on machine learning to classify, for the first nozzle and the second nozzle, each frame contained in the captured image into a category representing a state of ejecting or stopping the processing liquid, and based on the classification result, calculate the time difference between the start time of starting to eject the processing liquid from the second nozzle and the stop time of stopping ejecting the processing liquid from the first nozzle. When the control unit determines that the time difference is outside a specified range, the control unit adjusts at least either the start time or the stop time in a manner that makes the time difference fall within the specified range.

[0039] Effects of the Invention

[0040] According to the first aspect of the substrate processing method and the first aspect of the substrate processing apparatus, the time difference between the start time and the stop time is obtained based on image processing of the captured image, so that the time difference can be obtained with high accuracy. Therefore, the time difference can be adjusted to a specified range with high accuracy.

[0041] As a result, for example, it is possible to substantially avoid splashing of the processing liquid on the substrate and liquid drying up in which the substrate is partially dried.

[0042] According to the second aspect of the substrate processing method, the time difference can be set within a predetermined range without changing the length of the processing period during which the processing liquid is supplied from the first nozzle.

[0043] According to the third aspect of the substrate processing method, since the pixel values ​​of the first discharge determination area and the second discharge determination area are used, the processing can be reduced compared to the case where the image processing is performed on the entire captured image.

[0044] According to the fourth aspect of the substrate processing method, the start time and the stop time can be determined by a simple process.

[0045] According to the fifth aspect of the substrate processing method, the operator can visually check the temporal change of the statistic and adjust the time difference based on the temporal change.

[0046] According to the sixth aspect of the substrate processing method, the second aspect of the substrate processing apparatus, and the aspect of the substrate processing system, the time difference can be obtained with high accuracy through machine learning.

[0047] According to the seventh aspect of the substrate processing method, the time difference can be appropriately determined.

[0048] According to the eighth aspect of the substrate processing method, the time difference can be appropriately determined.

[0049] According to the ninth aspect of the substrate processing method, the operator can recognize that the time difference is outside the predetermined range.

[0050] According to the tenth aspect of the substrate processing method, the time difference can be adjusted so as to reduce the generation of splashing.

[0051] According to the eleventh aspect of the substrate processing method, the presence or absence of liquid splashing can be determined with high accuracy.

[0052] According to the twelfth aspect of the substrate processing method, the classification accuracy can be improved.

[0053] According to the thirteenth aspect of the substrate processing method and the third aspect of the substrate processing apparatus, each frame can be classified with high accuracy.

[0054] According to the fourteenth aspect of the substrate processing method and the fourth aspect of the substrate processing apparatus, the operator can input information into the input unit. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 FIG. 1 is a diagram showing an example of a schematic structure of a substrate processing apparatus.

[0056] Figure 2 This is a plan view schematically showing an example of the structure of the processing unit.

[0057] Figure 3 This is a plan view schematically showing an example of the structure of the processing unit.

[0058] Figure 4 This is a flowchart showing an example of the operation of the control unit.

[0059] Figure 5 FIG. 1 is a diagram schematically showing an example of a frame generated by a camera.

[0060] Figure 6 FIG. 1 is a diagram schematically showing an example of a frame generated by a camera.

[0061] Figure 7 FIG. 1 is a diagram schematically showing an example of a frame generated by a camera.

[0062] Figure 8 FIG. 1 is a diagram schematically showing an example of a frame generated by a camera.

[0063] Fig. 9 1 is a flowchart showing a specific example of the monitoring process.

[0064] Fig.10 This is a graph schematically showing an example of brightness distribution in the discharge determination area.

[0065] Fig.11 This is a diagram schematically showing an example of brightness distribution in the discharge determination area.

[0066] Fig.12 This is a diagram schematically showing an example of temporal changes in statistics of pixel values ​​within the discharge determination area.

[0067] Fig.13 This is a diagram schematically showing an example of temporal changes in statistics of pixel values ​​within the discharge determination area.

[0068] Fig.14 This is a functional block diagram schematically showing an example of the internal structure of the control unit.

[0069] Fig.15 1 is a flowchart showing a specific example of the monitoring process.

[0070] Fig.16 This is a diagram schematically showing an example of a plurality of frames of captured images.

[0071] Fig.17 This is a functional block diagram schematically showing an example of the internal structure of the control unit.

[0072] Fig.18 This is a diagram schematically showing an example of an input screen.

[0073] Fig.19 This is a functional block diagram schematically showing an example of a substrate processing apparatus and a server.

[0074] Fig. 20 This is a diagram schematically showing an example of a deep learning model.

[0075] Fig.21FIG. 1 is a diagram schematically showing an example of a frame generated by a camera.

[0076] Fig. 22 This is a diagram schematically showing an example of temporal changes in statistics of pixel values ​​within the liquid splash determination area.

[0077] Fig.23 1 is a flowchart showing a specific example of the monitoring process.

[0078] Fig.24 This is a plan view schematically showing an example of the structure of the processing unit.

[0079] Fig.25 FIG. 1 is a diagram schematically showing an example of a frame generated by a camera. DETAILED DESCRIPTION

[0080] Below, while referring to the attached Figure 1 The embodiments are described below. In addition, the drawings are schematically shown, and for the convenience of description, the structure is appropriately omitted or simplified. In addition, the size and position of the structures shown in the drawings are not necessarily accurately described and can be appropriately changed.

[0081] In the following description, the same components are denoted by the same reference numerals, and the names and functions of these components are also the same. Therefore, in order to avoid duplication, the detailed description of these components may be omitted.

[0082] First implementation method.

[0083] <Overview of substrate processing apparatus>

[0084] Figure 1 1 is a diagram showing the overall structure of a substrate processing apparatus 100. The substrate processing apparatus 100 is an apparatus for processing a substrate W by supplying a processing liquid to the substrate W. The substrate W is, for example, a semiconductor substrate. The substrate W has a substantially disk shape.

[0085] The substrate processing device 100 can supply at least two processing liquids to the substrate W in sequence. For example, the substrate processing device 100 can supply a rinsing liquid to the substrate W after supplying a cleaning liquid to the substrate W, thereby performing a cleaning process. The liquid can typically use SC1 liquid (a mixture of ammonia water, hydrogen peroxide water and water), SC2 liquid (a mixture of hydrochloric acid, hydrogen peroxide water and water), DHF liquid (dilute hydrofluoric acid), etc. As the rinsing liquid, for example, pure water can be used. In this specification, the liquid and the rinsing liquid are collectively referred to as "processing liquid". In addition, not only the cleaning process, but also coating liquids such as photoresist liquid used for film forming processing, liquids used to remove unnecessary films, liquids used for etching, etc. are also included in the "processing liquid".

[0086] The substrate processing device 100 has a indexer 102, a plurality of processing units 1, and a main transport robot 103. The indexer 102 has the function of moving unprocessed substrates W received from outside the device into the device, and moving processed substrates W after the cleaning process out of the device. The indexer 102 carries a plurality of carriers and has a transfer robot (all omitted in the figure). As the carrier, a FOUP (front opening unified pod) or a SMIF (Standard Mechanical Interface) box that stores the substrate W in a closed space, or an OC (open cassette) that exposes the substrate W to the outside air in a stored state can be used. The transfer robot transfers the substrate W between the carrier and the main transport robot 103.

[0087] The substrate processing apparatus 100 is provided with 12 processing units 1. In detail, four towers are arranged to surround the main transport robot 103, and the towers are formed by stacking three processing units 1. In other words, the four processing units 1 arranged to surround the main transport robot 103 are stacked in three stages. Figure 1 One of the layers is shown. In addition, the number of the processing units 1 mounted in the substrate processing apparatus 100 is not limited to 12, and may be 8 or 4, for example.

[0088] The main transport robot 103 is installed in the center of four towers including the stacked processing units 1. The main transport robot 103 carries unprocessed substrates W received from the indexer 102 into each processing unit 1, and carries out processed substrates W from each processing unit 1 and delivers them to the indexer 102.

[0089] Next, the processing unit 1 will be described. Hereinafter, one of the 12 processing units 1 mounted in the substrate processing apparatus 100 will be described, but the same applies to the other processing units 1 . Figure 2 is a top view of the processing unit 1. In addition, Figure 3 is a longitudinal sectional view of the processing unit 1. In addition, Figure 2 2 shows a state where the substrate W is not held by the substrate holding portion 20. Figure 3 The state where the substrate W is held by the substrate holding portion 20 is shown.

[0090] The processing unit 1 has the following parts as main elements in the chamber 10: a substrate holding part 20, which holds the substrate W in a horizontal position (a position in which the normal line of the substrate W is along the vertical direction); three processing liquid supply parts 30, 60, and 65, which are used to supply processing liquid to the upper surface of the substrate W held by the substrate holding part 20; a processing cup 40, which surrounds the substrate holding part 20; and a camera 70, which takes an image of the space above the substrate holding part 20. In addition, a partition plate 15 is provided around the processing cup 40 in the chamber 10 to partition the inner space of the chamber 10 into upper and lower parts.

[0091] The chamber 10 has a side wall 11 extending in the vertical direction, a top wall 12 that closes the upper side of the space surrounded by the side wall 11, and a bottom wall 13 that closes the lower side. The space surrounded by the side wall 11, the top wall 12, and the bottom wall 13 serves as a processing space for the substrate W. In addition, a loading and unloading port for the main transport robot 103 to load and unload the substrate W in and out of the chamber 10 and a gate for opening and closing the loading and unloading port are provided in a portion of the side wall 11 of the chamber 10 (both are omitted in the figure).

[0092] A fan filter unit (FFU) 14 is installed on the top wall 12 of the chamber 10, and is used to further clean the air in the clean room where the substrate processing device 100 is installed and supply it to the processing space in the chamber 10. The fan filter unit 14 has a fan and a filter (such as a HEPA filter (High Efficiency Particulate Air Filter)) for taking in the air in the clean room and sending it to the chamber 10, and forms a downflow of clean air in the processing space in the chamber 10. In order to evenly disperse the clean air supplied from the fan filter unit 14, a punching plate with a plurality of blowing holes may also be provided directly below the top wall 12.

[0093] The substrate holding portion 20 is, for example, a spin chuck. The substrate holding portion 20 has a disk-shaped spin base 21, which is fixed in a horizontal position to the upper end of a rotation shaft 24 extending in a vertical direction. A rotation motor 22 is provided below the spin base 21 to rotate the rotation shaft 24. The rotation motor 22 rotates the spin base 21 in a horizontal plane via the rotation shaft 24. In addition, a cylindrical cover member 23 is provided to surround the rotation motor 22 and the rotation shaft 24.

[0094] The outer diameter of the disk-shaped rotating base 21 is slightly larger than the diameter of the circular substrate W held by the substrate holding portion 20. Therefore, the rotating base 21 has a holding surface 21a facing the entire lower surface of the substrate W to be held.

[0095] A plurality of (four in the present embodiment) chuck pins 26 are vertically provided on the peripheral edge of the holding surface 21a of the rotating base 21. The plurality of chuck pins 26 are arranged at equal intervals along the circumference corresponding to the outer circumference of the circular substrate W (if there are four chuck pins 26 as in the present embodiment, they are arranged at 90° intervals). The plurality of chuck pins 26 are driven in conjunction with a linkage mechanism (not shown) housed in the rotating base 21. The substrate holding portion 20 holds the substrate W by bringing each of the plurality of chuck pins 26 into contact with the outer circumferential end of the substrate W, thereby being able to hold the substrate W above the rotating base 21 in a horizontal posture close to the holding surface 21a (see FIG. 2 ). Figure 3 ), and can make multiple chuck pins 26 each leave the outer peripheral end of the substrate W to release the grip.

[0096] The substrate W can be rotated about the rotation axis CX extending in the vertical direction passing through the center of the substrate W by rotating the rotation shaft 24 by the rotation motor 22 while the substrate holding portion 20 holds the substrate W by gripping it with the plurality of chuck pins 26 .

[0097] The processing liquid supply unit 30 is composed of a nozzle 31 mounted on the front end of a nozzle arm 32 (see Figure 2 The base end of the nozzle arm 32 is fixedly connected to the nozzle base 33. The nozzle base 33 can be rotated around an axis along the vertical direction by a motor (not shown). Figure 2 As shown by arrow AR34 in FIG. 1 , the discharge nozzle 31 moves in an arc shape in the horizontal direction between a processing position above the substrate holding unit 20 and a standby position outside the processing cup 40 .

[0098] The processing liquid supply unit 30 is configured to supply a plurality of processing liquids. Specifically, the processing liquid supply unit 30 includes a plurality of spray nozzles 31. Figure 2 and Figure 3 In the example of FIG. 1 , two ejection nozzles 31a and 31b are shown as the ejection nozzles 31. The ejection nozzles 31a and 31b are fixed to a nozzle base 33 via a nozzle arm 32. Therefore, the ejection nozzles 31a and 31b move synchronously with each other. The ejection nozzles 31a and 31b are arranged adjacent to each other in a horizontal plane.

[0099] like Figure 3As shown in the example, the ejection nozzle 31a is connected to the processing liquid supply source 37a via the pipe 34a, and the ejection nozzle 31b is connected to the processing liquid supply source 37b via the pipe 34b. On-off valves 35a and 35b are provided in the middle of the pipes 34a and 34b, respectively. By opening the on-off valve 35a, the processing liquid Lq1 from the processing liquid supply source 37a flows through the inside of the pipe 34a and is ejected from the ejection nozzle 31a, and by opening the on-off valve 35b, the processing liquid Lq2 from the processing liquid supply source 37b flows through the inside of the pipe 34b and is ejected from the ejection nozzle 31b. For example, SC1 liquid is ejected from the ejection nozzle 31a, and for example, pure water is ejected from the ejection nozzle 31b. The processing liquids Lq1 and Lq2 ejected from the ejection nozzles 31a and 31b in a state of being stopped at the processing position fall on the upper surface of the substrate W held by the substrate holding portion 20.

[0100] Suck back valves 36a and 36b may be provided in the middle of the pipes 34a and 34b, respectively. The suck back valve 36a sucks the processing liquid Lq1 in the pipe 34a when the ejection of the processing liquid Lq1 stops, thereby introducing the processing liquid Lq1 from the front end of the ejection nozzle 31a. As a result, it is difficult for the processing liquid Lq1 to drip from the front end of the ejection nozzle 31a in the form of relatively large lumps (droplets) when the ejection stops. The same applies to the suck back valve 36b.

[0101] In addition, in the processing unit 1 of the present embodiment, in addition to the above-mentioned processing liquid supply unit 30, two processing liquid supply units 60 and 65 are also provided. The processing liquid supply units 60 and 65 of the present embodiment have the same structure as the above-mentioned processing liquid supply unit 30. That is, the processing liquid supply unit 60 is configured by installing a spray nozzle 61 at the front end of a nozzle arm 62, and the spray nozzle 61 moves in an arc shape between the processing position above the substrate holding unit 20 and the standby position outside the processing cup 40 through a nozzle base 63 connected to the base end side of the nozzle arm 62 as shown by an arrow AR64. Similarly, the processing liquid supply unit 65 is configured by installing a spray nozzle 66 at the front end of a nozzle arm 67, and the spray nozzle 66 moves in an arc shape between the processing position above the substrate holding unit 20 and the standby position outside the processing cup 40 as shown by an arrow AR69 through a nozzle base 68 connected to the base end side of the nozzle arm 67. The processing liquid supply parts 60 and 65 may be configured to supply a plurality of processing liquids, or may be configured to supply a single processing liquid.

[0102] The processing liquid supply parts 60 and 65 spray the processing liquid onto the upper surface of the substrate W held by the substrate holding part 20 when their respective spray nozzles 61 and 66 are located at the processing position. In addition, at least one of the processing liquid supply parts 60 and 65 may also be a two-fluid nozzle that mixes a cleaning liquid such as pure water with a pressurized gas to generate droplets and sprays the mixed fluid of the droplets and the gas onto the substrate W. In addition, the processing liquid supply parts provided in the processing unit 1 are not limited to three, as long as there are more than one. However, in the present embodiment, since it is based on the premise of switching and spraying two processing liquids in sequence, more than two spray nozzles are generally provided. Each spray nozzle of the processing liquid supply parts 60 and 65 can also be connected to the processing liquid supply source via a pipe in the same way as the processing liquid supply part 30, and an opening and closing valve is provided in the middle of the pipe, and a back-sucking valve is further provided. Below, the processing using the processing liquid supply part 30 is described representatively.

[0103] The processing cup 40 is arranged in a manner to surround the substrate holding portion 20. The processing cup 40 has an inner cup 41, a middle cup 42, and an outer cup 43. The inner cup 41, the middle cup 42, and the outer cup 43 are arranged in a manner that can be raised and lowered. When the inner cup 41, the middle cup 42, and the outer cup 43 are in a state of being raised, the processing liquid scattered from the periphery of the substrate W hits the inner circumferential surface of the inner cup 41 and falls. The fallen processing liquid is appropriately recovered by the first recovery mechanism (not shown). When the inner cup 41 is lowered and the middle cup 42 and the outer cup 43 are raised, the processing liquid scattered from the periphery of the substrate W hits the inner circumferential surface of the middle cup 42 and falls. The fallen processing liquid is appropriately recovered by the second recovery mechanism (not shown). When the inner cup 41 and the middle cup 42 are lowered and the outer cup 43 is raised, the processing liquid scattered from the periphery of the substrate W hits the inner circumferential surface of the outer cup 43 and falls. The fallen processing liquid is appropriately recovered by the third recovery mechanism (not shown). Thereby, different processing liquids can be respectively recovered appropriately.

[0104] The partition plate 15 is provided around the processing cup 40 so as to partition the inner space of the chamber 10 up and down. The partition plate 15 may be a single plate-shaped member surrounding the processing cup 40, or may be formed by joining a plurality of plate-shaped members. In addition, a through hole or a notch penetrating in the thickness direction may be formed on the partition plate 15. In the present embodiment, a through hole (not shown) for inserting a support shaft is formed, and the support shaft is used to support the nozzle bases 33, 63, 68 of the processing liquid supply parts 30, 60, 65.

[0105] The outer peripheral end of the partition plate 15 is connected to the side wall 11 of the chamber 10. In addition, the edge of the partition plate 15 surrounding the processing cup 40 is formed in a circular shape with a diameter larger than the outer diameter of the outer cup 43. Therefore, the partition plate 15 does not hinder the lifting and lowering of the outer cup 43.

[0106] In addition, an exhaust duct 18 is provided at a part of the side wall 11 of the chamber 10 and near the bottom wall 13. The exhaust duct 18 is connected to an exhaust mechanism (not shown). The clean air supplied from the fan filter unit 14 and flowing down in the chamber 10, the air passing between the processing cup 40 and the partition plate 15 is discharged from the exhaust duct 18 to the outside of the device.

[0107] The camera 70 is disposed in the chamber 10 above the partition plate 15. The camera 70 includes, for example, an optical system such as a photographing element (e.g., a CCD (Charge Coupled Device)), an electronic shutter, and a lens. The ejection nozzle 31 of the processing liquid supply unit 30 is disposed at a processing position ( Figure 3 The solid line position) and the standby position ( Figure 3 The processing position is a position where the processing liquid is sprayed from the processing liquid supply unit 30 onto the upper surface of the substrate W held by the substrate holding unit 20 for cleaning. The standby position is a position where the processing liquid supply unit 30 stops spraying the processing liquid and stands by when no cleaning is performed. At the standby position, a standby box for storing the spray nozzle 31 of the processing liquid supply unit 30 may also be provided.

[0108] The camera 70 is set in such a manner that at least the front end of the ejection nozzle 31 at the processing position is included in its photographing area. More specifically, the camera 70 is set in such a manner that the front end of the ejection nozzle 31 and the processing liquid ejected from the front end are included in the photographing area. Figure 3 As shown in FIG. 1 , the camera 70 is set at a position to shoot the ejection nozzle 31 at the processing position from the front and top. Therefore, the camera 70 can shoot the shooting area including the front end of the ejection nozzle 31 at the processing position. Similarly, the camera 70 can also shoot the shooting area including the front end of the ejection nozzle 61, 66 of the processing liquid supply part 60, 65 at the processing position. In addition, when the camera 70 is set at Figure 2 When the position shown in FIG. 1 is shown, the ejection nozzles 31 and 66 of the processing liquid supply unit 30 and 65 move laterally within the field of view of the camera 70, so the movement near the processing position can be properly photographed, while the ejection nozzle 61 of the processing liquid supply unit 60 moves in the depth direction within the field of view of the camera 70, so the movement near the processing position may not be properly photographed. In this case, a camera dedicated to the processing liquid supply unit 60 may be provided separately from the camera 70.

[0109] In addition, if Figure 3As shown, an illumination unit 71 is provided in the chamber 10 and above the partition plate 15. Usually, the chamber 10 is a dark room, so when the camera 70 takes pictures, the illumination unit 71 irradiates light to the ejection nozzles 31, 61, 66 of the processing liquid supply units 30, 60, 65 near the processing position. The captured image generated by the camera 70 is output to the control unit 9.

[0110] The control unit 9 controls various structures of the substrate processing apparatus 100 to process the substrate W. In addition, the control unit 9 performs image processing on the captured image generated by the camera 70. The control unit 9 obtains the time difference between the start time and the stop time of ejecting the processing liquid from each ejection nozzle through the image processing. The image processing will be described in detail below.

[0111] The hardware structure of the control unit 9 is the same as that of a general computer. That is, the control unit 9 includes: a CPU (Central Processing Unit) for performing various calculations, a read-only memory for storing basic programs, namely a ROM (Read Only Memory), a freely readable and writable memory for storing various information, namely a RAM (Random Access Memory), and a disk that pre-stores control software or data. The CPU of the control unit 9 executes a predetermined processing program, so that each action mechanism of the substrate processing device 100 is controlled by the control unit 9, and the substrate processing device 100 is processed. In addition, the CPU of the control unit 9 executes a predetermined processing program to perform image processing. In addition, part or all of the functions of the control unit 9 can also be realized by dedicated hardware.

[0112] The user interface 90 has a display and an input unit. The display is, for example, a liquid crystal display or an organic EL (ElectroLuminescence) display. The input unit is, for example, a touch panel, a mouse, or a keyboard. The user interface 90 is connected to the control unit 9. The display displays an image based on a display signal from the control unit 9. The displayed image includes, for example, an image captured by the camera 70. The input unit outputs input information input by the user to the control unit 9. The control unit 9 can control various structures based on the input information.

[0113] <Operation of the control unit>

[0114] Figure 4 1 is a flowchart showing an example of the operation of the control unit 9. Here, as an example, the process using the processing liquid supply unit 30 is described. First, in step S1, the main transfer robot 103 transfers the substrate W to the substrate holding unit 20. The substrate holding unit 20 holds the transferred substrate W.

[0115] Next, in step S2 , the control unit 9 rotates the nozzle base 33 to move the ejection nozzles 31 a and 31 b to the processing position. When the ejection nozzles 31 a and 31 b are stopped at the processing position, the tip of the ejection nozzle 31 a and the tip of the ejection nozzle 31 b are included in the imaging area of ​​the camera 70 .

[0116] Next, in step S3, the control unit 9 controls the camera 70 to start photographing. As a result, the camera 70 can more reliably photograph the front end of the ejection nozzle 31a and the front end of the ejection nozzle 31b. The camera 70 photographs the photographing area at a predetermined frame rate (e.g., 60 frames / second), and sequentially outputs each frame of the generated photographed image to the control unit 9. In addition, the photographing of the camera 70 may be started when the movement of the ejection nozzles 31a and 31b in step S2 is started.

[0117] Figure 5 FIG. 1 is a diagram schematically showing an example of a frame IM1 of a captured image generated by the camera 70. Figure 5 In the illustrated frame IM1, the tip of the discharge nozzle 31a and the tip of the discharge nozzle 31b are captured, and a portion of the substrate W is also captured. In the frame IM1, the processing liquid Lq1 has not yet been discharged from the discharge nozzle 31a, and similarly, the processing liquid Lq2 has not yet been discharged from the discharge nozzle 31b.

[0118] Next, in step S4, the control unit 9 starts the ejection from the ejection nozzle 31a. Specifically, the control unit 9 outputs an opening signal to the on-off valve 35a. The on-off valve 35a performs an opening action based on the opening signal to open the pipe 34a. As a result, the processing liquid from the processing liquid supply source 37a is ejected from the ejection nozzle 31a and supplied to the upper surface of the substrate W. In addition, a delay time is generated from the output of the opening signal to the actual ejection of the processing liquid Lq1. The delay time depends on various factors such as the moving speed of the valve body caused by the opening action of the on-off valve 35a, the pipe length of the pipe 34a, and the pressure loss.

[0119] In addition, the control unit 9 rotates the rotation motor 22 to rotate the substrate W immediately before step S4.

[0120] Figure 6 FIG. 1 is a diagram schematically showing an example of a frame IM2 of a captured image generated by the camera 70. Figure 6In the illustrated frame IM2, the processing liquid Lq1 is ejected from the ejection nozzle 31a, and the processing liquid Lq2 is not ejected from the ejection nozzle 31b. The processing liquid Lq1 ejected from the ejection nozzle 31a is a so-called continuous flow, and has a liquid column shape extending in the vertical direction in the area from the front end of the ejection nozzle 31a to the upper surface of the substrate W. The processing liquid Lq1 lands on the approximate center of the substrate W, and is spread on the upper surface of the substrate W by the centrifugal force accompanying the rotation of the substrate W. Then, it scatters from the periphery of the substrate W. As a result, the processing liquid Lq1 acts on the entire upper surface of the substrate W, and the processing based on the processing liquid Lq1 is performed.

[0121] For example, after a predetermined time has passed since step S4, the control unit 9 switches the nozzle that ejects the treatment liquid from the ejection nozzle 31a to the ejection nozzle 31b in step S5. That is, the control unit 9 stops ejecting the treatment liquid Lq1 from the ejection nozzle 31a, and starts ejecting the treatment liquid Lq2 from the ejection nozzle 31b. That is, the control unit 9 sends a closing signal to the on-off valve 35a and sends an opening signal to the on-off valve 35b. As a specific example, the control unit 9 outputs an opening signal to the on-off valve 35b when the elapsed time from step S4 reaches a first reference time, and outputs a closing signal to the on-off valve 35a when the elapsed time from step S4 reaches a second reference time. For example, the second reference time can be set to be longer than the first reference time.

[0122] The on-off valve 35b performs an opening action based on the opening signal to open the pipe 34b. As a result, the processing liquid Lq2 from the processing liquid supply source 37b is ejected from the ejection nozzle 31b and lands on the upper surface of the substrate W. In addition, a delay time occurs from the output of the opening signal to the actual ejection of the processing liquid Lq2. The delay time depends on various factors such as the moving speed of the valve body caused by the opening action of the on-off valve 35b, the pipe length of the pipe 34b, and the pressure loss.

[0123] The on-off valve 35a performs a closing action based on the closing signal to close the pipe 34a. In addition, when a back-sucking valve 36a is provided, the control unit 9 sends a suction signal to the back-sucking valve 36a. The back-sucking valve 36a performs a suction action based on the suction signal to suck the processing liquid in the pipe 34a. The closing action of the on-off valve 35a and the suction action of the back-sucking valve 36a are performed simultaneously with each other. As a result, the processing liquid Lq1 on the front end side of the ejection nozzle 31a is pulled back, and the ejection of the processing liquid Lq1 is appropriately stopped. In addition, a delay time is generated from the output of the closing signal to the actual stop of the ejection of the processing liquid Lq1. The delay time depends on various factors such as the movement speed of the valve body caused by the closing action of the on-off valve 35a, the movement speed of the valve body of the back-sucking valve 36a, the piping length of the pipe 34a, and the pressure loss.

[0124] Figure 7 and Figure 8An example of a frame of a captured image generated by the camera 70 is schematically shown. Figure 7 and Figure 8 The frames IM3 and IM4 respectively illustrated in the figure represent the frames when the ejection nozzle 31a is switched to the ejection nozzle 31b. Frame IM3 is a frame when the on-off valves 35a and 35b are performing closing and opening actions, respectively. Therefore, in frame IM3, the processing liquid Lq1 and the processing liquid Lq2 are being ejected from both the ejection nozzle 31a and the ejection nozzle 31b, respectively. However, the width of the processing liquid Lq1 ejected from the ejection nozzle 31a is narrower than that in frame IM2. This is because the flow rate of the processing liquid Lq1 is reduced by the closing action of the on-off valve 35a. In frame IM3, the on-off valve 35b has not been fully opened, so the width of the processing liquid Lq2 ejected from the ejection nozzle 31b is also narrow.

[0125] Frame IM4 is a frame in a state where the on-off valve 35a is closed and the on-off valve 35b is open. Therefore, in frame IM4, the processing liquid Lq1 is not ejected from the ejection nozzle 31a, and the processing liquid Lq2 is ejected from the ejection nozzle 31b. The processing liquid Lq2 ejected from the ejection nozzle 31b is also a continuous flow, and has a liquid column shape extending in the vertical direction in the area from the front end of the ejection nozzle 31b to the upper surface of the substrate W. The processing liquid Lq2 lands on the approximate center of the substrate W, and is spread on the upper surface of the substrate W by the centrifugal force accompanying the rotation of the substrate W. Then, it scatters from the periphery of the substrate W. As a result, the processing liquid Lq2 acts on the entire upper surface of the substrate W, and a process based on the processing liquid Lq2 is performed.

[0126] After a predetermined time has passed since step S5, in step S6, the control unit 9 stops ejecting the processing liquid Lq2 from the ejection nozzle 31b. As a specific example, the control unit 9 sends a closing signal to the on-off valve 35b when the time elapsed from the time point when the opening signal is output to the on-off valve 35b reaches the third reference time. The on-off valve 35b performs a closing action based on the closing signal to close the pipe 34b. In addition, when a back-suction valve 36b is provided, a suction signal is sent to the back-suction valve 36b. Thus, the back-suction valve 36b performs a suction action simultaneously with the closing action of the on-off valve 35a to suck the processing liquid Lq2 in the pipe 34b. Thus, the ejection of the processing liquid Lq2 from the ejection nozzle 31b is appropriately stopped. In addition, at this time, a delay time will be generated from the output of the signal to the actual end of the ejection of the processing liquid Lq2.

[0127] The control unit 9 may also stop the rotation of the rotary motor 22 after step S6 to stop the rotation of the substrate W. Alternatively, the control unit 9 may increase the rotation speed of the rotary motor 22, and after the processing liquid Lq2 on the substrate W is scattered from the periphery of the substrate W by the rotational force to dry the substrate W, the rotation of the rotary motor 22 is stopped.

[0128] Next, in step S7, the control unit 9 ends the imaging by the camera 70. Next, in step S8, the control unit 9 controls the nozzle base 33 to move the ejection nozzles 31a and 31b to the standby position.

[0129] Through the above-described operation, a series of processes using the processing liquids Lq1 and Lq2 can be performed in sequence.

[0130] In addition, if Figure 4 As shown in the example, the control unit 9 performs a monitoring process in step S10 simultaneously with steps S4 to S6 in order to monitor the ejection / stop timing of the processing liquid. This monitoring process is a process for monitoring whether the time difference between the start time tb of starting ejection of the processing liquid Lq2 from the ejection nozzle 31b and the stop time ta of stopping ejection of the processing liquid Lq1 from the ejection nozzle 31a in step S5 is appropriate.

[0131] Fig. 9 1 is a flowchart showing an example of specific operations of the monitoring process. First, the control unit 9 performs image processing on the captured image generated by the camera 70 to determine the start time tb of the ejection nozzle 31b and the stop time ta of the ejection nozzle 31a (step S11).

[0132] Here, the start time tb of the ejection nozzle 31b is first described. When a frame is input from the camera 70, the control unit 9 cuts out the ejection determination area Rb1 from the frame. The ejection determination area Rb1 mentioned here is an area extending from the front end of the ejection nozzle 31b to the ejection direction of the processing liquid Lq2 in each frame of the captured image (see also Figures 5 to 8 ). Here, the processing liquid Lq2 extends vertically downward, so the ejection determination area Rb1 has a long strip shape (e.g., rectangular shape) extending in the longitudinal direction of the captured image. The lateral width of the ejection determination area Rb1 is set to be wider than the width of the processing liquid Lq2, and the longitudinal length of the ejection determination area Rb1 is set to a length such that the ejection determination area Rb1 does not include the landing position of the processing liquid Lq2.

[0133] Furthermore, the pixel value in the discharge determination region Rb1 when the processing liquid Lq2 is not discharged from the discharge nozzle 31 b is different from the pixel value in the discharge determination region Rb1 when the processing liquid Lq2 is discharged from the discharge nozzle 31 b. Fig.10 and Fig.11 This is a diagram schematically showing an example of the lateral brightness distribution in the discharge determination region Rb1. Fig.10 exemplifies the brightness distribution when the processing liquid Lq2 is not ejected. Fig.11 exemplifies the brightness distribution when the processing liquid Lq2 is being ejected.

[0134] When the processing liquid Lq2 is being ejected, the liquid column portion of the processing liquid Lq2 is captured in the ejection determination area Rb1. When the illumination light is incident from the same direction as the shooting direction of the camera 70, the surface of the liquid column of the processing liquid Lq2 appears to glow brightly. Fig.11 As shown in the example, the brightness of the portion corresponding to the liquid column is higher than the brightness of the surrounding. Specifically, the brightness distribution has a shape that is convex upward on the liquid column portion. That is, the brightness distribution has a feature caused by the shape of the liquid column of the processing liquid Lq2.

[0135] On the other hand, when the processing liquid Lq2 is not ejected, the shape of the liquid column of the processing liquid Lq2 is not captured in the ejection determination area Rb1. Fig.10 As shown in the example, the brightness distribution does not have the characteristics caused by the shape of the liquid column of the processing liquid Lq2. Although the brightness varies due to diffuse reflection caused by the pattern on the upper surface of the substrate W or the reflection of the internal components of the chamber 10, it has a relatively uniform distribution.

[0136] In addition, the camera 70 may be a camera of a type that generates grayscale captured images or a camera of a type that generates color captured images. In the former case, it can be said that the pixel value of the captured image represents the brightness value. Below, a camera of a type that generates grayscale captured images is cited as an example for explanation, but in the case of color, it is sufficient to calculate the brightness value based on the pixel value and use the brightness value.

[0137] The control unit 9 determines whether the treatment liquid Lq2 is being ejected from the ejection nozzle 31b based on the pixel values ​​in the ejection determination area Rb1. Specifically, the control unit 9 calculates a statistic A2 of the pixel values ​​in the ejection determination area Rb1. The statistic A2 is a value reflecting the ejection state of the treatment liquid Lq2 from the ejection nozzle 31b, for example, the sum (integral value) of the pixel values ​​in the ejection determination area Rb1. The reason for this is that the sum of the pixel values ​​when the treatment liquid Lq2 is being ejected is greater than the sum of the pixel values ​​when the treatment liquid Lq2 is not being ejected.

[0138] As the statistic A2, the variance of the pixel values ​​can also be used instead of the sum of the pixel values. The reason is that Fig.10 and Fig.11 As shown, the brightness distribution when the processing liquid Lq2 is ejected is not uniform compared to the brightness distribution when the processing liquid Lq2 is not ejected. As the variance, for example, the standard deviation can be used. Alternatively, the variance for all pixel values ​​in the ejection determination area Rb1 can be used.

[0139] On the other hand, the processing liquid Lq2 has a liquid column shape along the vertical direction, so the change of the longitudinal brightness distribution in the ejection determination area Rb1 is small. Therefore, the pixels arranged in a row along the horizontal direction may be cut out, and the variance of the plurality of pixel values ​​may be used. Alternatively, the pixel values ​​arranged in a row along the vertical direction may be integrated for each column to calculate the integrated pixel value, and the variance of the integrated pixel value for each column may be used.

[0140] As an example of determination, a threshold value th1 is set for the statistic A2. When the statistic A2 is greater than the threshold value th1, it can be determined that the processing liquid Lq2 is being ejected from the ejection nozzle 31b. When the statistic A2 is less than the threshold value th1, it can be determined that the processing liquid Lq2 is not being ejected from the ejection nozzle 31b. The threshold value th1 can be set in advance through experiments or simulations.

[0141] Fig.12 To show Figure 4 FIG. 1 is a diagram showing an example of the temporal variation of the statistic A2 in steps S4 to S6. Fig.12 In FIG. 1 , the horizontal axis represents the frame number of the captured image generated by the camera 70, and the vertical axis represents the statistic. The frame number increases with the passage of time, so it can also be said that the horizontal axis represents time. The statistic A2 about the ejection determination area Rb1 is represented by a dotted line, and the statistic A1 about the ejection determination area Ra1 described later is represented by a solid line.

[0142] exist Fig.12 The statistic A2 is smaller than the threshold value th1 in the initial stage. This is because the discharge nozzle 31b does not discharge the treatment liquid Lq2 at the beginning of the treatment (see Figure 4 In step S4 of the embodiment of the present invention, when the ejection nozzle is switched in step S5, the statistic A2 increases and exceeds the threshold value th1. That is, the statistic A2 changes from a state less than the threshold value th1 to a state greater than the threshold value th1. The moment when the statistic A2 exceeds the threshold value th1 corresponds to the start time tb. Therefore, the start time tb can be determined based on the change in the statistic A2. The following is a detailed description.

[0143] The control unit 9 calculates the statistic A2 for each frame and determines whether the statistic A2 is greater than the threshold value th1 for each frame. Furthermore, the control unit 9 stores the determination result in a storage medium. When the statistic A2 in the previous frame is less than the threshold value th1 and the statistic A2 in the current frame is greater than the threshold value th1, the control unit 9 determines that the statistic A2 exceeds the threshold value th1.

[0144] The control unit 9 determines the start time tb based on the previous frame and the current frame. That is, the control unit 9 determines the start time tb based on the previous frame whose statistic A2 is less than the threshold value th1 and the current frame which is the next frame of the frame and whose statistic A2 is greater than the threshold value th1. For example, the control unit 9 may determine the generation time of the previous frame as the start time tb, or may determine the generation time of the current frame as the start time tb, or may determine the average of the generation times of the previous and current frames as the start time tb.

[0145] Next, the stop time ta of the ejection nozzle 31a is described. When a frame is input, the control unit 9 cuts out the ejection determination area Ra1 from the frame. The ejection determination area Ra1 mentioned here is an area extending from the front end of the ejection nozzle 31a to the ejection direction of the processing liquid Lq1 in each frame of the captured image (see also Figures 5 to 8 ). Here, the processing liquid Lq1 extends vertically downward, so the ejection determination area Ra1 has a long strip shape (e.g., rectangular shape) extending in the longitudinal direction of the captured image. The lateral width of the ejection determination area Ra1 is set to be wider than the width of the processing liquid Lq1, and the longitudinal length of the ejection determination area Rb1 is set to be a length such that the ejection determination area Rb1 does not include the landing position of the processing liquid Lq1.

[0146] Similar to the ejection determination area Rb1, the brightness distribution in the ejection determination area Ra1 is different depending on whether the processing liquid Lq1 is ejected. Therefore, similar to the determination of whether the processing liquid Lq2 is ejected, the control unit 9 determines whether the processing liquid Lq1 is ejected based on the pixel value of the ejection determination area Ra1. More specifically, the control unit 9 calculates the statistic A1 of the pixel values ​​in the ejection determination area Ra1. Similar to the statistic A2, the statistic A1 is a value reflecting the ejection state of the processing liquid Lq1 from the ejection nozzle 31a, for example, the sum or variance of the pixel values ​​in the ejection determination area Ra1.

[0147] When the statistic A1 is large, it can be determined that the treatment liquid Lq1 is being ejected from the ejection nozzle 31a, and when the statistic A1 is small, it can be determined that the treatment liquid Lq1 is not being ejected from the ejection nozzle 31a. Therefore, a threshold value for the statistic A1 used for these determinations is set. Here, the threshold value th1 for the statistic A2 is used as the threshold value for the statistic A1. In addition, a value different from the threshold value th1 may be used as the threshold value for the statistic A1.

[0148] like Fig.12 As shown in the example, the statistic A1 is larger than the threshold value th1 at the initial stage. This is because the processing liquid Lq1 is ejected from the beginning of the processing (see Figure 4Then, when the ejection nozzle is switched in step S5, the statistic A1 decreases and becomes lower than the threshold value th1. The moment when the statistic A1 becomes lower than the threshold value th1 corresponds to the stop time ta. Therefore, the stop time ta can be determined based on the change of the statistic A1. The following is a detailed description.

[0149] The control unit 9 calculates the statistic A1 for each frame and determines whether the statistic A1 is greater than the threshold value th1 for each frame. Furthermore, the control unit 9 stores the determination result in a storage medium. When the statistic A1 is greater than the threshold value th1 in the previous frame and the statistic A1 is less than the threshold value th1 in the current frame, the control unit 9 determines that the statistic A1 is less than the threshold value th1.

[0150] Furthermore, the control unit 9 determines the stop time ta based on the previous frame and the current frame. That is, the control unit 9 determines the stop time ta based on the previous frame whose statistic A1 is greater than the threshold value th1 and the current frame which is the next frame of the frame and whose statistic A2 is less than the threshold value th1. For example, the control unit 9 may determine the generation time of the previous frame as the stop time ta, or may determine the generation time of the current frame as the stop time ta, or may determine the average of the generation times of the previous and current frames as the stop time ta.

[0151] Next, in step S12, the control unit 9 calculates the time difference between the start time tb and the stop time ta. Specifically, the control unit 9 calculates the time difference by subtracting the start time tb from the stop time ta.

[0152] Next, in step S13, the control unit 9 determines whether the time difference is outside a predetermined range. The predetermined range may be preset, for example, and stored in a storage medium. The lower limit value and the upper limit value of the predetermined range may have positive values, for example.

[0153] When the time difference is outside the prescribed range, in step S14, the control unit 9 performs an error notification process. For example, the control unit 9 causes the error to be displayed on the display of the user interface 90. Alternatively, when a sound output unit such as a buzzer or a speaker is provided, the control unit 9 may also cause the sound output unit to output the error. The display and sound output unit of the user interface 90 are an example of a notification unit. In short, the control unit 9 causes the notification unit to notify the error. Through such notification, the operator can recognize that the time difference is outside the prescribed range.

[0154] Next, in step S15 , the control unit 9 adjusts at least one of the start time tb and the stop time ta so that the time difference falls within a predetermined range.

[0155] When the time difference is greater than the upper limit of the prescribed range, in order to reduce the time difference and limit it within the prescribed range, the control unit 9, for example, updates the start time tb of the ejection nozzle 31b to a later time. As a more specific example, the control unit 9 updates the second reference time for the opening time of the opening and closing valve 35b to a longer value in a manner that makes the time difference within the prescribed range, and stores the updated second reference time in the storage medium. As a result, when switching to step S5 after the next time, the time of outputting the opening signal to the opening and closing valve 35b is delayed, and thus the start time tb is delayed. Therefore, the time difference in step S5 after the next time (step S5 for the next substrate) can be reduced and set within the prescribed range.

[0156] Thus, the stop timing ta of the discharge of the processing liquid Lq1 does not change, and thus the length of the processing period for the processing liquid Lq1 does not change. Therefore, the processing using the processing liquid Lq1 can be appropriately performed.

[0157] Alternatively, the control unit 9 may also update the stop time ta of the ejection nozzle 31a to an earlier time. As a more specific example, the control unit 9 updates the first reference time that specifies the closing time of the on-off valve 35a to a shorter value in such a way that the time difference is within the specified range, and stores the updated first reference time in the storage medium. As a result, when switching to step S5 next time or later, the time of outputting the closing signal to the on-off valve 35a is advanced, and thus the stop time ta is advanced. Therefore, the time difference in step S5 next time or later can be reduced and set within the specified range.

[0158] Alternatively, the start time tb and the stop time ta may be adjusted to reduce the time difference.

[0159] In addition, when the time difference is greater than the upper limit of the specified range, the overlap period of the two sprayed processing liquids Lq1 and Lq2 is long. That is, the total amount of processing liquid sprayed onto the substrate W temporarily increases. As a result, for example, a splash may be generated in which the processing liquid Lq2 collides with the processing liquid Lq1 on the substrate W and a part of the processing liquid splashes on the substrate W. In the first embodiment, when the time difference is greater than the upper limit of the specified range, the time difference in the next step S5 can be limited to the specified range. Therefore, by adopting a value that does not generate splashes as the upper limit of the specified range, the generation of splashes can be roughly avoided.

[0160] On the other hand, when the time difference is less than the lower limit of the prescribed range, the control unit 9, for example, updates the start time tb of the ejection nozzle 31b to an earlier time in order to increase the time difference and limit it within the prescribed range. As a more specific example, the control unit 9 updates the second reference time that specifies the opening time of the on-off valve 35b to a shorter value so that the time difference is within the prescribed range, and stores the updated second reference time in the storage medium. In this way, the time difference in the next step S5 can be increased and set within the prescribed range.

[0161] Thus, the stop timing ta of the discharge of the processing liquid Lq1 does not change, and thus the length of the processing period for the processing liquid Lq1 does not change. Therefore, the processing using the processing liquid Lq1 can be appropriately performed.

[0162] Alternatively, the control unit 9 may also update the stop time ta of the ejection nozzle 31a to a later time. As a more specific example, the control unit 9 updates the first reference time for specifying the closing time of the on-off valve 35a to a longer value so that the time difference is within the specified range, and stores the updated first reference time in the storage medium. Thus, the time difference in the next step S5 and thereafter may be increased and set within the specified range.

[0163] Alternatively, the start time tb and the stop time ta may be adjusted to increase the time difference.

[0164] In addition, when the time difference is less than the lower limit of the specified range, the overlapping period of the two spraying treatment liquids Lq1 and Lq2 becomes shorter. That is, the treatment liquid Lq2 starts to be sprayed when the spraying of the treatment liquid Lq1 is almost stopped. Since the substrate W is rotating, the treatment liquid Lq1 supplied to its upper surface is moved to the peripheral side of the substrate W by the centrifugal force. Therefore, if the spraying amount of the treatment liquid Lq1 decreases when the treatment liquid Lq2 is not sprayed, the treatment liquid Lq1 on the upper surface of the substrate W decreases, so that the substrate W may be locally dried (liquid drying up). In particular, near the landing position of the treatment liquid Lq1, the upper surface of the substrate W may be locally dried. Such drying may cause an undesirable condition (such as the adhesion of particles, etc.) on the upper surface of the substrate W, and is therefore not preferred. In the first embodiment, even when the time difference is less than the lower limit of the specified range, the time difference in the next and subsequent steps S5 can be limited to a specified range. Therefore, by setting the lower limit value of the predetermined range to a value at which local drying of the substrate W does not occur, the occurrence of local drying of the substrate W can be substantially avoided.

[0165] As described above, according to the first embodiment, when the time difference is outside the predetermined range, it is adjusted to be within the predetermined range. Therefore, it is possible to substantially avoid the occurrence of inconveniences (liquid splashing and liquid drying) caused by the time difference being outside the predetermined range.

[0166] In addition, according to the first embodiment, the start time tb of the ejection nozzle 31b and the stop time ta of the ejection nozzle 31a are determined by image processing of the captured image from the camera 70. That is, the start time tb and the stop time ta can be determined based on the actual ejection state of the processing liquid Lq1 and Lq2, so that the determination accuracy can be improved. Moreover, the time difference is calculated based on the time determined with high accuracy, so the calculation accuracy of the time difference is also high. Therefore, the time difference can be limited to a specified range with higher accuracy.

[0167] Furthermore, since the pixel values ​​of the discharge determination areas Ra1 and Rb1 are used, the processing can be reduced compared to the case where the image processing is performed on the entire captured image.

[0168] Furthermore, since the time difference can be obtained based on the comparison between the statistic and the threshold value th1, the processing is simple.

[0169] In addition, in the above example, the upper limit of the specified range is set to a value at which no splashing occurs, and the lower limit is set to a value at which no liquid drying occurs. However, depending on other factors, the upper limit may be set to a smaller value, and the lower limit may be set to a larger value. For example, the time difference may sometimes affect whether the processing result for the substrate W is good. In this case, the upper limit and lower limit of the specified range of the time difference may be set by experiments or simulations to make the processing result good.

[0170] Due to manufacturing deviations of the plurality of processing units 1, the delay time for the ejection control may be different for each processing unit 1. However, by performing the above-mentioned operation on each processing unit 1 by the control unit 9, the time difference can be adjusted to within a specified range for each processing unit 1. In the past, in order to adjust the start time tb and the stop time ta for each processing unit 1, the processing recipe information of the substrate W was changed for each processing unit 1. As a result, the management of the processing recipe information becomes complicated. In the first embodiment, even if the common processing recipe information is used, each processing unit 1 can be processed with the optimal time difference, and stable processing performance can be achieved.

[0171] In addition, in the above-described example, after the discharge of the processing liquid Lq2 from the discharge nozzle 31 b starts, the discharge of the processing liquid Lq1 from the discharge nozzle 31 a stops. Fig.13 is a graph showing another example of the time variation of a statistic. Fig.13In the example, the start time tb of the ejection nozzle 31b occurs after the stop time ta of the ejection nozzle 31a, and the time difference is relatively large. In this case, since the processing liquid Lq1 on the upper surface of the substrate W decreases before the processing liquid Lq2 is ejected, the substrate W may be partially dried.

[0172] In this case, the time difference obtained by subtracting the start time tb from the stop time ta in step S12 has a negative value. Therefore, the time difference is less than the lower limit of the prescribed range, and the control unit 9 determines that the time difference is outside the prescribed range in step S13. Therefore, in step S15, the control unit 9 adjusts at least one of the start time tb and the stop time ta so that the time difference is within the prescribed range. Therefore, the time difference can be set within the prescribed range in the next step S5.

[0173] In addition, the above description can be understood as an action during processing of the substrate W, and can also be understood as an action during initial setting performed when setting the substrate processing apparatus 100. That is, if a temporary time is set during initial setting to actually process the substrate W, then when the temporary time is not appropriate, an appropriate time is set.

[0174] <User Interface>

[0175] In the above example, the control unit 9 determines whether the time difference is outside the predetermined range. However, the determination may be made by an operator.

[0176] The control unit 9 causes the display of the user interface 90 to display a graph showing the temporal changes of the statistics A1 and A2. Preferably, the control unit 9 causes the threshold value th1 to be also displayed on the graph. Fig.12 or Fig.13 The graph shown is displayed on the display. This allows the operator to visually check the temporal changes of the statistics A1 and A2 and determine the magnitude of the time difference.

[0177] The input unit of the user interface 90 receives input for adjusting the start time tb and the stop time ta. For example, when the operator determines that the time difference is greater than the upper limit of the specified range, the operator inputs at least one of an input for delaying the start time tb and an input for advancing the stop time ta to the input unit. The input unit outputs the input information to the control unit 9. The control unit 9 adjusts at least one of the start time tb and the stop time ta according to the input information. The same is true when the operator determines that the time difference is less than the lower limit of the specified range.

[0178] As described above, the operator can visually check the temporal changes of the statistics A1 and A2 and adjust the time difference based on the temporal changes.

[0179] Machine Learning

[0180] In the above example, the control unit 9 obtains the time difference between the start time tb and the stop time ta based on the statistics of the pixel values, but the present invention is not limited thereto. The control unit 9 may obtain the time difference between the start time tb and the stop time ta by machine learning during the monitoring process.

[0181] Fig.14 : This is a diagram schematically showing an example of the internal structure of the control unit 9. The control unit 9 includes a classifier 91 and a machine learning unit 92. The classifier 91 sequentially inputs each frame of the captured image from the camera 70. The classifier 91 classifies each input frame into the following four categories C1 to C4 related to the ejection / stop of the ejection nozzles 31a and 31b. A category may also be referred to as a class.

[0182] Four categories C1 to C4 represent Figures 5 to 8 More specifically, category C1 indicates a state where both the ejection nozzles 31a and 31b do not eject the processing liquid ( Figure 5 ) category, and category C2 indicates a state where only the ejection nozzle 31a is ejecting the processing liquid ( Figure 6 ) category, and category C3 indicates that both the ejection nozzles 31a and 31b are ejecting the processing liquid ( Figure 7 ) category, and category C4 indicates that only the ejection nozzle 31b is ejecting the processing liquid ( Figure 8 ) category.

[0183] The classifier 91 is generated by a machine learning unit 92 using a plurality of teaching data. That is, the classifier 91 can also be said to be a classifier that has been learned by machine learning. The machine learning unit 92 uses, for example, the nearest neighbor method, support vector machine, random forest, or neural network (including deep learning) as a machine learning algorithm.

[0184] The teaching data includes learning data and a label indicating which category the learning data is classified into. The learning data is a frame of a captured image captured by the camera 70 and is generated in advance. Each learning data is assigned a correct category as a label. This assignment can be performed by an operator operating, for example, the user interface 90. The machine learning unit 92 performs machine learning based on these teaching data to generate a classifier 91.

[0185] As an example, a classifier 91 that classifies frames by the nearest neighbor method is described. The classifier 91 has a feature vector extraction unit 911, a determination unit 912, and a storage medium storing a determination database 913. Each frame of the captured image from the camera 70 is sequentially input to the feature vector extraction unit 911. The feature vector extraction unit 911 extracts a feature vector of the frame according to a prescribed algorithm. The feature vector is a vector that easily represents a feature quantity corresponding to the ejection state of the ejection nozzles 31a and 31b. The algorithm can adopt a known algorithm. The feature vector extraction unit 911 outputs the feature vector to the determination unit 912.

[0186] The judgment database 913 stores a plurality of feature vectors (hereinafter referred to as reference vectors) generated from a plurality of teaching data by the machine learning unit 92, and the reference vectors are classified into categories C1 to C4. Specifically, the machine learning unit 92 applies the same algorithm as the feature vector extraction unit 911 to the plurality of teaching data to generate a plurality of reference vectors. In addition, the machine learning unit 92 assigns the label (correct category) of the teaching data to the reference vector.

[0187] The determination unit 912 classifies the frame based on the feature vector input from the feature vector extraction unit 911 and the plurality of reference vectors stored in the determination database 913. For example, the determination unit 912 may determine the reference vector closest to the feature vector and classify the frame into the class of the determined reference vector (nearest neighbor algorithm). Thus, the determination unit 912 can classify the frame input to the classifier 91 (feature vector extraction unit 911) into one of the classes C1 to C4.

[0188] The control unit 9 classifies each frame using the classifier 91 , and obtains the time difference between the start time tb of the discharge nozzle 31 b and the stop time ta of the discharge nozzle 31 a based on the classification result.

[0189] Fig.15 2 is a flowchart showing an example of the operation of the control unit 9 in the monitoring process. In step S40, the control unit 9 determines the start time tb and the stop time ta by machine learning as image processing.

[0190] Fig.16 FIG. 1 is a diagram schematically showing an example of a plurality of frames F of a captured image. Fig.16 In the example of , the frame F is arranged in a time series. Here, the reference numerals of the frames are changed for convenience of explanation, and the frame F is the same type of frame as the above-mentioned frames IM1 to IM4.

[0191] like Fig.16As shown in the example, at the initial stage of the operation of the substrate processing apparatus 100, neither the discharge nozzles 31a nor 31b discharges the processing liquid. Fig.16 The illustrated frames F[ 1 ] to F[k] are classified into the class C1 by the classifier 91 .

[0192] Next, when the treatment liquid Lq1 is ejected from the ejection nozzle 31a (step S4), Fig.16 As illustrated, the subsequent frames F[k+1] to F[m] are classified into class C2 by the classifier 91 .

[0193] Next, the nozzle for ejecting the processing liquid is switched from the ejection nozzle 31a to the ejection nozzle 31b (step S5). That is, the processing liquid Lq2 is ejected from the ejection nozzle 31b. Fig.16 As shown in the example, the subsequent frames F[m+1] to F[n] are classified into the class C3 by the classifier 91. Then, the ejection of the processing liquid Lq1 from the ejection nozzle 31a is stopped. Fig.16 As illustrated, the frames after the subsequent frame F[n+1], that is, the frames until the discharge of the processing liquid Lq2 from the discharge nozzle 31 b is completed, are classified into the class C4 by the classifier 91 .

[0194] As described in detail below, the control unit 9 determines the start time tb and the stop time ta based on the classification result of each frame.

[0195] That is, the control unit 9 determines the start time tb of the ejection nozzle 31b based on the m-th frame F[m] in which the ejection nozzle 31b is classified as stopped (class C2) and the (m+1)-th frame after the frame F[m], that is, the frame F[m+1] in which the ejection nozzle 31b is classified as ejecting (class C3). For example, the control unit 9 may determine the generation time of the frame F[m] as the start time tb, the generation time of the frame F[m+1] as the start time tb, or the average of the generation times of the frames F[m] and F[m+1] as the start time tb.

[0196] Similarly, the control unit 9 determines the stop time ta of the ejection nozzle 31a based on the nth frame F[n] in which the ejection nozzle 31a is classified as ejection (class C3) and the (n+1)th frame after the frame F[n], that is, the frame F[n+1] in which the ejection nozzle 31a is classified as stop (class C4). For example, the control unit 9 may determine the generation time of the frame F[n] as the stop time ta, the generation time of the frame F[n+1] as the stop time ta, or the average of the generation times of the frames F[n] and F[n+1] as the stop time ta.

[0197] As described above, the control unit 9 can determine (classify) the ejection state of the ejection nozzles 31a and 31b by machine learning. Therefore, each frame can be classified with high accuracy. Furthermore, the time difference between the stop time ta and the start time tb can be obtained with high accuracy.

[0198] Steps S41 to S44 after the stop time ta and the start time tb are determined are respectively the same as steps S12 to S15 of the first embodiment.

[0199] In addition, when the ejection of the ejection nozzle 31b is started before the ejection of the ejection nozzle 31a is stopped, the overlapping period of the ejection of the processing liquids Lq1 and Lq2 is equivalent to the time difference. Therefore, in this case, the control unit 9 may also calculate the time difference based on the number of frames classified into the category C3. For example, the control unit 9 may also calculate the time difference by multiplying the time between frames by the number of frames.

[0200] <Input to the classifier>

[0201] In the above example, the entire area of ​​each frame F of the captured image is used as input data to the classifier 91, but the present invention is not necessarily limited to this. For example, the control unit 9 may cut out images representing the ejection determination areas Ra1 and Rb1 in the frame F and input the images to the classifier 91. In this case, images representing the ejection determination areas Ra1 and Rb1 are also used as learning data to be input to the machine learning unit 92.

[0202] Thus, the classifier 91 can remove the influence of the area with low correlation with the ejection state to perform classification, thereby improving its classification accuracy. Furthermore, the accuracy of determining the start time tb and the stop time ta can be improved. In addition, if the ejection determination areas Ra1 and Rb1 are used, the processing can be reduced compared to the case where the classifier 91 processes the entire frame of the captured image.

[0203] In addition, the classifier 91 may also classify the ejection state for each image of the ejection determination area Ra1, Rb2. That is, the classifier 91 may also classify the image into the following two categories Ca1 and Ca2 based on the image of the ejection determination area Ra1. That is, the category Ca1 indicates the state in which the ejection nozzle 31a is not ejecting the processing liquid Lq1, and the category Ca2 indicates the state in which the ejection nozzle 31a is ejecting the processing liquid Lq1. Similarly, the classifier 91 may also classify the image into the following two categories Cb1 and Cb2 based on the image of the ejection determination area Rb2. That is, the category Cb1 indicates the state in which the ejection nozzle 31b is not ejecting the processing liquid Lq2, and the category Cb2 indicates the state in which the ejection nozzle 31b is ejecting the processing liquid Lq2.

[0204] The control unit 9 can determine the start time tb of the ejection nozzle 31b based on the frame including the image classified as the class Cb1 and the frame including the next image classified as the class Cb2. The same applies to the stop time ta of the ejection nozzle 31a.

[0205] In addition, as input data to the classifier 91, for example, a pixel value group of pixels arranged in a row in each ejection determination area Ra1, Rb1 may be used. Fig.10 and Fig.11 As shown, the pixel value group of the pixels arranged in the horizontal direction changes depending on whether the processing liquid is ejected. Alternatively, as input data of the classifier 91, an integral value group including an integral value for each column can be used, wherein the integral value for each column is the sum of the pixel values ​​arranged in a column in the vertical direction in each ejection determination area Ra1, Rb1.

[0206] <Multiple classifiers>

[0207] Fig.17 FIG. 1 is a functional block diagram schematically showing an example of the internal structure of the control unit 9. The control unit 9 includes a plurality of classifiers 91. Fig.14 These multiple classifiers 91 are also generated by the machine learning unit 92.

[0208] For example, the machine learning unit 92 generates a plurality of classifiers 91 for each type of the processing liquids Lq1 and Lq2. Fig.17 In the example of FIG. 1 , three classifiers 91A to 91C are shown as the plurality of classifiers 91. The classifiers 91A to 91C are classifiers generated using different teaching data.

[0209] For example, each frame of a captured image when a certain treatment liquid Lq1 or Lq2 is used for treatment is used as learning data. The machine learning unit 92 generates a classifier 91A based on the teaching data containing the learning data. Thus, a classifier 91A for a group (hereinafter referred to as the first group) of the types of the treatment liquids Lq1 and Lq2 can be generated. Similarly, the machine learning unit 92 generates a classifier 91B for the second group based on the teaching data of the treatment liquids Lq1 and Lq2 of the second group different from the first group, and generates a classifier 91C for the third group based on the teaching data of the treatment liquids Lq1 and Lq2 of the third group different from the first group and the second group.

[0210] exist Fig.17In the example of FIG. 1 , the feature vector extracting unit 911 and the judging unit 912 are provided in common in the classifiers 91A to 91C, and the classifiers 91A to 91C are distinguished by the judging database 913. In short, the machine learning unit 92 generates the judging databases 913A to 913C for the first group to the third group, respectively, thereby generating the classifiers 91A to 91C. In the judging database 913A, the reference vectors when the processing liquids Lq1 and Lq2 of the first group are used are recorded, and the correct categories thereof are recorded. The same is true for the judging databases 913B and 913C.

[0211] The user interface 90 is connected to the determination unit 912. The operator inputs the types of the processing liquids Lq1 and Lq2 into the user interface 90. Fig.18 is a diagram showing an example of an input screen 90a displayed on the display of the user interface 90. On the input screen 90a, a table 901 related to the processing liquids Lq1 and Lq2 is displayed. In the table 901, the type of processing liquid, the flow rate, the rotation speed of the substrate W, and the processing time are displayed. Various information in the table 901 can be changed by inputting to the input unit by the operator. For example, by clicking (or touching) the part of the table 901, the information of the part can be input. For example, by the click, a plurality of information is displayed in a drop-down form in the part, and the operator can select one of them to input information.

[0212] exist Fig.18 The illustrated input screen 90a also displays a soft key 902 for selecting the type of substrate W. The operator can select the soft key 902 by clicking or touching the soft key 902. As the type of substrate W, there are, for example, a silicon (Si) substrate and a silicon carbide (SiC) substrate. In addition, it is also possible to select a determination database based on whether a film is formed on the upper surface of the substrate W or the type of the film formed on the upper surface of the substrate W (for example, SiO2, SiN, TiN, etc.).

[0213] In addition, Fig.18 The illustrated input screen 90a shows an area 903 for displaying a captured image generated by the camera 70. This allows the operator to visually check the captured image by the camera 70.

[0214] The user interface 90 outputs input information (such as the types of the processing liquids Lq1 and Lq2 ) input by the operator to the determination unit 912 .

[0215] The feature vector extraction unit 911 extracts a feature vector of the input frame and outputs the feature vector to the determination unit 912. The determination unit 912 selects a determination database 913 corresponding to the type of the processing liquid Lq1, Lq2 to be processed. The determination unit 912 classifies the frame using the input feature vector and the reference vector of the selected determination database 913.

[0216] Thus, by classifying the frames using the classifier 91 corresponding to the types of the processing liquids Lq1 and Lq2, the accuracy of the frame classification can be improved. Furthermore, the accuracy of the time difference calculation can be improved.

[0217] In addition, in the above-mentioned example, the classifier 91 is generated for each type of the processing liquid Lq1, Lq2, but, for example, the classifier 91 may be generated for each type of substrate W or each flow rate of the processing liquid Lq1, Lq2. In addition, for example, the position of the ejection nozzle in the captured image of the camera 70 may be different. For example, in the captured image, the positional relationship of the ejection nozzle 31 of the processing liquid supply unit 30 and the positional relationship of the ejection nozzle 66 of the processing liquid supply unit 65 may be different. In this case, the classifier 91 may also be generated for each position of the ejection nozzle. In addition, these situations may also be combined. For example, a classifier 91 corresponding to the type of the group of processing liquids Lq1, Lq2 and the type of substrate W may also be generated. For example, when there are N types of groups of processing liquids Lq1, Lq2 and M types of substrates W, (N×M) classifiers 91 corresponding to the combination thereof may also be generated.

[0218] The user interface 90 only needs to accept the input of the information required to select the classifier 91 (at least one of the type of processing liquid Lq1, Lq2, the type of substrate W, the flow rate of processing liquid Lq1, Lq2 and the position of the ejection nozzle), and output the input information to the control unit 9.

[0219] In addition, the information required for selecting the classifier 91 does not necessarily need to be input by the operator. For example, sometimes substrate information about the substrate W is sent to the control unit 9 from the upstream side of the substrate processing device 100, and this information is included in the substrate information. In this case, the control unit 9 only needs to select the classifier 91 based on the information in the substrate information.

[0220] <Server>

[0221] In the above example, the control unit 9 provided in the substrate processing apparatus 100 generates the classifier 91 by machine learning, and classifies the frame by the classifier 91. However, at least part of the machine learning function performed by the control unit 9 may be provided in a server.

[0222] Fig.191 is a functional block diagram schematically showing an example of the electrical structure of a substrate processing system. The substrate processing system includes a substrate processing apparatus 100 and a server 200. Fig.19 As illustrated, the control unit 9 of the substrate processing apparatus 100 communicates with the server 200 via the communication unit 93. The communication unit 93 is a communication interface, and can communicate with the server 200 by wire or wirelessly.

[0223] exist Fig.19 In the example of , the server 200 has a machine learning unit 210 and a storage medium storing a determination database 220. The machine learning unit 210 has the same function as the machine learning unit 92 and can generate the determination database 220 based on the teaching data. The determination database 220 is the same as the determination database 913.

[0224] The determination unit 912 transmits a request signal for requesting the determination database 220 to the server 200 via the communication unit 93. In response to the request signal, the server 200 transmits the determination database 220 to the communication unit 93. Thus, the determination unit 912 can use the determination database 220 stored in the server 200. In this case, the machine learning unit 92 and the determination database 913 are not necessarily required.

[0225] In addition, in the above example, the determination unit 912 is provided in the control unit 9 of the substrate processing apparatus 100, but the determination unit 912 may also be provided in the server 200. In this case, the feature vector extraction unit 911 sends the feature vector to the server 200 via the communication unit 93. The server 200 classifies the frame based on the received feature vector and the determination database 220, and sends the classification result to the communication unit 93. The communication unit 93 outputs the classification result to the control unit 9.

[0226] In addition, in the above-mentioned example, the feature vector extraction unit 911 is provided in the control unit 9 of the substrate processing apparatus 100, but the feature vector extraction unit 911 may also be provided in the server 200. That is, the classifier 91 itself may also be provided in the server 200. In this case, the control unit 9 transmits the frame of the captured image generated by the camera 70 to the server 200 via the communication unit 93. The server 200 extracts a feature vector from the frame, and classifies the frame using the extracted feature vector and the determination database 220, and transmits the classification result to the communication unit 93. The communication unit 93 outputs the classification result to the control unit 9.

[0227] According to the above-described aspect, since the server 200 is provided with the determination processing function, it is possible to perform common determination on a plurality of substrate processing units.

[0228] In addition, the machine learning unit 210 of the server 200 may also generate a determination database 220 for at least one of the type of each processing liquid Lq1, Lq2, the type of each substrate W, the flow rate of each processing liquid Lq1, Lq2, and the position of each ejection nozzle 31a, 31b. In this case, the control unit 9 only needs to send information specifying the determination database 220 to be used to the server 200 via the communication unit 93.

[0229] In short, the substrate processing apparatus 100 and the server 200 may perform the following functions as a whole, that is, classify each frame included in the captured image into each category using a classifier obtained through machine learning, and calculate the time difference based on the classification result.

[0230] Deep Learning

[0231] As machine learning, deep learning can also be used. Fig. 20 The model of the neural network (including deep learning) NN1 is shown. In this model, an input layer, an intermediate layer (hidden layer) and an output layer are provided. Each layer has a plurality of nodes (artificial neurons), and in each node, the output data of the nodes of the previous layer are respectively weighted and input. That is, in each node, the multiplication result obtained by multiplying the output of the node of the previous stage by the respective weighting coefficients is respectively input. Each node outputs the result of a known function, for example. The number of intermediate layers is not limited to 1 and can be set arbitrarily.

[0232] The machine learning unit 92 determines the weighting coefficient for weighting between each node by learning based on the teaching data. The weighting coefficient is stored as a determination database. Thus, the machine learning unit 92 can substantially generate the classifier 91.

[0233] The input layer receives the frame F generated by the camera 70. The classifier 91 uses the weighting coefficients stored in the determination database to perform arithmetic processing from the input layer to the output layer via the intermediate layer based on the frame F, thereby calculating the probability that the frame F corresponds to each category C1 to C4. The classifier 91 classifies the frame into the category with the highest probability.

[0234] As described above, frames can be classified by a neural network. In a neural network, the classifier 91 automatically generates feature quantities, so the designer does not need to determine feature vectors.

[0235] In addition, a plurality of classifiers 91 may also be generated in the neural network. For example, the machine learning unit 92 may perform machine learning using teaching data for each type of the processing liquids Lq1 and Lq2 to generate a determination database (weighting coefficient) for each type.

[0236] The classifier 91 determines the type of the processing liquids Lq1 and Lq2 based on the input information from the user interface 90 , and classifies the frame F using a determination database corresponding to the type. This can improve classification accuracy.

[0237] Of course, the classifier 91 is not limited to the type of the processing liquids Lq1 and Lq2, and may be generated for at least one of the type of each substrate W, the flow rate of each processing liquid Lq1 and Lq2, and the position of each discharge nozzle 31a and 31b.

[0238] Second embodiment

[0239] An example of the structure of the substrate processing apparatus 100 of the second embodiment is the same as that of the first embodiment, and an example of its operation is as follows: Figure 4 However, the specific example of the monitoring process is different from that of the first embodiment. In the second embodiment, the control unit 9 performs image processing on the captured image captured by the camera 70, thereby detecting the splashing generated when the ejection nozzle 31a is switched to the ejection nozzle 31b. That is, in the monitoring process, monitoring is performed to determine whether splashing occurs.

[0240] Fig.21 is a diagram showing an example of a frame of a captured image generated by the camera 70. Fig.21 In the illustrated frame IM5, the processing liquids Lq1 and Lq2 are ejected from the ejection nozzles 31a and 31b, respectively, and splashes are generated. The splashes are generated near the landing positions of the ejection nozzles 31a and 31b.

[0241] Since the area where the splashing occurs is known in advance, the splashing determination area R2 can be set in the captured image to detect the splashing. Specifically, the splashing determination area R2 is set near the ejection nozzles 31a and 31b in each frame of the captured image. The splashing determination area R2 is set in the area where the substrate W is captured, that is, the area that does not overlap with the ejection determination areas Ra1 and Rb1. Fig.21 In the example of , the splash determination area R2 has determination areas R21 and R22, which are set near the ejection nozzles 31a and 31b. More specifically, the determination areas R21 and R22 are located on opposite sides of a set of ejection nozzles 31a and 31b in the horizontal direction of the captured image. Fig.21 In the example of FIG. 1 , the determination regions R21 and R22 have a rectangular shape.

[0242] When splashing liquid is captured in the judgment areas R21 and R22, the average value of the brightness value in the judgment areas R21 and R22 becomes higher due to the reflection of the illumination light at the splashing liquid, and the brightness distribution becomes largely non-uniform. In other words, it is possible to determine whether splashing liquid has occurred based on the sum or variance (e.g., standard deviation) of the pixel values ​​in the judgment areas R21 and R22. Therefore, the control unit 9 calculates the sum or variance (e.g., standard deviation) of the pixel values ​​in the judgment area R21, i.e., the statistic B1, and the sum or variance (e.g., standard deviation) of the pixel values ​​in the judgment area R22, i.e., the statistic B2. If the statistics B1 and B2 of the pixel values ​​in the judgment areas R21 and R22 become larger, it can be determined that splashing liquid has occurred.

[0243] Fig. 22 : is a diagram showing an example of temporal changes in statistics B1 and B2. Fig. 22 The example shows the statistics B1 and B2 when splashing occurs in step S5. Fig. 22 As shown in the example, when splashing occurs, the statistics B1 and B2 increase and exceed the threshold value th2. Conversely, when at least one of the statistics B1 and B2 is greater than the threshold value th2, splashing can be determined. Such a threshold value th2 can be set in advance, for example, through experiments or simulations.

[0244] <Operation of the control unit 9>

[0245] An example of the operation of the control unit 9 is Figure 4 However, the specific content of the monitoring process in step S10 is different. In the second embodiment, the monitoring process is a process of monitoring whether splashing occurs based on the captured image generated by the camera 70.

[0246] Fig.23 Flowchart showing an example of the monitoring process of the second embodiment. First, in step S30, the control unit 9 initializes the value nn to 1. Next, in step S31, the control unit 9 determines whether splashing occurs based on the nnth frame F[nn]. Specifically, the control unit 9 calculates the statistics B1 and B2 of the pixel values ​​in the determination areas R21 and R22 of the frame F[nn].

[0247] Next, the control unit 9 determines whether at least one of the statistics B1 and B2 is greater than the threshold value th2. When both the statistics B1 and B2 are less than the threshold value th2, it is determined that no splashing occurs. In step S32, the control unit 9 adds 1 to the value nn to update it, and executes step S31 for the updated value nn. That is, when both the statistics B1 and B2 are less than the threshold value th2, it is determined that no splashing occurs, and the determination of step S31 for the next frame is performed.

[0248] On the other hand, when at least one of the statistics B1 and B2 is greater than the threshold value th2, in step S33, the control unit 9 performs an error notification process. For example, the control unit 9 displays the error on the display of the user interface 90. Alternatively, when a sound output unit such as a buzzer or a speaker is provided, the control unit 9 may also cause the sound output unit to output the error. Through such notification, the operator can recognize that splashing occurs.

[0249] Next, in step S34, the control unit 9 adjusts either the start time tb or the stop time ta. That is, when at least either the statistics B1 or B2 is greater than the threshold th2, it is determined that splashing is detected, and the control unit 9 adjusts the time difference. Specifically, at least either the start time tb or the stop time ta is adjusted in such a way that the time difference becomes smaller.

[0250] The amount of reduction of the time difference may also be preset. That is, in step S34, the control unit 9 may also reduce the time difference by a predetermined amount. Then, the control unit 9 executes again Figure 4 At this time, when splashing is detected again in step S31, the time difference is reduced again by the reduction amount in step S34. By repeating this series of actions, the time difference is adjusted to a value that does not cause splashing.

[0251] As described above, in the second embodiment, when the occurrence of splashing is detected, the time difference is adjusted so that splashing does not occur. This makes it possible to avoid or suppress the occurrence of splashing in subsequent processing.

[0252] Machine Learning

[0253] In the above example, the control unit 9 detects liquid splashing based on the statistics B1 and B2 of the pixel values, but the present invention is not limited thereto. The control unit 9 may detect liquid splashing by machine learning.

[0254] An example of the internal structure of the control unit 9 is Fig.14 However, in the second embodiment, the classifier 91 classifies each frame input from the camera 70 into the following two categories C11 and C12. That is, category C11 is a category indicating a state where no splashing occurs, and category C12 is a category indicating a state where splashing occurs.

[0255] As learning data, frames of images captured by the camera 70 are used, and correct categories are assigned as labels to the learning data, thereby generating teaching data. The machine learning unit 92 generates a determination database 913 based on the teaching data.

[0256] A flowchart showing an example of the operation of the control unit 9 in the monitoring process is shown in FIG. Fig.23However, in step S31, the control unit 9 determines whether splashing is detected based on the classification result of the frame F[nn] by the classifier 91. Specifically, when the control unit 9 classifies the frame F[nn] into the category C11, it determines that splashing is not detected, and when the control unit 9 classifies the frame F[nn] into the category C12, it determines that splashing is detected.

[0257] As described above, the frame F[nn] is classified into the class C11 or the class C12 by the classifier 91 generated by machine learning. Therefore, the frame can be classified with high classification accuracy. Furthermore, the splashing liquid can be detected with high detection accuracy.

[0258] In the second embodiment, a plurality of classifiers 91 may be generated similarly to the first embodiment, and part or all of the classifiers 91 and the machine learning unit 92 may be provided in the server 200 .

[0259] Modifications

[0260] In the above-described example, the process is performed using the two ejection nozzles 31 a and 31 b provided vertically above the substrate W. However, the present invention is not necessarily limited to this. Fig.24 1 is a diagram showing an example of the structure of a processing unit 1A. The processing unit 1A is the same as the processing unit 1 except for the presence or absence of a processing liquid supply unit 80. The processing liquid supply unit 80 has a spray nozzle 81, which is arranged on the side of the substrate W and at a position higher than the upper surface of the substrate W. The spray nozzle 81 sprays the processing liquid Lq3 in a substantially horizontal direction in such a manner that the processing liquid lands on the upper surface of the substrate W. The processing liquid Lq3 is released in an arc shape from the front end of the spray nozzle 81 and lands approximately near the center of the upper surface of the substrate W. The spray nozzle 81 is connected to the processing liquid supply source 84 via a pipe 82. An on-off valve 83 is provided in the middle of the pipe 82. By opening the on-off valve 83, the processing liquid Lq3 from the processing liquid supply source 84 flows through the inside of the pipe 82 and is sprayed from the spray nozzle 81.

[0261] In such a processing unit 1A, a discharge nozzle (for example, the discharge nozzle 31a) located above the substrate W and a discharge nozzle 81 located on the side of the substrate W are sometimes used to sequentially supply the processing liquid to the substrate W. Here, a case is described in which the substrate W is processed by the processing liquid Lq1 from the discharge nozzle 31a, and then the nozzle that discharges the processing liquid is switched from the discharge nozzle 31a to the discharge nozzle 81, and the substrate W is processed by the processing liquid Lq3 from the discharge nozzle 81.

[0262] The camera 70 is installed at a position where the tip of the discharge nozzles 31a, 31b, and 81 is included in the imaging area. Each frame of the imaging image generated by the camera 70 is sequentially output to the control unit 9.

[0263] Fig.25 1 is a diagram schematically showing an example of a frame IM6 of a captured image generated by the camera 70. In the frame IM6, the ejection nozzles 31a and 81 eject the processing liquids Lq1 and Lq3, respectively. That is, the frame IM6 is a frame at the time when the ejection nozzles 31a and 81 are switched and both eject the processing liquids.

[0264] In the frame IM6, liquid splashes are generated on the substrate W. The liquid splashes are also caused by the ejection of the processing liquid Lq1 and the processing liquid Lq3 by the ejection nozzles 31a and 81, respectively. In addition, the processing liquid Lq3 from the ejection nozzle 81 is ejected from the side of the substrate W toward the center of the substrate W, so that liquid splashes are likely to be generated on the side opposite to the ejection nozzle 81 relative to the center of the substrate W.

[0265] In the frame IM6, a discharge determination area Rc1 is set. The discharge determination area Rc1 is provided on the discharge path of the processing liquid Lq3 from the front end of the discharge nozzle 81 to the landing position on the substrate W, for example, extending from the front end of the discharge nozzle 81 to the direction in which the processing liquid Lq3 extends. The discharge determination area Rc1 has, for example, a rectangular shape.

[0266] Similar to the ejection nozzles 31a and 31b, the control unit 9 determines whether the treatment liquid Lq3 is ejected from the ejection nozzle 81 based on the size of the statistic of the pixel values ​​in the ejection determination area Rc1, and determines the start time of the ejection nozzle 81. The statistic mentioned here can be the sum or variance of the pixel values ​​in the ejection determination area Rc1. In addition, the ejection nozzle 81 ejects the treatment liquid Lq3 in the horizontal direction, so the unevenness of the brightness distribution in the horizontal direction is very small, and the characteristics of the treatment liquid Lq3 appear in the brightness distribution in the vertical direction. Therefore, the statistic can also be the variance of the pixels arranged in a row in the vertical direction. Alternatively, the pixel values ​​of the pixels arranged in the horizontal direction can be integrated for each row, and the variance of the integrated value of each row can be used.

[0267] Moreover, similarly to the first embodiment, the control unit 9 also determines the stop timing of the discharge nozzle 31 a.

[0268] The control unit 9 calculates the time difference between the start time of the ejection nozzle 81 and the stop time of the ejection nozzle 31a. For example, the time difference is calculated by subtracting the start time from the stop time. Then, the control unit 9 determines whether the time difference is outside the prescribed range. When the time difference is outside the prescribed range, the control unit 9 adjusts either the stop time of the ejection nozzle 31a or the stop time of the ejection nozzle 81 so that the time difference is within the prescribed range.

[0269] As described above, in the processing unit 1A, the time difference between the stop timing of the discharge nozzle 31 a and the start timing of the discharge nozzle 81 can be adjusted.

[0270] In addition, similarly to the second embodiment, it may be determined whether or not the processing liquid Lq3 is ejected from the ejection nozzle 81 by a classifier using machine learning.

[0271] In addition, Fig.25 In the example of FIG. 1 , a splash determination area R3 is provided. The splash determination area R3 is provided in the area where the substrate W is imaged and in an area on the opposite side of the ejection nozzle 81 with respect to the center of the substrate W. The splash determination area R3 has, for example, a rectangular shape.

[0272] Similar to the splash determination area R2, the control unit 9 determines whether there is splash based on the size of the statistic of the pixel values ​​in the splash determination area R3. The statistic mentioned here can be the sum or variance of the pixel values ​​in the splash determination area R3. Moreover, when splash occurs, the control unit 9 adjusts the time difference between the start time of the ejection nozzle 81 and the stop time of the ejection nozzle 31a in a way that does not cause splash.

[0273] In addition, similarly to the fourth embodiment, the presence or absence of liquid splashing may be determined by a classifier through machine learning.

[0274] The above is a description of the embodiment of the substrate processing device, but the embodiment can be modified in various ways other than the above embodiment as long as it does not deviate from its main purpose. The above various embodiments and modifications can be appropriately combined for implementation. For example, both the first embodiment and the second embodiment can be implemented, thereby performing both time difference monitoring and splash monitoring.

[0275] In addition, although a semiconductor substrate is used as the substrate W, the present invention is not limited thereto. For example, a glass substrate for a photomask, a glass substrate for a liquid crystal display, a glass substrate for a plasma display, a substrate for an FED (Field Emission Display), a substrate for an optical disk, a substrate for a magnetic disk, or a substrate for a magneto-optical disk may also be used.

[0276] Furthermore, the present embodiment can be applied to any device that ejects a processing liquid from a movable nozzle onto a substrate to perform a predetermined process. For example, in addition to the single-wafer cleaning processing device of the above-mentioned embodiment, the technology of the present embodiment can also be applied to a spin coating device (spin coater) that ejects a photoresist liquid from a nozzle onto a rotating substrate to coat an anti-etching agent, an apparatus that ejects a film removal liquid from a nozzle onto the edge of a substrate having a film formed on the surface, or an apparatus that ejects an etching liquid from a nozzle onto the surface of a substrate.

[0277] Description of Reference Numerals

[0278] 20: substrate holding part,

[0279] 30, 60, 65: Processing liquid supply unit,

[0280] 31a: first nozzle (spray nozzle),

[0281] 31b: second nozzle (spray nozzle),

[0282] 70: Camera,

[0283] 90: User interface,

[0284] 91, 91A~91C: Classifier,

[0285] 100: substrate processing equipment,

[0286] W: substrate.

Claims

1. A substrate processing device, wherein: have: a substrate holding portion for holding the substrate, a processing liquid supply unit including a first nozzle for spraying the processing liquid onto the substrate and a second nozzle for spraying the processing liquid onto the substrate, a camera that captures a capture area including a front end of the first nozzle and a front end of the second nozzle to generate a captured image, and Control Department; The control unit controls the processing liquid supply unit in such a manner that after starting to eject the processing liquid from the first nozzle to the substrate, the processing liquid starts to be ejected from the second nozzle to the substrate, and stops ejecting the processing liquid from the first nozzle to the substrate, and obtains a time difference between a start time of starting to eject the processing liquid from the second nozzle and a stop time of stopping ejecting the processing liquid from the first nozzle based on image processing of the captured image, and when it is determined that the time difference is outside a prescribed range, adjusts at least one of the start time and the stop time so that the time difference is within the prescribed range; The control unit classifies each frame included in the captured image into a category indicating a state of discharge or stop of the processing liquid for the first nozzle and the second nozzle using a classifier obtained through machine learning, and obtains the time difference based on the classification result.

2. The substrate processing apparatus according to claim 1, wherein: The control unit selects a classifier from a plurality of machine-learned classifiers corresponding to at least any one of the type of the substrate, the type of the processing liquid, the positions of the first nozzle and the second nozzle, and the flow rate of the processing liquid, and classifies each frame contained in the captured image based on the selected classifier.

3. The substrate processing apparatus according to claim 2, wherein: An input unit is provided for inputting at least one of the type of the substrate, the type of the processing liquid, the positions of the first nozzle and the second nozzle, and the flow rate of the processing liquid, The control unit selects a classifier from the plurality of classifiers based on an input to the input unit.

4. A substrate processing system, wherein: A substrate processing device and a server communicating with the substrate processing device are provided. The substrate processing device comprises: a substrate holding portion for holding the substrate, a processing liquid supply unit including a first nozzle for spraying the processing liquid onto the substrate and a second nozzle for spraying the processing liquid onto the substrate, a camera that captures a capture area including a front end of the first nozzle and a front end of the second nozzle to generate a captured image, and a control unit that controls the processing liquid supply unit in such a manner that, after starting to eject the processing liquid from the first nozzle toward the substrate, the processing liquid starts to be ejected from the second nozzle toward the substrate, and the ejection of the processing liquid from the first nozzle toward the substrate is stopped; The substrate processing device and the server use a machine-learned classifier to classify, for the first nozzle and the second nozzle, each frame included in the captured image into a category indicating a state of ejection or stop of the processing liquid, and based on the classification result, calculate a time difference between a start time of starting to eject the processing liquid from the second nozzle and a stop time of stopping ejecting the processing liquid from the first nozzle, When the control unit determines that the time difference is outside a predetermined range, the control unit adjusts at least one of the start time and the stop time so that the time difference falls within the predetermined range.

Images