Program stored in recording medium and substrate processing method

Abstract

The present invention provides a program stored in a recording medium, for selecting a reference wavelength and a reference time point for detecting an end point at which processing of a substrate by plasma is terminated. The program may perform: a data classification operation of classifying a plurality of pieces of sensor data collected by a sensor installed in a plasma chamber for processing the substrate into a first group and a second group that is a group different from the first group, wherein each piece of sensor data is data including a change in intensity of light according to time for each wavelength generated in the plasma chamber; a group selection operation of selecting a group to which a larger number of pieces of sensor data belongs from among the first group and the second group; a wavelength selection operation of selecting the reference wavelength for detecting the end point on the basis of selection sensor data that is sensor data belonging to a selection group selected in the group selection operation; and a time point selection operation of selecting the reference time point for detecting the end point on the basis of the selection sensor data.

Description

Method for processing programs and substrates stored in recording media

[0001] The present invention relates to a program stored in a recording medium and a substrate processing method, and more specifically, to a program stored in a recording medium for selecting a reference for an end point of a process when processing a substrate using plasma, and a substrate processing method for terminating a processing process for a substrate based on a reference for an end point selected from the program.

[0002] Semiconductor devices are typically manufactured by selectively and repeatedly performing processes such as photolithography, deposition, ion implantation, etching, and ashing on substrates such as wafers. Among these processes, etching and ashing, which remove film-like structures on the substrate, can be accomplished by exciting process gases into plasma and transferring the excited plasma to the film-like structures on the substrate.

[0003] In plasma processes that use plasma to remove film from substrates, detecting the end point of the process is crucial. Failure to accurately detect the end point can result in the process ending with less of the desired film removed, more film removed than intended, or excessive removal of the desired film.

[0004] In order to detect the end point (so-called endpoint) of the above process, an OES sensor (Optical Emission Spectroscopy Sensor) is attached to the process chamber where the plasma process is performed. Then, the OES sensor detects the light generated when the plasma and the film on the substrate react, and the endpoint is detected through the intensity of the detected light or the slope of the light intensity change graph. The criteria for detecting the endpoint (e.g., light intensity, slope, etc.) are generally stored in advance in the controller for controlling the process chamber. During the process, if the intensity of the light detected by the OES sensor reaches a preset reference value, the controller determines that the endpoint of the plasma process has been reached and stops the process in the process chamber.

[0005] However, these criteria are manually set by the operator. This means that the endpoint criteria used to determine that the plasma process has reached its endpoint can vary slightly depending on the operator's skill level and subjective judgment.

[0006] Furthermore, the wavelength of light primarily emitted when plasma and the film react varies depending on the type of film and the process gas used to remove it. Therefore, operators must have prior knowledge of the wavelength of light emitted based on the type of film and the process gas used to establish appropriate endpoint criteria.

[0007] The present invention aims to provide a program capable of selecting an endpoint criterion for detecting an endpoint of a process performed in a plasma chamber, and a substrate processing method capable of effectively processing a substrate through a reference wavelength and reference time point selected through the program.

[0008] In addition, the present invention aims to provide a program stored in a recording medium that can appropriately select an end point reference even without background knowledge about the type of film to be removed from a substrate and the wavelength of light generated depending on the type of process gas used, and a substrate processing method that processes a substrate through a reference wavelength and reference time point selected through the program.

[0009] The problems to be solved by the present invention are not limited to the problems described above, and problems not mentioned can be clearly understood by a person having ordinary skill in the technical field to which the present invention pertains from this specification and the attached drawings.

[0010] The present invention provides a program stored in a recording medium for selecting a reference wavelength and a reference time point for detecting an endpoint, which is a time point at which processing of a substrate by plasma is terminated. The program can perform a data classification operation for classifying a plurality of sensor data collected by a sensor installed in a plasma chamber for processing a substrate, each of the sensor data including a change in intensity of light over time for each wavelength occurring within the plasma chamber, into a first cluster and a second cluster which is a different cluster from the first cluster; a cluster selection operation for selecting a cluster to which a greater number of sensor data belongs among the first cluster and the second cluster; a wavelength selection operation for selecting the reference wavelength for detecting the endpoint based on selected sensor data, which is sensor data belonging to a selected cluster selected in the cluster selection operation; and a time selection operation for selecting a reference time point for detecting the endpoint based on the selected sensor data.

[0011] According to one embodiment, the wavelength selection operation may set a first section of the selection sensor data as a representative section, set a second section that is earlier in time than the first section as a comparison section, and select a wavelength with the greatest light intensity in the second section as the reference wavelength.

[0012] According to one embodiment, a plurality of second sections are set, and the wavelength selection operation may select the reference wavelength based on the second section having the farthest distance from the first section among the second sections.

[0013] In one embodiment, the wavelength selection operation may set the first section to the last 10% section of the selection sensor data.

[0014] According to one embodiment, a dead time is set to be a time set at the beginning of the selected sensor data and a time set at the end of the selected sensor data, and the wavelength selection operation can set the first section and the second section to a time other than the dead time.

[0015] According to one embodiment, the time point selection operation may select a time point at which the slope of the intensity of light of the reference wavelength in the first section of the selection sensor data enters a set slope as the reference time point.

[0016] According to one embodiment, the time point selection operation may select a time point at which the area of ​​the lower portion of the graph per unit time enters a set area in a graph for the intensity of light of the reference wavelength in the first section of the selection sensor data enters a set area as the reference time point.

[0017] According to one embodiment, the data classification operation may include, for each of the sensor data, dividing the entire portion of the sensor data into a plurality of sections, extracting a statistical value for each section - the statistical value including the maximum value, minimum value, and median value of the intensity of light by wavelength - and using a principal component analysis (PCA) technique for the extracted statistical value to reduce it into a one-dimensional vector, and using a clustering technique of any one of K-Means, GMM (Gaussian Mixture Model), or Agglomerative Hierarchical Clustering for the reduced one-dimensional vector to classify the sensor data into the first cluster or the second cluster.

[0018] In addition, the present invention provides a method for processing a substrate. The substrate processing method comprises: placing the substrate in a chamber, generating plasma to remove a film on the substrate, and terminating the process based on a reference wavelength or reference time point selected by a program stored in a recording medium; and collecting sensor data through an optical sensor installed in the chamber, wherein the sensor data is data including a change in intensity of light over time for each wavelength that occurs in the process of processing the substrate in the chamber; setting a first section as a representative section; and setting a second section preceding the first section in time as a comparison section; selecting a wavelength at which the intensity of light is the greatest in the second section as the reference wavelength; and selecting a time point at which the change in intensity of light at the reference wavelength in the first section becomes stable as the reference time point.

[0019] According to one embodiment, the stabilization point in time may be selected as the reference point in time when the slope of the light intensity enters a set slope.

[0020] According to one embodiment, the stabilization point in time may be selected as the reference point in time when the area of ​​the lower part of the graph for the intensity of the light per unit time enters a set area.

[0021] According to one embodiment, the sensor data may be collected in multiple pieces, the sensor data may be classified into multiple clusters based on similarity, and the sensor data used to select the reference wavelength and the reference time point may be data belonging to a cluster having the largest number of sensor data belonging to the clusters among the multiple clusters.

[0022] According to one embodiment, the point in time when the change in the intensity of light of the reference wavelength reaches a preset standard is determined as the endpoint, and if the difference between the point in time when the endpoint is reached and the reference point exceeds a threshold, it can be determined that the processing of the substrate is abnormal.

[0023] In one embodiment, if it is determined that the processing of the substrate has been performed abnormally, an interlock signal may be generated to stop the operation of the chamber or a visual or audible warning may be provided to the user.

[0024] According to one embodiment of the present invention, a substrate can be effectively treated.

[0025] In addition, according to one embodiment of the present invention, an endpoint criterion for detecting an endpoint of a process performed by a plasma chamber can be effectively selected even without any special background knowledge.

[0026] In addition, according to one embodiment of the present invention, the endpoint criterion for detecting the endpoint of a process performed by a plasma chamber can be selected relatively accurately regardless of the level of skill or background knowledge of the operator.

[0027] The effects of the present invention are not limited to the effects described above, and effects not mentioned can be clearly understood by a person having ordinary skill in the art to which the present invention pertains from this specification and the attached drawings.

[0028] Figure 1 is a conceptual diagram schematically illustrating a substrate processing device according to one embodiment of the present invention.

[0029] Figure 2 is a flow chart showing a substrate processing method according to one embodiment of the present invention.

[0030] Figure 3 is a flow chart showing the operations of a program stored in a recording medium that performs the endpoint criterion selection step of Figure 2.

[0031] Figure 4 is a diagram showing how sensor data is clustered into a first cluster and a second cluster when the cluster selection operation of Figure 3 is performed.

[0032] Figures 5 and 6 are graphs showing sensor data belonging to stable clusters.

[0033] Figures 7 and 8 are graphs showing sensor data belonging to an unstable cluster.

[0034] Fig. 9 is a graph for explaining a method of performing the wavelength selection operation of Fig. 3.

[0035] Figure 10 is a graph for explaining a method of performing the point selection operation of Figure 3.

[0036] The various features and advantages of the non-limiting embodiments of this disclosure will become more apparent upon review of the detailed description in conjunction with the accompanying drawings. The accompanying drawings are provided for illustrative purposes only and should not be construed as limiting the scope of the claims. The accompanying drawings are not to scale unless explicitly stated otherwise. Various dimensions in the drawings may be exaggerated for clarity.

[0037] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. These exemplary embodiments are provided so that this disclosure will be thorough and will fully convey the scope of the present disclosure to those skilled in the art. To provide a thorough understanding of the embodiments of the present disclosure, numerous specific details, such as examples of specific components, devices, and methods, are set forth. It will be apparent to those skilled in the art that specific details are not necessarily required, and that the exemplary embodiments can be implemented in many different forms, and neither should be construed as limiting the scope of the present disclosure. In some exemplary embodiments, well-known processes, well-known device structures, and well-known techniques are not described in detail.

[0038] The terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting of the example embodiments. As used herein, the singular or non-plural forms are intended to include the plural forms as well, unless the context clearly indicates otherwise. The terms "comprises," "comprising," "including," and "having" are open-ended and thus specify the presence of stated features, elements, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, elements, steps, operations, elements, components, and / or groups thereof. The method steps, processes, and operations herein are not necessarily to be construed as necessarily being performed in the particular order discussed or described, unless such order is explicitly stated. Additionally, additional or alternative steps may be selected.

[0039] When an element or layer is referred to as being "on," "connected," "joined," "attached," "adjacent," or "covering" another element or layer, it is intended that it is directly on, connected, joined, attached, adjacent, or covering said other element or layer, or that intermediate elements or layers may be present. Conversely, when an element is referred to as being "directly on," "directly connected to," or "directly coupled to" another element or layer, it should be understood that no intermediate elements or layers are present. Like reference numerals refer to like elements throughout the specification. The term "and / or" as used herein includes all combinations and subcombinations of one or more of the listed items.

[0040] Although terms such as first, second, third, etc. may be used herein to describe various elements, regions, layers, and / or sections, it should be understood that these elements, regions, layers, and / or sections are not limited by these terms. These terms are used merely to distinguish one element, region, layer, or section from another element, region, layer, or section. Thus, a first element, a first region, a first layer, or a first section discussed below could also be referred to as a second element, a second region, a second layer, or a second section without departing from the teachings of the exemplary embodiments.

[0041] Spatially relative terms (e.g., "beneath," "beneath," "lower," "above," "top," etc.) may be used for convenience of description to describe the relationship of one element or feature to other element(s) or features as depicted in the drawings. It should be understood that spatially relative terms are intended to encompass not only the orientation depicted in the drawings, but also other orientations of the device in use or operation. For example, if the device in the drawings were turned over, elements described as "beneath" or "below" other elements or features would then be oriented "above" the other elements or features. Thus, the term "beneath" can encompass both above and below orientations. The device can be oriented differently (rotated 90 degrees, or at other orientations), and the spatially relative descriptive phrases used herein can be interpreted accordingly.

[0042] When using the terms "same" or "same" in the description of embodiments, it should be understood that there may be some inaccuracy. Therefore, when one element or value is referred to as being the same as another element or value, it should be understood that the element or value is the same as the other element or value within a manufacturing or operating tolerance (e.g., ±10%).

[0043] When the terms "approximately" or "substantially" are used herein in connection with a numerical value, it should be understood that the numerical value includes manufacturing or operating tolerances (e.g., ±10%) of the stated value. Furthermore, when the terms "typically" and "substantially" are used in connection with geometrical shapes, it should be understood that geometrical accuracy is not required, but that latitude in the shape is within the disclosed scope.

[0044] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the exemplary embodiments pertain. Furthermore, terms, including terms defined in commonly used dictionaries, should be interpreted to have a meaning consistent with their meaning within the context of the relevant art, and will not be interpreted in an idealized or overly formal sense unless explicitly defined herein.

[0045] In the following, an example will be given in which the substrate (W) to be processed is a wafer.

[0046] Figure 1 is a conceptual diagram schematically illustrating a substrate processing device according to one embodiment of the present invention.

[0047] Referring to FIG. 1, a substrate processing device (1) according to one embodiment of the present invention may include a plasma chamber (10), a sensor (20), and a controller (30).

[0048] The plasma chamber (10) can process the substrate (W). The plasma chamber (10) may be a chamber that processes the substrate (W) by delivering plasma to the substrate (W). The plasma chamber (10) may be a device configured to perform a process of processing the substrate (W) using plasma. The process performed in the plasma chamber (10) may be a process of removing a film on the substrate (W) or a process of forming a film on the substrate (W). For example, the process may be an etching process or an ashing process that removes a film on the substrate (W). In addition, the process may be a deposition process or a passivation process that forms a film on the substrate (W).

[0049] The plasma chamber (10) may include various components for processing a substrate (W) using plasma. For example, the plasma chamber (10) may provide a processing space (11) in which a substrate (W) is processed, and may include a susceptor for supporting the substrate (W) in the processing space (11), a gas supply module for supplying process gas to the processing space, a plasma source for generating plasma in the processing space (11), and an exhaust unit for exhausting the processing space (11) to maintain the pressure of the processing space at a pressure close to vacuum.

[0050] The process gas supplied by the gas supply module can be varied according to the user's choice, depending on the type of film to be removed from the substrate (W) or the type of film to be formed on the substrate (W). The exhaust unit may include known devices capable of reducing the pressure in the processing space (11), such as a pump.

[0051] The plasma source may be a CCP (Conductively Coupled Plasma) type or an ICP (Inductively Coupled Plasma) type. In addition, the plasma chamber (10) can remove a film on the substrate (W) using a remote plasma method. For example, the plasma chamber (10) can remove ions from the generated plasma and remove a mask remaining on the substrate (W) using only radicals.

[0052] The optical sensor (20) can detect light generated in the processing space (11). The optical sensor (20) may be an OES (Optical Emission Spectroscopy) sensor. The optical sensor (20) may be installed in the plasma chamber (10) so as to sense light generated in the processing space (11) through a view port formed on the side of the plasma chamber (10). The optical sensor (20) may be a monochrome sensor, a CCD-OES (Charge Coupled Device - Optical Emission Spectroscopy) sensor, or a SCM-OES (Sensor Cluster Manager - Optical Emission Spectroscopy). However, the present invention is not limited thereto, and the optical sensor (20) may be variously modified into a known OES sensor.

[0053] The light sensor (20) can be electrically / communicatively connected to the controller (30). The light sensor (20) can sense the wavelength and intensity of light generated in the processing space (11) and transmit the sensed result to the controller (30). The controller (30) can collect data transmitted by the light sensor (20) over time. That is, the controller (30) can store sensor data (SD), which is data collected through the light sensor (20), in a form that includes changes in light intensity over time for each wavelength of light generated in the processing space (11).

[0054] The controller (30) may include a memory, a processor, a display, an interface unit, and a bus.

[0055] Various components such as memory, processor, display, and interface unit can be connected and communicated with each other (i.e., control message transmission and data transmission) by the bus.

[0056] The memory may include volatile memory (e.g., DRAM, SRAM, or SDRAM) and / or nonvolatile memory (e.g., one time programmable ROM (OTPROM), PROM, EPROM, EEPROM, mask ROM, flash ROM, flash memory, PRAM, RRAM, MRAM, a hard drive, or a solid state drive (SSD)). The memory may include internal memory and / or external memory. The memory may store, for example, instructions or data related to at least one other component of the electronic device. Additionally, the memory may store software and / or programs. The programs may include, for example, a kernel, middleware, an application programming interface (API), and / or an application program (or "application"). At least a portion of the kernel, middleware, or API may be referred to as an operating system.

[0057] In addition, the controller (30) may be provided with a non-transitory computer-readable medium. A non-transitory computer-readable medium refers to a medium that stores data semi-permanently and can be read by a computer, rather than a medium that stores data for a short period of time, such as a register, cache, or memory. Specifically, the various applications or programs described above may be stored and provided in a non-transitory computer-readable medium, such as a CD, DVD, hard disk, Blu-ray disk, USB, memory card, or ROM. Examples of program instructions include not only machine language codes created by a compiler, but also high-level language codes that can be executed by a computer using an interpreter, etc. The above-described hardware device may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

[0058] The processor may include one or more of a central processing unit, an application processor, or a communication processor (CP). The processor (20) may, for example, perform calculations or data processing related to control and / or communication of at least one other component of a computing device or a non-transitory computer-readable medium.

[0059] The display may include, for example, a liquid crystal display (LCD), a light emitting diode (LED) display, an organic light emitting diode (OLED) display, a micro electro mechanical systems (MEMS) display, or an electronic paper display. The display may, for example, display various content (e.g., text, images, videos, icons, and / or symbols) to a user. The display may include a touch screen and may receive touch, gesture, proximity, or hovering inputs, for example, using an electronic pen or a part of the user's body.

[0060] The interface unit enables the computing device to communicate with the outside world via a network. Here, the network includes both wired and wireless methods. In particular, the wireless communication may include cellular communication using at least one of LTE, LTE-A (LTE Advance), CDMA (code division multiple access), WCDMA (wideband CDMA), UMTS (universal mobile telecommunications system), WiBro (Wireless Broadband), or GSM (Global System for Mobile Communications). Alternatively, the wireless communication may include at least one of WiFi (wireless fidelity), LiFi (light fidelity), Bluetooth, Bluetooth low energy (BLE), Zigbee, near field communication (NFC), Magnetic Secure Transmission, Radio Frequency (RF), or Body Area Network (BAN). Alternatively, the wireless communication may include GNSS. GNSS may be, for example, GPS (Global Positioning System), Glonass (Global Navigation Satellite System), Beidou Navigation Satellite System (hereinafter "Beidou"), or Galileo, the European global satellite-based navigation system.Wired communication may include at least one of, for example, USB (universal serial bus), HDMI (high definition multimedia interface), RS-232 (recommended standard232), power line communication, or POTS (plain old telephone service), a computer network (e.g., LAN or WAN), etc.

[0061] The substrate processing method described below can be performed by the substrate processing device (1). In addition, the endpoint criteria used to detect the endpoint described below can be selected by an algorithm of a program recorded in a memory provided to the controller (30) or a non-transitory computer-readable medium. In addition, the endpoint criteria used to detect the endpoint can be selected by a program recorded in the controller (30), or alternatively, can be selected by a program stored in a computer device other than the controller (30) or in a medium readable by the computer device.

[0062] Figure 2 is a flow chart showing a substrate processing method according to one embodiment of the present invention.

[0063] Referring to FIG. 2, a substrate processing method according to an embodiment of the present invention may include a data collection step (S10), an endpoint criterion selection step (S20), and a process step (S30). The data collection step (S10), the endpoint criterion selection step (S20), and the process step (S30) may be performed sequentially.

[0064] The data collection step (S10) may be a step in which the light sensor (20) collects information about light generated in the processing space (11) of the plasma chamber (10). The data collected by the light sensor (20) may be sensor data (SD). The sensor data (SD) collected by the light sensor (20) may include information about changes in light intensity over time. In addition, the sensor data (SD) may include information about changes in light intensity over time for each wavelength of light generated in the processing space (11). For example, the sensor data (SD) may include changes in light intensity (Intensity) over time of light of wavelength A, changes in light intensity over time of light of wavelength B, changes in light intensity over time of light of wavelength C, etc.

[0065] The light generated in the processing space (11) may be light generated by a reaction between a film on a substrate (W) provided in the processing space (11) and plasma transferred to the substrate (W). The sensor data may include data on lights generated in the processing space (11) while processing one substrate (W). One piece of sensor data (SD) may include data on light generated while processing one substrate (W) in the processing space (11).

[0066] A plurality of sensor data (SD) can be collected. For example, the plasma chamber (10) can process a plurality of substrates (W), and each data collected while processing each substrate (W) can constitute one sensor data (SD). The controller (30) can collect tens to hundreds of data, or more sensor data (SD). The sensor data (SD) can all be data collected while removing the same recipe and the same type of film.

[0067] In the endpoint criterion selection step (S20), the endpoint criterion to be used in the process step (S30) for processing the substrate (W) can be selected based on the collected sensor data (SD).

[0068] The endpoint criteria selection step (S20) can select a reference wavelength and a reference time point. The endpoint criteria selection step (S20) can select a reference wavelength used to detect the endpoint and a reference time point used to determine that the endpoint has been reached.

[0069] The reference wavelength may be the wavelength of light primarily generated when the plasma reacts with the film to be removed on the substrate (W). The primarily generated wavelength here may refer to the wavelength at which the intensity of light generated when the film and plasma react is greatest. The reference point may be the point at which the removal of the film by the plasma is determined to be complete.

[0070] The endpoint criterion selection step (S20) can be selected by an algorithm of a program stored in the controller (30).

[0071] A detailed description of the endpoint criteria selection step (S20) is provided below.

[0072] In the process step (S30), the endpoint can be detected based on the reference wavelength and reference time point selected in the endpoint reference selection step (S20).

[0073] For example, in the process step (S30), a film on a substrate (W) is removed using plasma. When the film on the substrate (W) is removed using plasma, the light sensor (20) can detect light generated in the processing space (11). Information about the light detected by the light sensor (20) can be transmitted to the controller (30).

[0074] The controller (30) can monitor the change in intensity of light corresponding to a reference wavelength over time among the information about the light transmitted by the optical sensor (20). Then, when the change in light corresponding to the reference wavelength reaches a preset standard, it is determined that the end point of the process has been reached, and the plasma chamber (10) can be controlled to terminate the process performed in the plasma chamber (10) (for example, turning off the plasma). When the preset standard is reached, it may be when the slope of the change in intensity of the light of the reference wavelength reaches a preset slope. The preset standard for determining that the end point has been reached in the process step (S30) may be the same as the preset standard for determining that the end point has been reached in the time point selection operation (S24) described below.

[0075] In addition, if the time difference between the time when the change in light corresponding to the reference wavelength reaches a preset standard and the time when the reference time is pre-selected through sensor data (SD) is greater than a threshold value, the controller (30) determines that the process processing in the plasma chamber (10) has occurred abnormally, and can visually or audibly warn the user of this through a display or speaker provided in the controller (30), or can generate an interlock signal to stop the operation of the plasma chamber (10).

[0076] In some cases, the controller (30) may immediately terminate the process for the substrate (W) when the reference point is reached after the process has started.

[0077] Figure 3 is a flow chart showing the operations of a program stored in a recording medium that performs the endpoint criterion selection step of Figure 2.

[0078] Referring to FIG. 3, the end point criterion selection step (S20) can be performed by a program stored in a recording medium of the controller (30) or stored in a recording medium of a device other than the controller (30). The program for selecting the end point criterion can include a data classification operation (S21), a cluster selection operation (S22), a wavelength selection operation (S23), a time point selection operation (S24), and a reference value transmission operation (S25). The program for selecting the end point criterion can sequentially perform the data classification operation (S21), the cluster selection operation (S22), the wavelength selection operation (S23), the time point selection operation (S24), and the reference value transmission operation (S25).

[0079] The data classification operation (S21) can classify the sensor data (SD) collected in the data collection step (S10). In the data classification operation (S21), when selecting the endpoint criteria, data classification (S21) can be performed so that sensor data (SD) collected while performing an abnormal process can be excluded.

[0080] The sensor data (SD) collected in the data collection step (S10) are mostly data collected while normal processes are being performed, unless the plasma chamber (10) is defective. However, occasionally, an abnormal process may be performed in the plasma chamber (10) due to internal or external causes. The range of the light wavelength collected by the optical sensor (20), which is a commonly used OES sensor, may be 200 nm to 1000 nm, the number of variables may be more than 1500, and the process time may also vary depending on the recipe. Therefore, it takes a considerable amount of time for the user to classify the sensor data (SD) collected while normal processes are being performed and the sensor data (SD) collected while abnormal processes are being performed. Therefore, the following process may be performed to select the sensor data (SD) collected while an efficient normal process is being performed.

[0081] First, in the data classification operation (S21), the collected sensor data (SD) can be divided into multiple sections corresponding to 10% of the total process time. For example, the section of one sensor data (SD) can be divided into 10 sections. At this time, the total process time divided into multiple sections is excluded from the dead time by setting the set time after the plasma source is turned on and the set time before the plasma source is turned off. This is because, even in a normal process, a hunting phenomenon may occur due to a change in light intensity when the plasma source is turned on / off. In addition, for each section, statistical values ​​of the light intensity for each wavelength can be extracted. The statistical values ​​can include the maximum, minimum, and median values ​​of the light intensity for each wavelength for each section.

[0082] Second, the extracted statistical values ​​can be reduced to a one-dimensional vector using principal component analysis (PCA). PCA is a well-known dimensionality reduction method. PCA selects the unique principal components that best represent the characteristics of sensor data (SD) as representative principal component axes. It then projects the data along these principal component axes, identifying specific components that represent a wide range of data with minimal information loss.

[0083] Third, for the reduced one-dimensional vector, clusters are classified using one of the following clustering techniques: K-Means, GMM (Gaussian Mixture Model), or Agglomerative Hierarchical Clustering. These can be clustering techniques that divide similar data into a fixed number of clusters.

[0084] Figure 4 is a diagram showing how sensor data is clustered into a first cluster and a second cluster when the cluster selection operation of Figure 3 is performed.

[0085] Referring to FIGS. 3 and 4, the controller (30) classifies sensor data (SD: SD1, SD2, SD3, SDN) into a first cluster and a second cluster. If necessary, the sensor data (SD) may be classified into two or more clusters.

[0086] In general, the process in the plasma chamber (10) is generally performed as a normal process, but an abnormal process may rarely be performed. Therefore, the sensor data (SD) belonging to a cluster composed of a large number (for example, the largest number) will correspond to data collected while the normal process is performed, and the sensor data (SD) belonging to a cluster composed of a small number will correspond to data collected while the abnormal process is performed. For example, when the number of sensor data (SD) belonging to a first cluster is large and the number of sensor data (SD) belonging to a second cluster different from the first cluster is small, the sensor data (SD) belonging to the first cluster may correspond to data collected while the normal process is performed, and the sensor data (SD) belonging to the second cluster may correspond to data collected while the abnormal process is performed.

[0087] In the above example, the clustering technique of K-Means, GMM (Gaussian Mixture Model), or Agglomerative Hierarchical Clustering was used as an example to classify clusters, but it is not limited thereto, and various clustering methods that can classify sensor data (SD) into clusters based on similarity can also be applied.

[0088] The cluster selection operation (S22) may select a cluster that contains a greater number of sensor data (SD) among the clusters classified in the data classification operation (S21). For example, if the first cluster contains a greater number of sensor data (SD) among the first and second clusters, the program may select the first cluster. The sensor data (SD) belonging to the selected cluster selected in the cluster selection operation (S22) may be referred to as selected sensor data (SD) hereinafter.

[0089] Figures 5 and 6 are graphs showing sensor data belonging to a stable cluster, and Figures 7 and 8 are graphs showing sensor data belonging to an unstable cluster.

[0090] As can be seen from Figures 5 to 8, sensor data belonging to stable clusters (i.e., sensor data collected while a normal process is being performed) are identified in wavelengths where light intensity is significantly high. On the other hand, sensor data belonging to unstable clusters (i.e., sensor data collected while an abnormal process is being performed) have difficulty identifying wavelengths where light intensity is significantly high.

[0091] According to one embodiment of the present invention, there is an advantage in that the accuracy and reliability of the reference wavelength and reference time point can be further improved by classifying a plurality of sensor data (SD) collected in the data collection step (S10) according to similarity, selecting a cluster to which a plurality of sensor data belongs among the classified clusters, and performing the wavelength selection operation (S23) and time point selection operation (S24) described below through the sensor data (SD) in the selected cluster.

[0092] Fig. 9 is a graph for explaining a method of performing the wavelength selection operation of Fig. 3.

[0093] Referring to Fig. 9, the upper graph is a graph showing the change in light intensity in the wavelength range where the light intensity is observed to be the greatest among the information included in the sensor data (SD) collected by the light sensor (20). The lower graph is a graph showing the average value of the change in light intensity in the wavelength range where the light intensity is observed to be the greatest in each sensor data (SD), among the multiple sensor data (SD) provided within the selected cluster.

[0094] The wavelength selection operation (S23) selects a reference wavelength for detecting an endpoint based on sensor data belonging to a selected cluster selected in the cluster selection operation (S22).

[0095] More specifically, in the wavelength selection operation (S23), the total processing time of the selected sensor data (SD) is divided into multiple sections (SE1 to SE10). At this time, a time set at the beginning of the sensor data (SD) and a time set at the end of the sensor data (SD) are set as dead time, and the remaining time excluding the dead time is divided into multiple sections. The dead time can be specified as a set time after the plasma is turned on (i.e., the power of the plasma source is turned on) and a set time before the plasma is turned off (i.e., the power of the plasma source is turned off). The multiple sections divided in the wavelength selection operation (S23) may be the same as the multiple sections divided in the data classification operation (S21).

[0096] In the wavelength selection operation (S23), a section (SE10) belonging to the last 10% section among multiple sections (SE1 to SE10) can be set as a representative section (an example of the first section), and the remaining sections (SE 1 to SE 9) can be set as comparison sections (an example of the second section).

[0097] In the ashing process, over-ashing is performed to completely ash the thin film on the substrate (W), which typically occurs in the last 10% of the process. Therefore, the section (SE10) within the last 10% is set as the representative section.

[0098] In the representative section where over-ashing is performed, the film material to be removed remains almost non-existent. Conversely, in the comparison section that is the furthest in time from the representative section (e.g., SE1), the film material to be removed remains the greatest amount. Therefore, the intensity of the light generated by the plasma reaction with the film material is greatest in the SE1 section. Furthermore, the intensity of the light generated by the plasma reflection of other substances other than the film material to be removed may be difficult to detect.

[0099] Accordingly, in the wavelength selection operation (S23), the wavelength at which the light intensity is confirmed to be the greatest in the comparison section with the greatest temporal distance from the representative section (i.e., the wavelength at which the difference between the light intensity in the representative section and the light intensity in the comparison section with the greatest temporal distance from the representative section is confirmed to be the greatest) can be selected as the reference wavelength.

[0100] Through the wavelength selection operation (S23) as described above, the user can select a reference wavelength using only the sensor data (SD) collected while the plasma chamber (10) is being operated, and can also select a reference wavelength even without background knowledge about the wavelength of light generated when the film is removed.

[0101] Figure 10 is a graph for explaining a method of performing the point selection operation of Figure 3.

[0102] Referring to Fig. 10, in the time point selection operation (S24), a reference time point (ET) for detecting an endpoint is selected based on sensor data (SD). In the time point selection operation (S24), a time point at which the slope of the intensity of light of the reference wavelength (ET) in the first section (SE10) enters a set slope can be selected as the reference time point (ET). Alternatively, in the time point selection operation (S24), a time point at which the area under the graph per unit time of the graph enters a set area can be selected as the reference time point (ET). At this time,

[0103] Referring again to FIG. 3, in the reference value transmission operation (S25), the reference wavelength selected in the wavelength selection operation (S23) and the reference time point (ET) selected in the time point selection operation (S24) can be transmitted to a program for controlling the plasma chamber (10).

[0104] When using a program according to one embodiment of the present invention, a user can automatically and uniformly select an endpoint standard even without information about the type of film to be removed, the type of process gas used to remove the film, and the wavelength of light generated when the film is removed.

[0105] It should be understood that exemplary embodiments have been disclosed herein, and that other variations are possible. Individual elements or features of a particular embodiment are generally not limited to that particular embodiment, but, where applicable, may be interchangeable and used in a selected embodiment, even if not specifically illustrated or described. Such variations should not be considered a departure from the spirit and scope of the present disclosure, and all such modifications apparent to those skilled in the art are intended to be included within the scope of the following claims.

[0106] [Explanation of symbols]

[0107] Substrate processing unit: 1

[0108] Plasma Chamber: 10

[0109] Light sensor: 20

[0110] Controller: 30

[0111] Data collection phase: S10

[0112] Endpoint Criteria Selection Step: S20

[0113] Process stage: S30

[0114] Data classification operation: S21

[0115] Cluster selection operation: S22

[0116] Wavelength selection operation: S23

[0117] Point of View Selection Action: S24

[0118] Reference value transfer operation: S25

Claims

1. A program stored in a recording medium that selects a reference wavelength and reference time point for detecting an end point, which is the point at which treatment of a substrate by plasma ends. The above program: A data classification operation for classifying a plurality of sensor data collected by a sensor installed in a plasma chamber for processing a substrate, each of the sensor data including a change in intensity of light over time by wavelength occurring within the plasma chamber, into a first cluster and a second cluster that is a different cluster from the first cluster; A cluster selection operation for selecting a cluster to which a greater number of sensor data belongs among the first cluster and the second cluster; A wavelength selection operation for selecting the reference wavelength for detecting the endpoint based on the selected sensor data, which is sensor data belonging to the selected cluster selected in the above cluster selection operation; and A program stored in a recording medium that performs a point selection operation for selecting a reference point for detecting the endpoint based on the above-mentioned selected sensor data.

2. In paragraph 1, The above wavelength selection operation is, In the above-mentioned selected sensor data, the first section is set as the representative section, The second section, which is earlier in time than the first section, is set as the comparison section, A program stored in a recording medium that selects the wavelength with the greatest light intensity in the second section as the reference wavelength.

3. In paragraph 2, The above second section is set in multiple places, The above wavelength selection operation is a program stored in a recording medium that selects the reference wavelength based on the second section that is the farthest from the first section among the second sections.

4. In paragraph 2, The above wavelength selection operation is a program stored in a recording medium that sets the first section to the last 10% section of the selection sensor data.

5. In paragraph 2, Set the dead time to the time set at the beginning of the above-mentioned selected sensor data and the time set at the end of the above-mentioned selected sensor data, The above wavelength selection operation is a program stored in a recording medium that sets the first section and the second section to a time period other than the dead time.

6. In paragraph 2, The above time point selection operation is a program stored in a recording medium that selects a time point at which the slope of the light intensity of the reference wavelength in the first section of the selected sensor data enters a set slope as the reference time point.

7. In paragraph 2, The above time point selection operation is a program stored in a recording medium that selects a time point at which the area of ​​the lower part of the graph per unit time enters a set area in a graph for the intensity of light of the reference wavelength in the first section of the selected sensor data as the reference time point.

8. In paragraph 1, The above data classification operation is, For each of the above sensor data, the entire portion of the sensor data is divided into multiple sections, and for each section, statistical values ​​are extracted - the statistical values ​​include the maximum value, minimum value, and median value of the intensity of light by wavelength. The extracted statistical values ​​are reduced to a one-dimensional vector using the principal component analysis (PCA) technique, A program stored in a recording medium that classifies the sensor data into the first cluster or the second cluster by using any one of the clustering techniques of K_Means, GMM (Gaussian Mixture Model), or Agglomerative Hierarchical Clustering on the reduced one-dimensional vector.

9. In the method of processing the substrate, A process of placing the substrate in a chamber and generating plasma to remove a film on the substrate is performed, and the process is terminated based on a reference wavelength or reference time point selected by a program stored in a recording medium. In the sensor data collected through the optical sensor installed in the chamber - the sensor data is data including the change in light intensity over time by wavelength that occurs during the process of processing a substrate in the chamber - a first section is set as a representative section, and a second section that is earlier in time than the first section is set as a comparison section. In the second section above, the wavelength with the greatest light intensity is selected as the reference wavelength, A substrate processing method, wherein the point in time at which the change in the intensity of light of the reference wavelength is stabilized in the first section is selected as the reference point in time.

10. In paragraph 9, A substrate processing method in which the stabilization point is selected as the reference point when the slope of the light intensity enters a set slope.

11. In paragraph 9, A substrate processing method in which the stabilization point is selected as the reference point when the area of ​​the lower part of the graph per unit time enters a set area in the graph for the intensity of the light.

12. In paragraph 9, The above sensor data is collected in multiple pieces, The above sensor data is classified into multiple clusters based on similarity, A substrate processing method, wherein the sensor data used for selecting the reference wavelength and the reference time point is data belonging to a cluster having the largest number of sensor data among the plurality of clusters.

13. In paragraph 9, A substrate processing method, wherein the point in time when the change in the intensity of light of the reference wavelength reaches a preset standard is determined as an endpoint, and when the difference between the point in time when the endpoint is reached and the reference point exceeds a threshold, the processing of the substrate is determined to have been performed abnormally.

14. In paragraph 13, A substrate processing method, wherein, when it is determined that the processing of the above substrate is abnormal, an interlock signal is generated to stop the operation of the chamber or a visual or audible warning is given to the user.

Images