Monitoring method, monitoring program, monitoring device, wafer manufacturing method, and wafer

Abstract

To provide a monitoring method capable of improving quality of manufacture, a monitoring device, a manufacturing method of a wafer, and the wafer.SOLUTION: A monitoring method for monitoring a state of a processing device 10 based on achievement data acquired from the processing device 10 for processing a wafer W includes the steps of: inputting achievement data to an analysis model representing a relation between the achievement data and the state of the processing device 10 and acquiring at least one statistic representing the state of the processing device 10 outputted from the analysis model; determining whether or not the state of the processing device 10 is abnormal based on the at least one statistic; and identifying and outputting, as a cause item, an item which is a cause of abnormality, from among items of the achievement data in a case where it is determined that the state of the processing device 10 is abnormal.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a monitoring method, a monitoring program, a monitoring device, a wafer manufacturing method, and a wafer. [Background technology]

[0002] Conventionally, a method for predicting quality control value data of a wafer with high accuracy has been known (see Patent Document 1, etc.). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2007-242809 A Summary of the Invention [Problem to be solved by the invention]

[0004] Based only on the predicted results of product quality, process operators cannot understand the specific actions that should be taken on the processing equipment to improve the product quality. Product quality can be improved by enabling operators to understand the factors that affect product quality and take specific actions on the processing equipment.

[0005] Therefore, an object of the present disclosure is to propose a monitoring method, a monitoring program, a monitoring device, a wafer manufacturing method, and a wafer that can improve product quality. [Means for solving the problem]

[0006] One embodiment of the present disclosure that solves the above problem is as follows. [1] A monitoring method for monitoring a state of a processing device that processes a block or a wafer based on performance data acquired from the processing device, comprising: inputting the performance data into an analytical model that represents a relationship between the performance data and a state of the processing device, and acquiring at least one statistical value that represents the state of the processing device and is output from the analytical model; determining whether the state of the processing device is abnormal based on the at least one statistical value; a step of identifying an item causing the abnormality from among the items of the performance data as a cause item and outputting the item when it is determined that the state of the processing device is abnormal; A monitoring method including: [2] The monitoring method described in [1] above, further comprising a step of determining that the condition of the processing device is abnormal if the at least one statistical value is outside a control range. [3] estimating a quality of the wafer based on the at least one statistical value; and determining that the state of the processing device is abnormal when the estimated result of the quality of the wafer does not satisfy a quality standard; The monitoring method according to [1] or [2] above, further comprising: [4] acquiring lot data during a period in which the processing device processes one lot; dividing the lot data into a plurality of interval data for each interval into which a period for processing one lot is divided; inputting the data for each section as the performance data into the analysis model, and acquiring at least one statistical value representing a state of the processing device in each section; The monitoring method according to any one of [1] to [3] above, further comprising: [5] The monitoring method described in [4] above, further comprising a step of generating an interval model corresponding to each of the intervals as the analysis model by using the processed data of each of the intervals as teacher data, and inputting the interval data into the interval model. [6] The monitoring method described in [4] or [5] above, further comprising a step of determining whether the condition of the processing equipment is abnormal based on at least one statistical value obtained by inputting section data from the plurality of section data, which is in a section prior to the middle of the period for processing the one lot, into the analysis model. [7] The processing device is a wire saw device that cuts a block, The monitoring method according to any one of the above [4] to [6], further comprising a step of determining whether the state of the processing device is abnormal based on at least one statistical value obtained by inputting, into the analysis model, section data from among the plurality of section data, in a section in which cutting of the block proceeds a predetermined distance. [8] The monitoring method described in [7] above, wherein the specified distance is the length of the range in which values ​​of items representing characteristics of wafers obtained by cutting the block are calculated. [9] The monitoring method described in [8] above, wherein the item representing the characteristic of the wafer is the inclination angle of the cut surface of the wafer obtained by cutting the block at the start of cutting.

[10] A step of acquiring performance data during a period when the processing device is idle as idle operation data; inputting the idle operation data into the analysis model and obtaining at least one statistical value representative of a state of the processing device during idle operation; The monitoring method according to any one of [1] to [9] above, further comprising:

[11] The monitoring method described in

[10] above, further comprising a step of generating an idle operation model corresponding to a period during which the processing equipment is running idle as the analysis model by using the idle operation data as training data.

[12] A monitoring program that causes a processor to execute the monitoring method described in any one of [1] to

[11] above.

[13] A monitoring device comprising a control unit that executes the monitoring method described in any one of [1] to

[11] above.

[14] A method for manufacturing a wafer, comprising the step of processing a wafer using a wafer processing apparatus monitored by executing the monitoring method described in any one of [1] to

[11] above.

[15] A wafer processed by a wafer processing apparatus monitored by carrying out the monitoring method according to any one of [1] to

[11] above. Effect of the Invention

[0007] According to the monitoring method, monitoring program, monitoring device, wafer manufacturing method, and wafer disclosed herein, product quality can be improved. [Brief description of the drawings]

[0008] [Figure 1] 1 is a block diagram showing a configuration example of a monitoring system according to an embodiment of the present disclosure. [Diagram 2] FIG. 2 is a diagram showing a configuration example of a wire saw device as a processing device. [Diagram 3] 10 is a flowchart illustrating an example of a procedure for generating an analytical model in a monitoring method according to an embodiment of the present disclosure. [Figure 4] 10 is a flowchart illustrating an example of a procedure for determining a state of a processing device using an analytical model in a monitoring method according to an embodiment of the present disclosure. [Diagram 5] FIG. 13 is a diagram showing an example in which a period during which a block is cut by a wire saw device is divided into six sections. [Figure 6] FIG. 2 is a diagram showing an example of a surface shape of a wafer cut by a wire saw device. [Figure 7] 13 is a graph showing an example of the Q statistic value of each lot calculated based on the section data of the first section. [Figure 8] 13 is a graph showing an example of the Q statistic value of each lot calculated based on the section data of the second section. [Figure 9] 13 is a graph showing an example of the Q statistics of a cause item extracted from a lot in which the Q statistics is equal to or greater than a threshold value. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] (Example of monitoring system 1 configuration) As shown in FIG. 1, the monitoring system 1 includes a monitoring device 50 and a display device 60. The monitoring device 50 monitors the state of a processing device 10 that is set in a process for manufacturing a product. The processing device 10 processes the product. The processing device 10 has a sensor that measures processing data that represents the operating state of the components of the processing device 10. The sensor of the processing device 10 may measure the temperature or flow rate of auxiliary materials such as cooling water or slurry supplied to the processing device 10, or may measure items related to the process environment such as the process temperature or humidity. The sensor outputs the processing data measured when the processing device 10 operates to the monitoring device 50. The monitoring device 50 acquires the processing data from the processing device 10 and monitors the state of the processing device 10 based on the processing data.

[0010] The state of the processing apparatus 10 includes a normal state and an abnormal state. An abnormal state corresponds to a state that is not a normal state. A normal state corresponds to a state that does not deviate significantly from the state of the apparatus when there is no problem with the quality after processing of the wafer. For example, when focusing on one apparatus parameter, a state in which the deviation amount of the apparatus parameter from the average during a period in which there is no problem with the quality after processing of the wafer is less than three times the standard deviation of the apparatus parameter during a period in which there is no problem with the quality after processing of the wafer is considered to be a normal state. Also, even when a composite parameter such as T2 statistics or Q statistics is generated using multiple apparatus parameters, a state in which the deviation of the composite parameter calculated using the apparatus parameters during a period in which there is no problem with the quality after processing of the wafer is less than three times the standard deviation of the composite parameter during a period in which there is no problem with the quality after processing of the wafer is considered to be a normal state.

[0011] In this embodiment, the product processed by the processing device 10 is a wafer. The monitored object is installed in a wafer manufacturing process. The processing device 10 may be, for example, a wire saw device, a lapping processing device, an SMP (single-sided polishing device), a DSP (double-sided polishing device), an edge polishing device, or a heat treatment device.

[0012] The monitoring device 50 includes a control unit 52. The control unit 52 may include at least one processor. The processor may execute a program that realizes various functions of the control unit 52. The processor may be realized as a single integrated circuit. The integrated circuit is also called an IC (Integrated Circuit). The processor may be realized as a plurality of communicatively connected integrated circuits and discrete circuits. The processor may be realized based on various other known technologies.

[0013] The monitoring device 50 further includes a storage unit 54. The storage unit 54 may include an electromagnetic storage medium such as a magnetic disk, or may include a memory such as a semiconductor memory or a magnetic memory. The storage unit 54 may include a non-transitory computer-readable medium. The storage unit 54 stores various information such as processing data acquired from the processing device 10, and programs executed by the control unit 52. The storage unit 54 may function as a work memory for the control unit 52. At least a part of the storage unit 54 may be included in the control unit 52. At least a part of the storage unit 54 may be configured as a storage device separate from the monitoring device 50.

[0014] The monitoring device 50 may further include a communication unit that transmits and receives data to and from the processing device 10 or the display device 60. The communication unit is communicatively connected to the processing device 10 or the display device 60. The communication unit may be communicatively connected to the processing device 10 or the display device 60 via a network. The communication unit may be communicatively connected to the processing device 10 or the display device 60 by wire or wirelessly. The communication unit may include a communication module that connects to the network or the processing device 10 or the display device 60. The communication module may include a communication interface such as a LAN (Local Area Network). The communication module may realize communication by various communication methods such as 4G or 5G. The communication method executed by the communication unit is not limited to the above examples and may include various other methods. At least a part of the communication unit may be included in the control unit 52.

[0015] The display device 60 may include a display device that outputs visual information such as an image, text, or graphics. The display device may include, for example, an LCD (Liquid Crystal Display), an organic EL (Electro-Luminescence) display, an inorganic EL display, or a PDP (Plasma Display Panel). The display device is not limited to these displays and may include other types of displays. The display device may include a light-emitting device such as an LED (Light Emitting Diode) or an LD (Laser Diode). The display device is not limited to these displays and may include other types of devices.

[0016] (Configuration Example of Wire Saw Device as Processing Device 10) In this embodiment, it is assumed that the processing device 10 is a wire saw device. As shown in FIG. 2, the processing device 10 as a wire saw device includes a wire group 16 in which a wire 12 is stretched in parallel between a plurality of rollers 14 so as to be capable of reciprocating movement. The processing device 10 includes a workpiece holding mechanism 18 that holds the workpiece W and moves the workpiece W in a direction in which the workpiece W is pushed against the wire group 16. The processing device 10 includes a pair of nozzles 20 that supply slurry to an area of ​​the wire group 16 in which the workpiece W is pushed. The processing device 10 cuts the workpiece W with the wire group 16. It is assumed that the workpiece W is a block of silicon or the like (a single crystal ingot cut into a block shape). The processing device 10 delivers sliced ​​wafers of silicon or the like obtained by cutting the workpiece W as a processed product. Hereinafter, when the processing device 10 is a wire saw device, it is assumed that the processed product is a sliced ​​wafer.

[0017] The wire 12 is wound around a pair of wire reels 38A and 38B. The wire 12 is tensioned from one wire reel 38A through the guide roller 32, the roller 14, etc. to the other wire reel 38B.

[0018] The wire reels 38A and 38B are each rotated by a drive motor 36. When the drive motor 36 is driven to rotate the wire reels 38A and 38B, the wire 12 is unwound from one wire reel 38A and can run through the guide roller 32, roller 14, etc. to the other wire reel 38B. The wire 12 runs through a tension applying means including a dancer arm 33, dancer roller 34, etc. Tension is applied to the wire 12 as the wire 12 runs through the tension applying means. The wire 12 runs through a touch roller 35. The touch roller 35 follows the position of the wire 12 as it moves when it is unwound from the wire reels 38A and 38B or wound onto the wire reels 38A and 38B.

[0019] The wire 12 is wound helically multiple times across multiple rollers 14. The helically wound wire 12 constitutes a wire group 16 arranged in parallel between the rollers 14 in a direction perpendicular to the roller axial direction X. The rollers 14 are configured such that polyurethane resin is pressed around a steel cylinder, and grooves are cut into the surface at a fixed pitch. The wire 12 fits into the grooves cut into the surface of the rollers 14, allowing the wire group 16 to run stably.

[0020] The running direction of the wire 12 is controlled by the rotation direction of the drive motor 36. The wire 12 can be controlled to run in one direction, or can be controlled to run back and forth as necessary. The magnitude of tension applied to the wire 12 may be set as appropriate. The running speed of the wire 12 may be set as appropriate.

[0021] The slurry supplied from the nozzle 20 to the wire group 16 is stored in a slurry tank 40, and is sent from the slurry tank 40 to the nozzle 20 via a slurry chiller 42 that adjusts the temperature of the slurry.

[0022] The sliced ​​wafers discharged from the wire saw device are further processed in steps such as polishing, and are shipped as final wafer products.

[0023] (Example of operation of monitoring device 50) The monitoring device 50 statistically analyzes the processing data of the processing device 10 by multivariate analysis and calculates statistical values ​​of the processing data. The monitoring device 50 monitors the state of the processing device 10 based on the calculated statistical values. An example of the operation of the monitoring device 50 will be specifically described below.

[0024] <Creating an analysis model> The control unit 52 of the monitoring device 50 stores the processing data during the period when the processing device 10 is operating normally in the memory unit 54 as teacher data. The control unit 52 generates an analysis model used to statistically analyze the processing data of the processing device 10 and monitor the state of the processing device 10 based on the teacher data stored in the memory unit 54. The analysis model accepts input of the processing data of the processing device 10 to be analyzed. The processing data of the processing device 10 to be analyzed is also called performance data. The analysis model is configured to output a statistical value calculated by statistically analyzing the input performance data. In other words, the analysis model represents the relationship between the performance data and the state of the processing device 10. The analysis model is capable of determining at least whether or not the processing device 10 is abnormal.

[0025] In this embodiment, the analysis model is configured to calculate statistical values ​​of input performance data based on a PLS (Partial Least Square Regression) method. The control unit 52 generates the analysis model based on the PLS method as described below.

[0026] The control unit 52 selects explanatory variables and objective variables from each item of the processed data of the processing device 10 used as teacher data. The control unit 52 defines latent variables (principal components) based on the relationship between the values ​​of the explanatory variables and the objective variables included in the teacher data. Specifically, the latent variables are defined as new variables expressed by a relational expression that can calculate a value having a high correlation with the objective variable using the explanatory variables.

[0027] The control unit 52 generates a linear regression equation that expresses the relationship between the value of the latent variable and the value of the objective variable by applying the least squares method. The control unit 52 inputs the value of the explanatory variable, whose relationship with the value of the objective variable is known in advance, into the generated linear regression equation to obtain the calculation result of the linear regression equation. The control unit 52 verifies the linear regression equation by calculating the difference between the calculation result of the linear regression equation and the value of the objective variable. The control unit 52 verifies the linear regression equation by repeating the definition of latent variables, the generation of the linear regression equation, and the confirmation of the difference so that the difference between the calculation result of the linear regression equation and the value of the objective variable becomes small. In the verification of the linear regression equation, the control unit 52 may define new latent variables and replace them, or add newly defined latent variables to increase the number of latent variables. In the verification of the linear regression equation, the control unit 52 determines the number of latent variables so that the difference between the calculation result of the linear regression equation and the value of the objective variable becomes small, and determines the linear regression equation.

[0028] The control unit 52 uses the determined linear regression equation as an analysis model. Specifically, the control unit 52 calculates a T2 statistic and a Q statistic based on the determined linear regression equation and the performance data. The T2 statistic represents the position of the performance data in a subspace defined by the latent variable as a statistical value, and is calculated as the square of the distance from the origin of the subspace to a point in the performance data corresponding to the value of the latent variable. The Q statistic represents the position of the performance data in the orthogonal complement of the subspace as a statistical value, and is calculated as the square of the length of a vector orthogonally projected onto the subspace from a point corresponding to the value of an item in the performance data that is not used in the calculation of the latent variable.

[0029] The T2 statistical value is a statistical value suitable for determining whether something is normal or abnormal. The Q statistical value is a statistical value suitable for determining fluctuations that cannot be determined by the T2 statistical value. When only the T2 statistical value is monitored, if the T2 statistical value falls outside the control limit, the control unit 52 can determine that some abnormality has occurred in the processing device 10, but cannot analyze the specific content of the abnormality.

[0030] On the other hand, when the control unit 52 monitors both the T2 statistic and the Q statistic, if the Q statistic is within the control limit even when the T2 statistic exceeds the control limit, it can be determined that the correlation between variables has not broken down and no serious abnormality such as a device failure has occurred. Further, the control unit 52 can determine that the T2 statistic has increased as a result of changing or making a mistake in the processing conditions or operation program set in the processing device 10, etc., as the cause of the T2 statistic exceeding the control limit. When the Q statistic exceeds the control limit, the control unit 52 can determine that a serious abnormality such as a device failure has occurred.

[0031] The control unit 52 may perform cross-validation on the generated analysis model to prevent overfitting of the analysis model and improve its generalization performance.

[0032] <State determination based on the T2 statistic and the Q statistic> The control unit 52 inputs the actual data into the generated analysis model and calculates the T2 statistic and the Q statistic. The control unit 52 sets a control range for each of the T2 statistic and the Q statistic. The control unit 52 specifies the control range by setting the upper control limit or the lower control limit. If both the T2 statistic and the Q statistic are within the control range, the control unit 52 determines that the state of the processing device 10 is normal. If either the T2 statistic or the Q statistic is outside the control range, the control unit 52 determines that the state of the processing device 10 is abnormal or there is a possibility of abnormality. The control unit 52 outputs the determination result of the state of the processing device 10 to the display device 60. The display device 60 displays the determination result of the state of the processing device 10 and notifies the user. The user may include an operator or administrator of the process, etc.

[0033] When the Q statistic value is outside the control range, the control unit 52 extracts items that have an influence on the Q statistic value (that have caused the value of the Q statistic value to fluctuate significantly) from among the items of the performance data. The items that have an influence on the Q statistic value can be said to be items that cause the state of the processing device 10 to become abnormal, and are also called cause items. The Q statistic value calculated by inputting the performance data into the analysis model corresponds to the sum of the Q statistic values ​​of the cause items. In other words, the Q statistic value is divided into the Q statistic values ​​of the cause items.

[0034] It can be said that the larger the Q statistic value of each cause item, the greater the influence that cause item has on the Q statistic value. The control unit 52 may rank the cause items based on the magnitude of the influence that cause item has on the Q statistic value. The control unit 52 may rank the cause items with larger Q statistic values ​​higher. The control unit 52 outputs the cause items to the display device 60. The display device 60 displays the cause items and notifies the user.

[0035] The user may determine the measures to be taken for the processing apparatus 10 based on the judgment result of the state of the processing apparatus 10 notified by the display device 60. For example, when the Q statistical value is out of the control range, the user may check the actual state of the processing apparatus 10 for the notified cause item. For example, when the T2 statistical value is out of the control range, the user may check the processing conditions or the operating program set in the processing apparatus 10.

[0036] (Example of monitoring procedure) The control unit 52 of the monitoring device 50 may generate the analysis model by executing a monitoring method including the steps of the flowchart illustrated in Fig. 3. The monitoring method may be realized as a monitoring program executed by the control unit 52.

[0037] The control unit 52 acquires processing data during a period when the processing device 10 is operating normally as training data (step S1). The control unit 52 generates an analysis model based on the training data (step S2).

[0038] The control unit 52 determines whether the analytical model is appropriate (step S3). For example, the control unit 52 may determine whether the difference between the calculation result of the linear regression equation and the value of the objective variable is less than a predetermined value. If the control unit 52 determines that the analytical model is not appropriate (step S3: NO), the control unit 52 returns to step S2 and regenerates the analytical model. If the control unit 52 determines that the analytical model is appropriate (step S3: YES), the control unit 52 determines the analytical model (step S4). After executing the procedure of step S4, the control unit 52 ends the execution of the procedure of the flowchart in FIG. 3.

[0039] The control unit 52 may execute a monitoring method including the steps of the flowchart illustrated in Fig. 4 to calculate statistical values ​​based on performance data of the processing device 10 and monitor the state of the processing device 10. The monitoring method may be realized as a monitoring program executed by the control unit 52.

[0040] The control unit 52 acquires performance data from the processing device 10 (step S11). The control unit 52 inputs the performance data into an analysis model and acquires statistical values ​​calculated by the analysis model (step S12). Specifically, the control unit 52 acquires a T2 statistical value and a Q statistical value as the statistical values.

[0041] The control unit 52 determines whether the statistical value is within the control range (step S13). Specifically, the control unit 52 determines whether both the T2 statistical value and the Q statistical value are within the control range. If the statistical value is within the control range (step S13: YES), the control unit 52 proceeds to the procedure of step S15. If the statistical value is not within the control range (step S13: NO), that is, if the statistical value is outside the control range, the control unit 52 analyzes the cause item (step S14).

[0042] The control unit 52 outputs the determination result (step S15). The control unit 52 outputs whether the statistical value was within the control range or outside the control range. If the statistical value is outside the control range, the control unit 52 also outputs the analysis result of the cause item. After executing the procedure of step S15, the control unit 52 ends the execution of the procedure of the flowchart in FIG. 4.

[0043] (Small summary) As described above, the monitoring device 50 according to the present embodiment monitors the state of the processing device 10 based on the performance data acquired from the processing device 10. The monitoring device 50 inputs the performance data into the analysis model. The analysis model outputs at least one statistical value representing the state of the processing device 10. The monitoring device 50 acquires at least one statistical value from the analysis model and determines whether the state of the processing device 10 is abnormal based on the at least one statistical value. When the monitoring device 50 determines that the state of the processing device 10 is abnormal, it identifies and outputs an item that causes the abnormality from among the items of the performance data as a cause item. The user takes action on the processing device 10 based on the output cause item. In this way, the user can understand the factors that affect the quality of the product and take specific measures for the processing device 10, thereby improving the state of the processing device 10 at an early stage. As a result, the quality of the product processed by the processing device 10 can be improved.

[0044] (Example of monitoring the status of a wire saw device) When the processing device 10 is a wire saw device, the control unit 52 of the monitoring device 50 monitors the state of the device as described below. The wire saw device cuts a cylindrical block to generate sliced ​​wafers. The sliced ​​wafers are turned into product wafers through post-processing such as polishing. Processing data, which is performance data of the wire saw device, includes the temperature of each part of the device, the position of the wire guide (guide roller 32 and roller 14, etc.), and the flow rate of cooling water and slurry. In addition, the processing data of the wire saw device includes data on the length, resistance value, oxygen concentration, and crystal orientation of the processed block. In addition, the processing data of the wire saw device includes data on warps, undulations, and processing scratches of the cut wafers (sliced ​​wafers). The control unit 52 generates an analysis model based on each of the above items.

[0045] As shown in FIG. 5, the wire saw device cuts a block as a workpiece W along a cutting direction. The cutting direction is along the diameter of the cylindrical shape of the workpiece W. It takes a predetermined time for the wire saw device to cut one workpiece W. The period during which the wire saw device cuts one workpiece W may be divided into a plurality of sections. In the example of FIG. 5, the period during which the wire saw device cuts one workpiece W is divided from a first section at the start of cutting to a sixth section at the end of cutting. The control unit 52 may collectively acquire processing data over the entire period during which one workpiece W is cut. The processing data over the entire period is also referred to as lot data. The control unit 52 may acquire processing data divided into each section. The processing data divided into each section is also referred to as section data. The control unit 52 can acquire section data for each of the six sections in the example of FIG. 5.

[0046] The control unit 52 acquires the section data as teacher data and generates an analysis model corresponding to each section. The analysis model corresponding to each section is also called a section model. Even if the wire saw device is in the same state throughout the entire period in which one workpiece W is cut, the processing data in each of the first section where the wire saw device starts cutting, the second to fifth sections where the wire saw device stably proceeds with cutting, and the sixth section where the cutting ends may differ. By generating a different analysis model for each section, the accuracy of determining the state of each section of the wire saw device as the processing device 10 may be improved.

[0047] The control unit 52 acquires lot data as performance data. The control unit 52 divides the lot performance data into section data for each section. The control unit 52 inputs the section data as performance data into a section model to acquire a statistical value for each section. The control unit 52 may determine the state of the wire saw device based on the statistical value for each section. The control unit 52 may determine that the state of the wire saw device is abnormal when the statistical value for at least one section is outside the control range.

[0048] The control unit 52 inputs the data of the temperature of each part of the device, the position of the wire guide (guide roller 32 and roller 14, etc.), the flow rate of the cooling water and the slurry, and the length, resistance value, oxygen concentration, and crystal orientation of the machined block as the performance data into the analysis model. The control unit 52 acquires two statistical values, the T2 statistical value and the Q statistical value, from the analysis model. If the acquired statistical value is within the control range, the control unit 52 judges that the state of the wire saw device is normal. If the acquired statistical value is outside the control range, the control unit 52 extracts the cause item and notifies the user through the display device 60. For example, if the position of the wire guide is extracted as the cause item, the user prepares to adjust the position of the wire guide after the processing of the current lot is completed, and adjusts the position of the wire guide before the processing of the next lot is started. By allowing the user to prepare for the adjustment in advance, the operation rate of the wire saw device can be increased.

[0049] <Selection of the section to be analyzed> The control unit 52 may determine the state of the wire saw device based on a statistical value obtained by inputting section data of a section before the intermediate stage of the period for processing one lot, among a plurality of section data included in the period for processing one lot, into the analysis model as actual data. In the example of FIG. 5, the first to third sections correspond to the sections before the intermediate stage. When the control unit 52 determines that the state of the wire saw device is abnormal based on the statistical value of the section data of the section before the intermediate stage, it can provide the user with information on the abnormality by ensuring a sufficient period before starting the processing of the next lot. The user can ensure a preparation period for performing a procedure on the wire saw device before starting the processing of the next lot. In this way, the operation rate of the wire saw device can be increased.

[0050] The control unit 52 may set the first second section of the second to fifth sections in which the wire saw device stably cuts as the section for analysis. The variation in the processing data in the section in which the wire saw device stably cuts may be smaller than the variation in the processing data in the section at the start of cutting. The accuracy of the judgment may be improved by calculating statistical values ​​using the processing data with small variation as actual data to judge the state of the wire saw device. In addition, the stability of the analysis model may be improved by generating the analysis model using the processing data with small variation as teacher data.

[0051] When the processing device 10 is a wire saw device, the control unit 52 may input section data of a section in which the cutting of the block proceeds by a predetermined distance, among the multiple section data, into the analysis model as performance data to obtain at least one statistical value. The control unit 52 may determine whether the state of the processing device 10 is abnormal based on the obtained at least one statistical value. In other words, the control unit 52 may associate the section in which the cutting of the block proceeds by a predetermined distance with the section for analysis.

[0052] The sections for analysis may be divided based on the characteristics of the wafer. For example, as shown in FIG. 6, which corresponds to the cross section AA in FIG. 5, a wafer produced by cutting a block with a wire saw device has a surface irregularity represented by WS along the cutting direction. In FIG. 6, the horizontal axis corresponds to the position along the cutting direction of the wafer. The vertical axis corresponds to the height of the wafer at each position along the cutting direction of the wafer.

[0053] The shape of the surface along the cutting direction is divided into a section at the start of the cut (W1), a section at the end of the cut (W2), and a section in between (W0), and each section is approximated by a plane. The cutting surface has a different tendency in each section. Specifically, the middle section (W0) is approximated by a straight line (L0) drawn with two dots and dashes. The section at the start of the cut (W1) is approximated by a straight line (L1) drawn with two dots and dashes. The section at the end of the cut (W2) is approximated by a straight line (L2) drawn with two dots and dashes.

[0054] In the section (W1) at the beginning of cutting, the elevation difference of the surface shape is calculated. The elevation difference of the surface shape in the section (W1) at the beginning of cutting is calculated as the sum of the maximum length of the perpendicular line drawn from the point located above the approximation line (L1) among the points of the surface shape in the section (W1) at the beginning of cutting to the approximation line (L1) and the maximum length of the perpendicular line drawn from the point located below the approximation line (L1) to the approximation line (L1). Specifically, in FIG. 6, a dashed line is shown passing through the point of the surface shape in the section (W1) at the beginning of cutting that is the furthest from the approximation line (L1) in the vertical direction and is parallel to the approximation line (L1). The elevation difference of the surface shape in the section (W1) at the beginning of cutting is calculated as the distance between the upper and lower dashed lines. The elevation difference of the surface shape in the section (W2) at the end of the cutting is calculated as the sum of the maximum length of the perpendicular line drawn from the point located above the approximation line (L2) among the points of the surface shape in the section (W2) at the end of the cutting to the approximation line (L2) and the maximum length of the perpendicular line drawn from the point located below the approximation line (L2) to the approximation line (L2). Specifically, in FIG. 6, a dashed line is shown that passes through the point on the surface shape in the section (W2) at the end of the cutting that is the furthest from the approximation line (L2) in both the vertical direction and parallel to the approximation line (L2). The elevation difference of the surface shape in the section (W2) at the end of the cutting is calculated as the distance between the upper and lower dashed lines.

[0055] When the control unit 52 determines the section for analysis as the section in which the cutting of the block is advanced by a predetermined distance, the predetermined distance corresponds to the length of the section for calculating the value of the item representing the characteristic of the wafer obtained by cutting the block. The item representing the characteristic of the wafer corresponds to the inclination angle of the cut surface at the start or end of cutting of the wafer obtained by cutting the block. The predetermined distance corresponds to the length of the section represented by W1 or W2. The control unit 52 may determine the state of the wire saw device based on the statistical value calculated for the section data of the section for calculating the value of the item representing the characteristic of the wafer. In this way, the accuracy of determining the state of the wire saw device can be improved. As a result, the quality of the product can be improved.

[0056] <Analysis of machining data during idle operation> Maintenance operations such as component replacement or cleaning can be performed on the processing device 10. In order to check the state of the processing device 10 after the maintenance operation, the processing device 10 may be operated without loading the workpiece W. The operation without loading the workpiece W is also referred to as an idle operation.

[0057] The control unit 52 may generate an analysis model using the processing data when the processing device 10 is in an idle operation as teaching data. The processing data when the processing device 10 is in an idle operation is also referred to as idle operation data. The analysis model generated using the idle operation data as teaching data is also referred to as an idle operation model. Further, the control unit 52 may input the idle operation data as performance data into the idle operation model to obtain a statistical value. The control unit 52 may determine the state of the processing device 10 based on the statistical value of the idle operation data. By doing so, it becomes easier to avoid abnormalities that occur when the workpiece W is loaded and the processing device 10 is operated after the maintenance operation. As a result, it becomes easier to avoid abnormalities during the processing of the workpiece W. By avoiding abnormalities during processing, the quality of the product can be improved.

[0058] The control unit 52 may also select the idle operation data when no abnormality has occurred over a predetermined period after the idle operation of the processing device 10 and generate an idle operation model using it as teaching data. In other words, the control unit 52 may generate an idle operation model using the idle operation data excluding the idle operation data in which an abnormality has occurred within a predetermined period after the idle operation of the processing device 10 as teaching data. By doing so, the stability of the idle operation model can be enhanced.

[0059] <Example of monitoring the state of a wire saw device based on statistical values> The control unit 52 may calculate statistical values ​​for the lot data of each lot of the wire saw device, and determine whether the statistical values ​​of each lot are within the control range. For example, as shown as graphs in Figs. 7 and 8, the control unit 52 may determine whether the Q statistical value of each lot is less than a Q statistical threshold value (QTH). In Figs. 7 and 8, the horizontal axis represents the lot, and the vertical axis represents the Q statistical value. The graph in Fig. 7 represents the Q statistical value of each lot calculated from the section data of the first section in Fig. 5. The graph in Fig. 8 represents the Q statistical value of each lot calculated from the section data of the second section in Fig. 5.

[0060] The control unit 52 determines that, for a Q statistic value corresponding to a lot represented by X1, both the Q statistic value in the first section shown in FIG. 7 and the Q statistic value in the second section shown in FIG. 8 are equal to or greater than QTH. In this case, the control unit 52 may determine that the state of the wire saw device is abnormal when the lot represented by X1 is being processed. The control unit 52 also determines that, for a Q statistic value corresponding to a lot represented by X2, the Q statistic value in the first section shown in FIG. 7 is less than QTH, but the Q statistic value in the second section shown in FIG. 8 is equal to or greater than QTH. In this case as well, the control unit 52 may determine that the state of the wire saw device is abnormal when the lot represented by X1 is being processed.

[0061] The control unit 52 extracts cause items for a lot whose Q statistic value is equal to or greater than QTH. For example, the control unit 52 extracts cause items for a lot represented by X1. As shown in FIG. 9, four items represented by F1 to F4 are extracted as cause items. The horizontal axis in FIG. 9 represents the magnitude of the Q statistic value of each cause item. The control unit 52 may output the Q statistic value of each cause item to the display device 60, or may output cause items whose Q statistic value is equal to or greater than a predetermined value to the display device 60. The display device 60 displays cause items whose Q statistic value is large and notifies the user. The user may check cause items whose Q statistic value is large and determine and execute measures for the wire saw device.

[0062] The QTH may be set to, for example, the average value of the Q values ​​in past normal device states plus three times the standard deviation (+3σ). The QTH is not limited to this and may be set to various other values.

[0063] (Other embodiments) Other embodiments are described below.

[0064] <Product pass / fail judgement> The control unit 52 of the monitoring device 50 may generate an analytical model so as to estimate the quality of a product processed and discharged by the processing device 10 as a final product when the final process is completed, based on the performance data of the processing device 10. The control unit 52 may estimate, based on the quality estimation result, whether the product processed and discharged by the processing device 10 will satisfy the specifications when it becomes a final product. In other words, the control unit 52 may estimate the final pass / fail of the product processed and discharged by the processing device 10, based on the performance data of the processing device 10.

[0065] The monitoring device 50 may determine that the state of the processing device 10 is abnormal if it estimates based on the performance data that the product discharged during the processing for which the performance data was obtained will not satisfy the specifications when it becomes a final product. If the final product is a wafer, the monitoring device 50 may determine that the state of the processing device 10 is abnormal if the estimated result of the wafer quality does not satisfy the quality standard.

[0066] The control unit 52 may estimate the quality of the final product based on the section data of each section from the first section to the sixth section illustrated in FIG. 5, for example. The estimation result of the quality of the final product when the product processed and discharged by the processing device 10 is completed to the final process was confirmed to satisfy the specifications when the final process is actually completed to become the final product. The estimation result based on the section data of each section was compared with the actual result. As a result, the agreement rate between the estimation result based on the section data of each section from the first section to the sixth section and the actual result was as follows: first section: 96%, second section: 99%, third section: 97%, fourth section: 87%, fifth section: 97%, sixth section: 81%. When the control unit 52 judges the state of the processing device 10 by estimation based on the section data of each section from the first section to the third section, it can be said that the judgment can be made correctly with an accuracy of 97 to 99%. According to this data, it can be said that the accuracy of judging the state of the wire saw device based on the section data of the section before the middle of the period of processing one lot is sufficiently high. Therefore, the advantage of being able to judge the state of the equipment at an early stage of processing a lot and prepare for the next lot can be emphasized.

[0067] <Wafer manufacturing method and wafer> The processing apparatus 10 monitored by the monitoring apparatus 50 according to this embodiment executing the monitoring method is used to process a product. The processing apparatus 10 processes a wafer as a product, thereby manufacturing the wafer. Therefore, a wafer manufacturing method including a step of processing a wafer by the processing apparatus 10 monitored by the monitoring apparatus 50 executing the monitoring method is realized. Also, a wafer processed by the processing apparatus 10 monitored by the monitoring apparatus 50 executing the monitoring method is realized.

[0068] <Examples of products processed by the monitored entity> In this embodiment, the product is a wafer, but the product is not limited to this and may be various industrial products such as industrial parts or materials, or various other products such as food.

[0069] <Device configuration example> In the monitoring system 1, the monitoring device 50 may be included as a part of the processing device 10. The monitoring device 50 may be configured separately from the processing device 10.

[0070] <Example of Processing Data When Processing Apparatus 10 is a Polishing Apparatus> As described above, the processing device 10 is not limited to a wire saw device. When the processing device 10 is a polishing device such as a lapping processing device, an SMP (single-sided polishing device), or a DSP (double-sided polishing device), the processing data, which is the performance data, may include the temperature of the polishing liquid, the flow rate of the polishing liquid, the polishing pressure, or the horizontal movement speed.

[0071] Although the embodiments according to the present disclosure have been described based on the drawings and examples, it should be noted that a person skilled in the art can make various modifications or alterations based on the present disclosure. Therefore, it should be noted that these modifications or alterations are included in the scope of the present disclosure. For example, the functions included in each component or each step can be rearranged so as not to be logically inconsistent, and multiple components or steps can be combined into one or divided. Although the embodiments according to the present disclosure have been described mainly with respect to the device, the embodiments according to the present disclosure can also be realized as a method including steps executed by each component of the device. The embodiments according to the present disclosure can also be realized as a method, a program, or a storage medium on which a program is recorded, executed by a processor included in the device. It should be understood that these are also included in the scope of the present disclosure.

[0072] The graphs included in this disclosure are schematic and do not necessarily correspond to the actual scale. [Industrial Applicability]

[0073] According to embodiments of the present disclosure, product quality may be improved. [Explanation of symbols]

[0074] 1. Surveillance System 10 Processing device (12: wire, 14: roller, 16: wire group, 18: work holding mechanism, 20: nozzle, 32: guide roller, 33: dancer arm, 34: dancer roller, 35: touch roller, 36: drive motor, 38A, 38B: wire reel, 40: slurry tank, 42: slurry chiller, W: work (block), X: roller axis direction) 50 monitoring device (52: control unit, 54: storage unit) 60 Display device

Claims

1. 1. A monitoring method for monitoring a state of a processing device that processes a block or a wafer based on performance data acquired from the processing device, comprising: a step of inputting the performance data into an analytical model based on a partial least square regression (PLS) method, which is generated by using the processing data of the processing device as teacher data so as to represent a relationship between the performance data and a state of the processing device, in which a latent variable is defined based on a relationship between a value of an explanatory variable included in the teacher data and a value of a target variable, and acquiring a T2 statistic value representing a position of the performance data in a subspace defined by the latent variable as at least one statistical value representing the state of the processing device output from the analytical model; determining whether the condition of the processing device is abnormal based on the at least one statistical value; a step of identifying an item causing the abnormality from among the items of the performance data as a cause item and outputting the item when it is determined that the state of the processing device is abnormal; A monitoring method including:

2. The monitoring method described in claim 1, further comprising a step of obtaining, as the at least one statistical value, a Q statistic value representing the position of the actual data in an orthogonal complement of a subspace defined by the latent variables.

3. estimating a quality of the wafer based on the at least one statistical value; and determining that the state of the processing device is abnormal when the estimated result of the quality of the wafer does not satisfy a quality standard; The monitoring method according to claim 1 or 2, further comprising:

4. acquiring lot data during a period in which the processing device processes one lot; dividing the lot data into a plurality of interval data for each interval obtained by dividing a period during which the one lot is processed into a plurality of intervals; inputting the data for each section as the performance data into the analysis model, and acquiring at least one statistical value representing a state of the processing device in each section; The monitoring method according to claim 1 , further comprising:

5. 1. A monitoring method for monitoring a state of a processing device that processes a block or a wafer based on performance data acquired from the processing device, comprising: inputting the performance data into an analytical model that represents a relationship between the performance data and a state of the processing device, and acquiring at least one statistical value that represents the state of the processing device and is output from the analytical model; estimating a quality of the wafer based on the at least one statistical value; and determining that the state of the processing device is abnormal when the estimated result of the quality of the wafer does not satisfy a quality standard; a step of identifying an item causing the abnormality from among the items of the performance data as a cause item and outputting the item when it is determined that the state of the processing device is abnormal; A monitoring method including:

6. A monitoring method for monitoring a state of a processing device, which is a wire saw device that cuts a block, based on performance data acquired from the processing device, comprising: acquiring lot data during a period in which the processing device processes one lot; dividing the lot data into a plurality of interval data for each interval obtained by dividing a period during which the one lot is processed into a plurality of intervals; inputting the data for each section as the performance data into an analytical model that represents a relationship between the performance data and a state of the processing device, and acquiring at least one statistical value that represents the state of the processing device in each section and is output from the analytical model; a step of inputting section data of the plurality of section data in a section in which cutting of the block proceeds by a predetermined distance into the analysis model and determining whether or not the state of the processing device is abnormal based on at least one statistical value acquired by the input data; a step of identifying an item causing the abnormality from among the items of the performance data as a cause item and outputting the item when it is determined that the state of the processing device is abnormal; Including, The predetermined distance is a length of a range for calculating values ​​of items representing characteristics of wafers obtained by cutting the block. Monitoring method.

7. 7. The monitoring method according to claim 6, wherein the item representing the characteristic of the wafer is an inclination angle of a cut surface of the wafer obtained by cutting the block at the start of cutting.

8. The monitoring method according to claim 6 or 7, further comprising the step of generating a section model corresponding to each of the sections as the analysis model by using the processed data of each of the sections as teacher data, and inputting the section data into the section model.

9. 9. The monitoring method according to claim 6, further comprising a step of determining whether the state of the processing device is abnormal based on at least one statistical value acquired by inputting section data from the plurality of section data, the section data being in a section preceding a midpoint of a period for processing the one lot, into the analysis model.

10. 1. A monitoring method for monitoring a state of a processing device that processes a block or a wafer based on performance data acquired from the processing device, comprising: A step of acquiring performance data during a period when the processing device is idle as idle operation data; A step of inputting the idle operation data as the performance data into an analysis model that represents a relationship between the performance data and a state of the processing device, and acquiring at least one statistical value that represents a state of the processing device during idle operation that is output from the analysis model; determining whether the condition of the processing device is abnormal based on the at least one statistical value; a step of identifying an item causing the abnormality from among the items of the performance data as a cause item and outputting the item when it is determined that the state of the processing device is abnormal; A monitoring method including:

11. The monitoring method according to claim 10, further comprising the step of generating, as the analysis model, an idle operation model corresponding to a period during which the processing device is idle by using the idle operation data as teacher data.

12. The monitoring method according to claim 1 , further comprising the step of determining that the state of the processing device is abnormal if the at least one statistical value is outside a control range.

13. A monitoring program that causes a processor to execute the monitoring method according to any one of claims 1 to 12.

14. A monitoring device comprising a control unit that executes the monitoring method according to any one of claims 1 to 12.

15. A method for manufacturing a wafer, comprising the step of processing a wafer with a wafer processing device monitored by carrying out the monitoring method according to any one of claims 1 to 12.

Images