State prediction device, state prediction method, and state prediction system

Abstract

An aspect relates to generating highly accurate state prediction results for a semiconductor manufacturing device, the state prediction device for the semiconductor manufacturing device includes a data acquisition unit for acquiring a first set of operational data for a first processing chamber and a second set of operational data for a second processing chamber, a feature management unit for generating first and second feature maps, a correlation calculation unit for calculating a normalized cross-correlation result indicative of a uniformity level of target features between the first and second feature maps, a ranking unit for ranking the target features based on the normalized cross-correlation result and selecting a subset of the target features that reach a ranking threshold, and a state prediction unit for generating a state prediction result characterizing a performance difference of the second processing chamber relative to the first processing chamber.

Description

Technical Field

[0001] The present disclosure relates to a state prediction device, a state prediction method, and a state prediction system.

Background Art

[0002] In recent years, in the manufacturing industry, efforts to improve productivity by effectively using data obtained from manufacturing devices have attracted attention.

[0003] As an example, in the field of semiconductor manufacturing, a plasma processing device may be equipped with a large number of sensors for data acquisition. The data obtained from these sensors can be used for early detection of device abnormalities and for improving productivity.

[0004] When processing a time-series signal obtained by a sensor, features representing individual measurable attributes or characteristics of the signal are extracted and analyzed. For example, "plasma impedance" can be a feature extracted from a time-series signal for analysis.

[0005] Generally, when performing analysis of a time-series signal to determine the operating state of a plasma processing device, for example, extracting more features can facilitate the generation of more reliable and accurate state prediction results. However, the larger the number of features to be analyzed, the more problems including an increase in computer resources may occur, the processing time may become longer, and the amount of training data may also increase.

[0006] Conventionally, techniques for reducing the number of features used for state prediction have been considered. As an example, Patent Document 1 discloses that "a state prediction device for predicting the state of a plasma processing apparatus is provided. A first set of features indicating the state of the plasma processing apparatus is determined based on monitored data of a normal-state plasma processing apparatus. A second set of features indicating the state of the plasma processing apparatus is determined based on the monitored data of the plasma processing apparatus. The features of the second set are calculated by using the features of the first set. A model for predicting the state of the plasma processing apparatus is generated by using a subset of the first set of features, which consists of the same type of features selected in descending order of the calculated features of the second set. The state of the plasma processing apparatus is predicted by using the generated model."

Prior Art Documents

Patent Documents

[0007]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0008] Patent Document 1 proposes a technique for predicting the state of a plasma processing device by using a subset of the features of the signals obtained from the plasma processing device. The subset of features can be ranked in the order of the standardized values indicating the degree of deviation of the test data from the normal device state.

[0009] However, the technique of Patent Document 1 performs device state prediction based on the standardized values indicating the degree of deviation of the test data from the normal device state, without considering the uniformity of the features between the data corresponding to the acceptable chamber state and the unacceptable chamber state. By selecting features considering the uniformity between the acceptable chamber state and the unacceptable chamber state, it is possible to obtain more accurate state prediction results.

Means for Solving the Problems

[0010] Accordingly, aspects of the present disclosure relate to a state prediction technique that can generate a high-precision state prediction result for a semiconductor manufacturing device based on a feature determined to have high uniformity between an acceptable chamber state and an unacceptable chamber state.

[0011] A representative example of the present disclosure is a state prediction device. The state prediction device includes a data acquisition unit configured to acquire a first set of operation data for a first semiconductor manufacturing device that reaches an operation threshold value and a second set of operation data for a second semiconductor manufacturing device that does not reach the operation threshold value; a feature management unit configured to generate a first feature map for a first target feature based on the first set of operation data and generate a second feature map for the first target feature based on the second set of operation data; a correlation calculation unit configured to calculate a normalized cross-correlation result indicating a uniformity level of the first target feature between the first feature map and the second feature map; a ranking unit configured to assign a rank to a first target feature indicating the suitability of the set of target features for semiconductor manufacturing device state prediction based on the normalized cross-correlation result and select a subset of target features that reach a ranking threshold value from among the set of target features; and a state prediction unit configured to generate a state prediction result characterizing the operation state of the second semiconductor manufacturing device based on the subset of target features.

Advantages of the Invention

[0012] According to the present disclosure, it is possible to provide a state prediction technique that can generate a high-precision state prediction result for a semiconductor manufacturing device based on a feature determined to have high uniformity between an acceptable chamber state and an unacceptable chamber state.

[0013] Problems, configurations, and effects other than those described above will become clear from the following description of embodiments for carrying out the present invention.

Brief Description of the Drawings

[0014]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

[0015] In this specification, embodiments of the present invention are described with reference to the drawings. It should be noted that the embodiments described in this specification are not intended to limit the present invention according to the claims, and it should be understood that each of the elements described in the embodiments and combinations thereof are not strictly necessary for implementing aspects of the present invention.

[0016] In the following description and the associated drawings, various aspects are disclosed. Alternative aspects can be devised without departing from the scope of the present disclosure. In addition, well-known elements of the present disclosure are not described in detail or are omitted in order not to obscure the relevant details of the present disclosure.

[0017] The terms "exemplary" and / or "example" are used herein to mean "serving as an example, instance, or illustration." Any aspect described herein as "exemplary" and / or "example" should not necessarily be construed as preferred or advantageous over other aspects. Similarly, the expression "aspect of the present disclosure" does not require that all aspects of the present disclosure include the features, advantages, or operational characteristics being discussed.

[0018] Furthermore, for example, many aspects are described regarding the order of actions performed by elements of an arithmetic unit. It will be recognized that the various actions described herein can be performed by a specific circuit (e.g., an application-specific integrated circuit (ASIC)), program instructions executed by one or more processors, or a combination of both. Additionally, the order of actions described herein can be embodied as a whole within any form of a computer-readable storage medium that stores a corresponding set of computer instructions that, when executed, can cause the associated processor to perform the functions described herein. Therefore, the various aspects of the present disclosure may be embodied in many different forms, all of which are intended to be within the scope of the subject matter recited in the claims.

[0019] In this specification, a detailed description of embodiments of the present disclosure is described with reference to the drawings.

[0020] Generally, the performance of semiconductor manufacturing devices (such as plasma processing devices) can change due to aging, component replacement, cleaning, or other factors. Such performance changes can appear in the performance of semiconductor devices manufactured by those semiconductor manufacturing devices. Accordingly, a state prediction system according to an embodiment of the present disclosure (e.g., the state prediction system 200 shown in FIG. 2) is configured to identify and correct performance differences between processing chambers (hereinafter also referred to as chambers) of one or more semiconductor manufacturing devices. This performance difference may occur in a single chamber due to the passage of time, component replacement or component cleaning of the chamber, or between different chambers (e.g., different chambers in different semiconductor manufacturing devices, etc.).

[0021] Here, the "performance difference" between chambers refers to the difference in processing results (e.g., difference in etching amount, difference in thickness of the formed film) between plasma processing (e.g., etching, film formation) performed using a reference chamber and processing using a target chamber. This performance difference between plasma processing using the plasma generated in the reference chamber and plasma processing using the plasma generated in the target chamber can be quantified by comparing semiconductor devices manufactured by a semiconductor manufacturing device including the reference chamber and semiconductor devices manufactured by a semiconductor manufacturing device including the target chamber. That is, the difference in performance between plasma processing using the plasma generated in the reference chamber and plasma processing using the plasma generated in the target chamber appears in the variation in performance of the manufactured semiconductor devices.

[0022] The performance difference due to the passage of time refers to the difference between the processing result of the processing performed in a specific chamber at a first time point and the processing result of the processing performed in the same chamber at a second time point later than the first time point.

[0023] The performance difference due to component replacement or component cleaning refers to the difference between the processing result of the processing performed in a specific chamber before component replacement or cleaning and the processing result of the processing performed in the same chamber after replacement or cleaning.

[0024] The state prediction device, system, and method according to the present disclosure relate to a technique for generating a state prediction result characterizing a performance difference of a target processing chamber (e.g., a second processing chamber) with respect to a reference processing chamber (e.g., a first processing chamber), and adjusting an operation parameter of the target processing chamber so as to reduce the performance difference of the target processing chamber with respect to the reference processing chamber using this state prediction result. As described herein, this state prediction result can be generated based on a feature having high uniformity between an acceptable chamber state and an unacceptable chamber state. Here, the "acceptable chamber state" refers to a processing chamber that has reached a predetermined performance level, and the "unacceptable chamber state" refers to a processing chamber that has not reached a predetermined performance level. In this way, by generating a state prediction result using a feature having high uniformity between an acceptable chamber state and an unacceptable chamber state, it becomes possible to align the target processing chamber with a reference processing chamber (for example, achieve performance results similar to those of the reference processing chamber).

[0025] Next, referring to the drawings, FIG. 1 is a schematic block diagram of a computer system 100 for implementing various embodiments of the present disclosure according to an embodiment. The mechanisms and apparatuses of the various embodiments disclosed herein are equally applicable to any suitable computing system. The main components of the computer system 100 include one or more processors 102, a memory 104, a terminal interface 112, a storage interface 113, an I / O (input / output) device interface 114, and a network interface 115, all of which are communicatively coupled, directly or indirectly, for component-to-component communication via a memory bus 106, an I / O bus 108, a bus interface unit 109, and an I / O bus interface unit 110.

[0026] The computer system 100 may include one or more general-purpose programmable central processing units (CPUs) 102A and 102B, generally referred to herein as processors 102. In an embodiment, the computer system 100 may include multiple processors, but in a particular embodiment, the computer system 100 may alternatively be a single CPU system. Each processor 102 executes instructions stored in the memory 104 and may include one or more levels of on-board cache.

[0027] In an embodiment, the memory 104 may include a random access semiconductor memory, a storage device, or a storage medium (either volatile or non-volatile) for storing or encoding data and programs. In a particular embodiment, the memory 104 represents the entire virtual memory of the computer system 100 and may further include the virtual memory of other computer systems coupled to or connected via a network to the computer system 100. Conceptually, the memory 104 can be viewed as a single monolithic entity, but in other embodiments, the memory 104 has a more complex configuration, such as a hierarchy of caches and other memory elements. For example, the memory may exist in multiple levels of cache, and those caches may be further partitioned by function, such that one cache holds instructions and another holds data other than the instructions used by the processor. The memory may further be distributed and associated with different CPUs or sets of CPUs, as is known in any of various so-called non-uniform memory access (NUMA) computer architectures.

[0028] Memory 104 may store all or a portion of various programs, modules, and data structures for processing the data transfers described herein. For example, memory 104 may store the state prediction application 150. In an embodiment, the state prediction application 150 may include instructions or statements that are executed on the processor 102, or instructions or statements that are interpreted by instructions or statements that are executed on the processor 102 to perform functions as further described below. In a particular embodiment, the state prediction application 150 may be implemented in hardware via semiconductor devices, chips, logic gates, circuits, circuit cards, and / or other physical hardware devices, instead of or in addition to a processor-based system. In an embodiment, the state prediction application 150 may include data in addition to instructions or statements. In a particular embodiment, a camera, sensor, or other data input device (not shown) may be provided in direct communication with the bus interface unit 109, processor 102, or other hardware of the computer system 100. In such a configuration, the need for the processor 102 to access the memory 104 and the state prediction application 150 may be reduced.

[0029] Computer system 100 may include a bus interface unit 109 that handles communication between a processor 102, a memory 104, a display system 124, and an I / O bus interface unit 110. The I / O bus interface unit 110 may be coupled to an I / O bus 108 to transfer data between various I / O units. The I / O bus interface unit 110 communicates via the I / O bus 108 with a plurality of I / O interface units 112, 113, 114, and 115, also known as I / O processors (IOPs) or I / O adapters (IOAs). The display system 124 may include a display controller, a display memory, or both. The display controller may provide data of types such as video, audio, or both to a display device 126. Further, computer system 100 may include one or more sensors or other devices configured to collect data and provide it to processor 102. By way of example, computer system 100 may include biometric sensors (e.g., collecting heart rate data, stress level data), environmental sensors (e.g., collecting humidity data, temperature data, pressure data), motion sensors (e.g., collecting acceleration data, movement data), etc. Other types of sensors are also possible. The display memory may be dedicated memory for buffering video data. The display system 124 may be coupled to a display device 126 such as a stand-alone display screen, a computer monitor, a television, a tablet, or a display of a handheld device. In one embodiment, the display device 126 may include one or more speakers for rendering audio. Alternatively, one or more speakers for rendering audio may be coupled to an I / O interface unit. In an alternative embodiment, one or more of the functions provided by the display system 124 may be implemented on an integrated circuit that also includes processor 102. Additionally, one or more of the functions provided by the bus interface unit 109 may be implemented on an integrated circuit that also includes processor 102.

[0030] The I / O interface unit supports communication with various storage devices and I / O devices. For example, the terminal interface unit 112 supports the connection of one or more user I / O devices 116 that may include a user output device (such as a video display device, a speaker, and / or a television receiver) and a user input device (such as a keyboard, a mouse, a keypad, a touchpad, a trackball, a button, a light pen, or other pointing device). The user may operate the user input device using the user interface to provide input data and commands to the user I / O device 116 and the computer system 100, and may further receive output data via the user output device. For example, the user interface may be presented via the user I / O device 116, such as a display on a display device, playback by a speaker, or printing by a printer.

[0031] The memory interface 113 supports the connection of one or more disk drives or direct access storage devices 117 (usually rotating a magnetic disk drive storage device, but alternatively, other storage devices including a disk drive configured to appear as a single large-capacity storage device to the host computer, or an array of solid state drives such as flash memory may also be used). In some embodiments, the storage device 117 may be implemented by any type of secondary storage device. The contents of the memory 104, or any part thereof, may be stored in the storage device 117 and may be retrieved from the storage device 117 as needed. The I / O device interface 114 provides an interface to various other I / O devices or any of other types of devices such as a printer or a fax machine. The network interface 115 provides one or more communication paths from the computer system 100 to other digital devices and computer systems, and these communication paths may include, for example, one or more networks 130.

[0032] The computer system 100 shown in FIG. 1 describes a specific bus structure that provides a direct communication path among a processor 102, a memory 104, a bus interface 109, a display system 124, and an I / O bus interface unit 110. However, in alternative embodiments, the computer system 100 may be configured in various forms such as a hierarchical configuration, a star configuration, or a web configuration, multiple hierarchical buses, parallel and redundant paths, or different buses or communication paths including point-to-point links in any other suitable type of configuration. Further, although the I / O bus interface unit 110 and the I / O bus 108 are shown as separate respective members, the computer system 100 may actually include multiple I / O bus interface units 110 and / or multiple I / O buses 108. Multiple I / O interface units are shown that separate the I / O bus 108 from various communication paths leading to various I / O devices, but in other embodiments, some or all of those I / O devices are directly connected to one or more system I / O buses.

[0033] In various embodiments, the computer system 100 may be a multi-user mainframe computer system, a single-user system, or a server computer or a similar device with little or no direct user interface, but receives requests from other computer systems (clients). In other embodiments, the computer system 100 may be implemented as a desktop computer, a portable computer, a laptop or notebook computer, a tablet computer, a pocket computer, a telephone, a smartphone, or any other suitable type of electronic device.

[0034] Next, referring to FIG. 2, a state prediction system according to an embodiment of the present disclosure will be described.

[0035] FIG. 0 is a diagram showing an exemplary functional configuration of a state prediction system 200 according to an embodiment of the present disclosure. As shown in FIG. 2, the state prediction system 200 mainly includes a semiconductor manufacturing device 210, a state prediction device 220, and a user terminal 240. The semiconductor manufacturing device 210, the state prediction device 220, and the user terminal 240 can be communicably connected via a communication network such as a local area network (LAN), the Internet, a wide area network (WAN), or the like.

[0036] The semiconductor manufacturing device 210 is a device used for manufacturing semiconductor devices. As an example, the semiconductor manufacturing device 210 can be a plasma processing apparatus including a processing chamber in which plasma processing is performed on a semiconductor substrate. For example, the semiconductor manufacturing device 210 may be a plasma etching device. The plasma processing performed by the semiconductor manufacturing device 210 can be controlled by adjusting a set of operating parameters (e.g., coil current, microwave intensity, pressure, high-frequency bias power, plasma impedance). The processing chamber of the semiconductor manufacturing device 210 may be provided with a plurality of sensors configured to monitor and measure the conditions of the processing chamber during plasma processing and transmit a set of target chamber data 211 collected by the sensors to the state prediction device 220. Note that, for convenience of explanation, an exemplary configuration of the state prediction system 200 including a single semiconductor manufacturing device 210 is shown in FIG. 2, but it should be noted that the state prediction system 200 according to the embodiment of the present disclosure is not limited herein, and a configuration including a plurality of semiconductor manufacturing devices is also possible.

[0037] The state prediction device 220 is a device configured to generate a state prediction result characterizing the performance difference between a target processing chamber (e.g., the second processing chamber) and a reference processing chamber (e.g., the first processing chamber), and to adjust the operating parameters of the target processing chamber using this state prediction result to reduce the performance difference between the target processing chamber and the reference processing chamber. Here, the target processing chamber and the reference processing chamber may be the same processing chamber within the semiconductor manufacturing device 210 at different times (e.g., before and after cleaning or component replacement), or different processing chambers within different semiconductor manufacturing devices (e.g., the semiconductor manufacturing device 210 and another semiconductor manufacturing device not shown in FIG. 2).

[0038] As shown in FIG. 2, the state prediction device 220 includes a set of reference chamber data 221, a data acquisition unit 222, a feature management unit 224, a correlation calculation unit 226, a ranking unit 228, and a state prediction unit 230. In an embodiment, the state prediction device 220 may be implemented using the computer system 100 shown in FIG. 1 such that a set of reference chamber data 221 is stored in the storage device 117 of the computer system 100, and the functions of the data acquisition unit 222, the feature management unit 224, the correlation calculation unit 226, the ranking unit 228, and the state prediction unit 230 are implemented using software modules of a state prediction application 150 stored in the memory 104 of the computer system 100. Alternatively, the functions of the data acquisition unit 222, the feature management unit 224, the correlation calculation unit 226, the ranking unit 228, and the state prediction unit 230 may be implemented using dedicated hardware or integrated circuits.

[0039] The data acquisition unit 222 is a functional unit configured to acquire a set of chamber data characterizing the operating conditions within the processing chamber of a semiconductor manufacturing device. For example, the data acquisition unit 222 may acquire a set of reference chamber data 221 (e.g., a second set of operating data) along with a set of target chamber data 211 (e.g., a first set of operating data) collected and transmitted by the semiconductor manufacturing device 210. As described herein, the set of target chamber data 211 may be a set of data characterizing operating conditions that do not reach the performance threshold of the target processing chamber during plasma processing (e.g., an unacceptable chamber state), and may be acquired from the semiconductor manufacturing device 210. The set of reference chamber data 221 may be a set of data characterizing operating conditions that do not reach the performance threshold of the reference processing chamber during plasma processing (e.g., an unacceptable chamber state), and may be pre-collected and stored within the state prediction device 220. Since the function of the data acquisition unit 222 will be described later, its description is omitted here.

[0040] The feature management unit 224 is a functional unit configured to generate a feature map based on the chamber data acquired by the data acquisition unit 222. For example, the feature management unit 224 may generate a first feature map for a first target feature based on the set of reference chamber data 221, and generate a second feature map for the first target feature based on the target chamber data 211. Here, "target feature" refers to an individual measurable characteristic or property of chamber data corresponding to an adjustable operating parameter of the processing chamber of a semiconductor manufacturing device. Therefore, the "target feature" can be considered to be data corresponding to operating parameters such as coil current, microwave intensity, pressure, high-frequency bias power, etc. of the semiconductor manufacturing device 210, or any other operating parameter. Since the function of the feature management unit 224 will be described later, its description is omitted here.

[0041] The correlation calculation unit 226 is a functional unit configured to calculate the normalized cross-correlation result between feature maps generated by the feature management unit 224 for specific target features. For example, the correlation calculation unit 226 may calculate a normalized cross-correlation result indicating the level of uniformity of the first target feature between the first feature map and the second feature map. Since the function of the correlation calculation unit 226 will be described later, its description is omitted here.

[0042] The ranking unit 228 is a functional unit configured to rank the features of the feature maps generated by the feature management unit 224 based on the normalized cross-correlation results generated by the correlation calculation unit 226. For example, the ranking unit 228 may assign a ranking to the first target feature indicating the suitability of the first target feature for a set of target features with respect to the prediction of the process chamber state based on the normalized cross-correlation result, and select a subset of target features that reach the ranking threshold from among the set of target features. Since the function of the ranking unit 228 will be described later, its description is omitted here.

[0043] The state prediction unit 230 is a functional unit for generating a state prediction result that characterizes the performance difference of the target process chamber with respect to the reference process chamber. In an embodiment, the state prediction result may indicate a change to a set of operating parameters of a semiconductor manufacturing device (e.g., semiconductor manufacturing device 210) including the target chamber so as to reduce the performance difference of the target process chamber with respect to the reference process chamber. The state prediction result may be transmitted to the user terminal 240 for presentation to the user and further to the semiconductor manufacturing device 210 to facilitate adjustment of the operating parameters. Since the function of the state prediction unit 230 will be described later, its description is omitted here.

[0044] The user terminal 240 is a device used by a user of the state prediction system 200. The user terminal 240 can be used to determine the state prediction result generated by the state prediction unit 230 and input data to the state prediction device 220 via a graphic user interface. As an example, the user terminal 240 can include a laptop computer, a desktop computer, a smartphone, a tablet, or other suitable computing device.

[0045] According to the state prediction system 200 described with reference to FIG. 2, based on the uniformity of a given target feature between an acceptable chamber state and an unacceptable chamber state, rank a set of target features present in a feature map corresponding to chamber data of a semiconductor manufacturing device, and then use the high-rank target features to generate a state prediction result that characterizes the performance difference of a target processing chamber with respect to a reference processing chamber. Next, this state prediction result can be used to adjust the operating parameters of a semiconductor manufacturing device including the target processing chamber so as to reduce the performance difference between the target processing chamber and the reference processing chamber.

[0046] Next, with reference to FIG. 3, an overview of a state prediction result generation process according to an embodiment of the present disclosure will be described.

[0047] FIG. 3 is a flowchart showing the flow of a state prediction result generation process 300 according to an embodiment of the present disclosure. The state prediction result generation process 300 is a process for generating a highly accurate state prediction result for a semiconductor manufacturing device based on features determined to have high uniformity between an acceptable chamber state and an unacceptable chamber state. The state prediction result generation process 300 can be executed by the functional units of the state prediction device 220 shown in FIG. 2.

[0048] First, in step S310, data acquisition unit 222 acquires reference chamber data 221. In an embodiment, in a reference chamber of a semiconductor manufacturing device, plasma may be generated while a plurality of operation parameter setting values change at specific intervals within a specific value range, the plasma process may be performed on a semiconductor substrate, and data characterizing the conditions of the reference chamber during the plasma process may be collected using various sensors mounted on the semiconductor manufacturing device and acquired by data acquisition unit 222 as reference chamber data 221. The collected reference chamber data 221 may be a time-series signal characterizing the conditions of the reference chamber during the plasma process. Here, the reference chamber may be a plasma processing chamber that reaches a performance threshold (e.g., generates a plasma processing result that achieves a desired quality standard). In this specification, a processing chamber that reaches a performance threshold may also refer to a chamber having an "acceptable chamber state".

[0049] Next, in step S315, feature management unit 224 generates a feature map of the reference chamber data 221. More specifically, feature management unit 224 may calculate a representative value of the sensor values collected during the plasma process at the operation parameter setting values that changed in step S310 using the reference chamber data 221. That is, feature management unit 224 may perform one or more operations on the operation parameter values in the reference chamber data 221 to calculate a plurality of representative values of the plasma generation characteristics under various processing conditions. Thereafter, feature management unit 224 performs an operation on the calculated representative values to calculate a feature quantity for a set of target features (e.g., features corresponding to adjustable operation parameters of the plasma processing chamber such as coil voltage, magnetron current, valve opening degree, etc.), and maps these feature quantities to a two-dimensional or higher-dimensional graph having an axis corresponding to a specific target feature to generate a first feature map 321.

[0050] In this way, the feature management unit 224 can generate a first feature map 321 that includes feature amounts for target features representing operation parameters characterized by the reference chamber data 221. As will be described later, the first feature map 321 can be used as a reference feature map for grasping the performance difference between the target chamber and the reference chamber. Here, for the sake of convenience of explanation, reference is made to the first feature map corresponding to the first target feature of the reference chamber data 221. In reality, however, feature maps for a plurality of target features characterizing the reference chamber data can be generated and processed according to the following steps.

[0051] In step S320, the data acquisition unit 222 acquires the target chamber data 211. In an embodiment, in the target chamber of the semiconductor manufacturing device, plasma may be generated while a plurality of operation parameter setting values change at specific intervals within a specific value range. The plasma process may be executed with respect to the semiconductor substrate, and data characterizing the conditions of the target chamber during the plasma process may be collected using various sensors mounted on the semiconductor manufacturing device and acquired by the data acquisition unit 222 as the target chamber data 211. The collected Target chamber data 211 may be a time series signal characterizing the conditions of the target chamber during the plasma process. Here, the target chamber may be a plasma processing chamber that does not reach the performance threshold (for example, generates a plasma processing result that does not reach the desired quality standard). In this specification, a processing chamber that does not reach the performance threshold may also refer to a chamber having an "unacceptable chamber state".

[0052] Next, in step S325, the feature management unit 224 generates a feature map of the target chamber data 211. More specifically, the feature management unit 224 may calculate a representative value of the sensor values collected during plasma processing at the operation parameter setting values changed in step S320 using the target chamber data 211. That is, the feature management unit 224 may execute one or more operations on the operation parameter values in the target chamber data 211 to calculate a plurality of representative values of the plasma generation characteristics under various processing conditions. Thereafter, the feature management unit 224 may perform operations on the representative values calculated to calculate feature amounts, and map these feature amounts onto a two-dimensional or higher-dimensional graph having axes corresponding to specific target features (for example, features corresponding to adjustable operation parameters of the plasma processing chamber such as coil voltage, magnetron current, valve opening degree, etc.), to generate a second feature map 311.

[0053] In this way, the feature management unit 224 can generate a second feature map 311 including feature amounts for target features representing operation parameters characterized by the target chamber data 211. As will be described later, this second feature map 311 can be compared with a first feature map 321 for grasping the performance difference of the target chamber with respect to the reference chamber. Here, for the sake of convenience of explanation, reference is made to the second feature map corresponding to the first target feature of the target chamber data 211 (for example, the same target feature as the first feature map 321), but in actuality, according to the following steps, feature maps for a plurality of target features characterizing the target chamber data can be generated and processed.

[0054] Next, in step S330, the correlation calculation unit 226 calculates a normalized cross-correlation result indicating the level of uniformity of the first target feature between the first feature map 321 and the second feature map 311. Here, the level of uniformity refers to the similarity of the first target feature between the first feature map and the second feature map. Using this normalized cross-correlation result, it is possible to identify those target features that best represent the conditions of the processing chamber. Details of the normalized cross-correlation result generation process 400 are described with reference to FIG. 4, and thus the description thereof is omitted here.

[0055] Next, in step S335, the ranking unit 228 filters the set of target features based on the normalized cross-correlation result in order to determine a subset of target features that reach the ranking threshold. In an embodiment, the ranking unit may assign higher ranks to those target features having a higher level of uniformity. Therefore, target features having higher uniformity between the first feature map 321 and the second feature map 311, and thus more accurately representing the conditions of the processing chamber, may be selected for use in predicting the state of the semiconductor manufacturing device. In this way, the rank assigned by the ranking unit indicates the suitability of each target feature relative to other target features for predicting the processing chamber state (for example, target features having higher uniformity are considered to be more suitable for evaluating the state of the processing chamber). Details of the target feature ranking process 800 are described with reference to FIG. 8, and thus the description thereof is omitted here.

[0056] Next, in step S340, the state prediction unit 230 generates a state prediction result for the target chamber based on the subset of target features determined in step S335. In an embodiment, this state prediction result may include an operation parameter revision recommendation indicating a change to a set of operation parameters of the semiconductor manufacturing device including the target chamber so as to reduce the performance difference of the target chamber relative to the reference chamber. The details of the state prediction result generation process 900 will be described with reference to FIG. 9, and thus the description thereof is omitted here.

[0057] Next, in step S345, the semiconductor manufacturing device including the target chamber may adjust its operation parameters according to the operation parameter revision recommendation generated in step S340 so as to reduce the performance difference between the target chamber and the reference chamber.

[0058] According to the state prediction result generation process 300 described with reference to FIG. 3, it is possible to reduce the performance difference between the target chamber and the reference chamber (caused by passage of time, component replacement, component cleaning, etc.). In this way, the variation in the performance of the semiconductor device manufactured by the semiconductor manufacturing device including the target chamber can be suppressed, and uniform performance of the semiconductor device can be achieved. Furthermore, note that since the state prediction results are generated based on those target features determined to have a high uniformity (e.g., reaching a ranking threshold) between acceptable chamber states and unacceptable chamber states, noise and inconsistent features can be excluded, enabling more reliable and accurate generation of state prediction results.

[0059] Next, with reference to FIGS. 4 to 7, the normalized cross-correlation result generation process 400 according to an embodiment of the present disclosure will be described. The normalized cross-correlation result generation process 400 is a process for generating a normalized cross-correlation result indicating the level of uniformity of specific target features between different feature maps (e.g., the first feature map 321 and the second feature map 311). The normalized cross-correlation result generation process 400 can be executed by the correlation calculation unit 226 and can substantially correspond to step S330 of the state prediction result generation process 300 shown in FIG. 3.

[0060] First, in step S405, the correlation calculation unit 226 normalizes the first feature map 321. Here, the correlation calculation unit 226 can calculate the average value of the features included in the first feature map 321 and normalize the first feature map 321 by subtracting this average value from each feature in the first feature map 321. In this way, the first feature map 321 can be normalized to correspond to different magnification factors between different sensors and the like.

[0061] Next, in step S410, the correlation calculation unit 226 normalizes the second feature map 311. Here, the correlation calculation unit 226 can calculate the average value of the features included in the second feature map 311 and normalize the second feature map 311 by subtracting this average value from each feature in the second feature map 311. In this way, the second feature map 311 can be normalized to correspond to different magnification factors between different sensors and the like.

[0062] Next, in step S415, the correlation calculation unit 226 calculates the cross - correlation between the normalized first feature map 321 and the normalized second feature map 311. Here, the cross - correlation refers to the degree of similarity between two data sets according to the displacement amount of one data set with respect to the other data set. For example, when two data sets x and y are given, the correlation between x and y can be calculated using the following formula 1.

Equation

[0063] Next, in step 420, the correlation calculation unit 226 calculates the autocorrelation of the normalized first feature map 321 (for example, the feature map corresponding to the reference chamber having an acceptable chamber state). Here, the autocorrelation refers to the degree of similarity between the data set and its delayed copy according to the delay. The autocorrelation of the normalized first feature map 321 can be calculated using conventional techniques and is not particularly limited herein.

[0064] Next, in step 425, the correlation calculation unit 226 calculates the autocorrelation of the normalized second feature map 311 (for example, the feature map corresponding to the target chamber having an unacceptable chamber state). The autocorrelation of the normalized second feature map 311 can be calculated using conventional techniques and is not particularly limited herein.

[0065] Next, in step S430, the correlation calculation unit 226 normalizes the cross-correlation result generated in step S415 based on the maximum value of the autocorrelation of the normalized first feature map 321 calculated in step S420 and the maximum value of the autocorrelation of the normalized second feature map 311 calculated in step S425. Here, the correlation calculation unit 226 may generate a normalized cross-correlation result using the following Equation 2.

Equation

[0066] FIG. 5 is a diagram showing an example of the normalized cross-correlation 510. In an embodiment, as shown in FIG. 5, the normalized cross-correlation result 510 can be represented as a bright image, and the brighter (e.g., white) the pixel, the greater the degree of correlation between the normalized first feature map 321 and the normalized second feature map 311.

[0067] Next, in step S435, the correlation calculation unit 226 can generate a graph based on the normalized cross-correlation result 510 calculated in step S430. As an example, the correlation calculation unit 226 can extract data points from the central region 515 of the normalized cross-correlation result 510 shown in FIG. 5, plot them on the graph, and generate a cross-correlation graph.

[0068] FIG. 5 is a diagram showing an example of the cross-correlation graph 520 generated based on the normalized cross-correlation result 510 calculated in step S430. As shown in FIG. 5, in the cross-correlation graph 520, the vertical axis can represent the degree of correlation between the normalized first feature map 321 and the normalized second feature map 311, and the horizontal axis can represent the displacement between the normalized first feature map 321 and the normalized second feature map 311.

[0069] In a cross-correlation graph 520 generated based on the normalized cross-correlation result between a normalized first feature map 321 and a normalized second feature map 311, the width of the correlation range (e.g., full width at half maximum) represents the uniformity level of a specific target feature between the normalized first feature map 321 and the normalized second feature map 311. Since the normalized first feature map 321 is generated based on chamber data for a reference chamber having an acceptable chamber state, and the normalized second feature map 311 is generated based on chamber data for a target chamber having an unacceptable chamber state, the width of the correlation range can be considered to represent the uniformity level of a specific target feature between an acceptable chamber state and an unacceptable chamber state. Aspects of the present disclosure rank target features based on their uniformity because target features having a high uniformity level between an acceptable chamber state and an unacceptable chamber state more accurately characterize the conditions of a processing chamber of a semiconductor manufacturing device, and by selecting target features that reach a ranking threshold (e.g., high uniformity), it is possible to eliminate noise and non-conforming features and facilitate the generation of more accurate and reliable state prediction results.

[0070] FIG. 6 is a diagram showing an example of a cross-correlation graph 620 generated for a set of feature maps 610 corresponding to target features of plasma impedance. FIG. 7 is a diagram showing an example of a cross-correlation graph 720 generated for a set of feature maps 710 corresponding to target features of magnetron current. As described herein, the full width at half maximum characteristics of the cross-correlation graphs 620 and 720 indicate the uniformity of specific target features between the respective sets of feature maps 610 and 710. As can be seen by comparing the cross-correlation graph 620 and 720, the full width at half maximum characteristic 725 of the cross-correlation graph 720 is larger than the full width at half maximum characteristic 625 of the cross-correlation graph 620. Therefore, it can be determined that the target features of the magnetron current have greater uniformity between acceptable chamber conditions and unacceptable chamber conditions than the target features of the plasma impedance.

[0071] Next, with reference to FIG. 8, a target feature ranking process according to an embodiment of the present disclosure will be described.

[0072] FIG. 8 is a flowchart showing an exemplary flow of a target feature ranking process 800 according to an embodiment of the present disclosure. The target feature ranking process 800 is a process for ranking target features based on their uniformity between feature maps corresponding to acceptable chamber states and unacceptable chamber states. The target feature ranking process 800 can be executed by a ranking unit 228 and can substantially correspond to step S335 of the state prediction result generation process 300 shown in FIG. 3.

[0073] First, at step S805, the ranking unit 228 calculates the width of the correlation range of the correlation graph generated at step S435 of the normalized cross-correlation result generation process 400 described with reference to FIG. 4. Here, the ranking unit 228 can identify the peak correlation value in the correlation graph, identify the points on the left and right of the peak where the correlation drops to half of the peak value, and measure the distance between those two points as the full width at half maximum characteristic of the correlation graph. By repeating step S805 and the steps of the cross-correlation result generation process 400 for other pairs of feature maps generated for other target features, the widths of the correlation ranges of a plurality of target features can be determined.

[0074] Next, in step S810, the ranking unit 228 assigns ranks to each target feature in a set of target features (e.g., a plurality of target features for which the full width at half maximum characteristic is calculated) based on its full width at half maximum as measured in step S805, such that a target feature having a larger width value is assigned a higher rank and a target feature having a lower width value is assigned a lower rank. In this way, a ranking list can be obtained in which the set of target features is ranked in order of the widths of their correlation values. As described herein, target features having larger width values have higher uniformity between an acceptable process chamber state and an unacceptable process chamber state.

[0075] Next, in step S815, the ranking unit 228 selects a subset of target features that reach a ranking threshold from the ranked set of target features obtained in S810. Here, the ranking threshold is a criterion, rule, or standard that defines a boundary between target features considered to have high uniformity and target features considered to have low uniformity. In an embodiment, the ranking unit 228 may utilize a ranking threshold that specifies a minimum width value and select as the subset of target features all those target features that meet the specified minimum width value. In an embodiment, the ranking unit 228 may utilize a ranking threshold that specifies a particular number (e.g., the top 20 features) or percentage (e.g., the top 10%) of the highest rank target features and select the highest rank target features equal to the specified number or percentage.

[0076] According to the target feature ranking process 800 described with reference to FIG. 8, it is possible to rank a set of target features based on their uniformity between an acceptable process chamber state and an unacceptable process chamber state. As described herein, target features having a high level of uniformity between an acceptable chamber state and an unacceptable chamber state more accurately characterize the conditions of a process chamber of a semiconductor manufacturing device. Therefore, by ranking target features based on their uniformity and selecting target features that reach a ranking threshold (e.g., a high uniformity level), it is possible to eliminate noise and incompatible features and facilitate the generation of more accurate and reliable state prediction results.

[0077] Next, with reference to FIG. 9, a state prediction result generation process according to an embodiment of the present disclosure will be described.

[0078] FIG. 9 is a flowchart showing an exemplary flow of a state prediction result generation process 900 according to an embodiment of the present disclosure. The state prediction result generation process 900 is a process for generating a state prediction result including an operation parameter revision recommendation indicating a change to an operation parameter of a semiconductor manufacturing device including a target chamber so as to reduce a performance difference of the target chamber with respect to a reference chamber. The state prediction result generation process 900 can be executed by the state prediction unit 230 shown in FIG. 2.

[0079] First, at step S905, the state prediction unit 230 extracts, as a feature sub-map, a part of the second feature map 311 (e.g., a feature map corresponding to a target chamber having an unacceptable chamber state) for the target features included in the subset of target features determined in step S815 of the target feature ranking process 800 described with reference to FIG. 8. For example, the state prediction unit 230 may extract a 3×3 region from the second feature map 311 as the feature sub-map. An example of feature sub-map extraction is described with respect to FIG. 13, and the description thereof is omitted here.

[0080] Next, in step S910, the state prediction unit 230 uses the feature sub-map extracted from the second feature map 311 and the first feature map 321 in step S905 to execute a template matching method, and determines the position on the first feature map 321 where the feature sub-map most closely matches the first feature map 321. More specifically, the state prediction unit 230 slides the feature sub-map on the first feature map 321, compares the features of the feature sub-map with the features of the corresponding regions of the first feature map 321, and calculates an offset value indicating the degree of difference between the feature sub-map and each corresponding region of the first feature map 321. This process can be repeated multiple times for different feature maps corresponding to different target features. Since an example of the template matching result is described with respect to FIG. 14, the description thereof is omitted here.

[0081] Next, in step S915, the state prediction unit 230 generates a state prediction result characterizing the performance difference of the target chamber with respect to the reference chamber based on the offset value calculated in step S910. Here, in the embodiment, the state prediction unit 230 may generate an operation parameter revision recommendation as the state prediction result. The operation parameter revision recommendation is information indicating a recommended change to a set of operation parameters (for example, operation parameters corresponding to the target features of the processed feature map) of the semiconductor manufacturing device including the target chamber so as to reduce the performance difference between the second processing chamber and the reference chamber. As an example, when the offset value is calculated between the feature sub-map and the first feature map corresponding to the target feature of "magnetron current", the state prediction unit 230 may generate a state prediction result indicating the amount by which the magnetron current of the target chamber should be adjusted to reduce the performance difference between the target chamber and the reference chamber.

[0082] Next, in step S920, the state prediction unit 230 may send a request including the state prediction result to the semiconductor manufacturing device including the target chamber, and instruct the semiconductor manufacturing device to adjust its operation parameters according to the operation parameter revision recommendation. Alternatively, in a specific embodiment, the state prediction unit 230 may send a request including the state prediction result to the user terminal 240, and instruct the user to adjust the operation parameters of the semiconductor manufacturing device including the target chamber according to the operation parameter revision recommendation. In this way, the operation parameters of the semiconductor manufacturing device including the target chamber can be adjusted so as to reduce the performance difference between the target chamber and the reference chamber.

[0083] According to the state prediction result generation process 900 described with reference to FIG. 9, based on the offset value calculated by using the template matching method for the feature map including the target feature having a high uniformity level between the acceptable chamber state and the unacceptable chamber state, it is possible to reduce the performance difference of the target chamber with respect to the reference chamber (generated by time elapse, component replacement, component cleaning, etc.). In this way, the variation in the performance of the semiconductor device manufactured by the semiconductor manufacturing device including the target chamber can be suppressed, and uniform performance of the semiconductor device can be achieved.

[0084] Next, with reference to FIG. 10, an example of the time-series data signal of each sensor value acquired from the semiconductor manufacturing device will be described.

[0085] FIG. 10 is a diagram showing an example of the time-series data signal 1000 of each sensor value acquired from the semiconductor manufacturing device according to an embodiment of the present disclosure. The time-series data signal 1000 may include the reference chamber data 221 or the target chamber data 211 collected in steps S310 and S320 of the state prediction result generation process 300 shown in FIG. 3.

[0086] In FIG. 10, the vertical axis represents the sensor value (VS), and the horizontal axis represents the time (t). Here, the time T401 indicates the transition section immediately after the operation parameter setting, the time T402 indicates the steady state section, and the times T403 and T404 indicate the section immediately before the end of the plasma process. As described in this specification, the representative value can be calculated from the time series data signal 1000 received from each sensor, and the feature can be calculated from the representative value. Here, the representative value may be statistical data such as an average value or a standard deviation in a specific time region of the time series data, or may be a sensor value at a specific time.

[0087] Next, with reference to FIGS. 11 and 12, an example of the feature map data format according to the embodiment of the present disclosure will be described.

[0088] FIG. 11 is a diagram showing an example of feature map data 1100 according to an embodiment of the present disclosure. The feature map data 1100 shows an example of a two-dimensional feature map that can be generated as the first feature map 321 or the second feature map 311 shown in FIG. 3. In the feature map data 1100, the height direction and the width direction are associated with value ranges for target features corresponding to different operation parameters of the semiconductor manufacturing unit.

[0089] For example, as shown in the feature map data 1100 of FIG. 11, the horizontal direction indicates the range between the minimum value and the maximum value of the target feature B, and the vertical direction indicates the range between the minimum value and the maximum value of the target feature A. As described in this specification, the target features A and B can correspond to adjustable operation parameters of the semiconductor manufacturing unit. For example, the target features A and B can correspond to operation parameters such as microwave intensity, coil current, pressure, or the like, but the target features and related operation parameters are not limited thereto, and other parameters can be used. In the feature map data 1100, each cell 1110 represents the plasma processing conditions by the operation parameters defined by the target features A and B having specific values.

[0090] FIG. 12 is a diagram showing an example of a set of feature amounts 1200 stored in the feature map data 1100 shown in FIG. 11. The set of feature amounts 1200 represents the values of specific target features (for example, target feature A) stored in the cells 1110 (for example, a0 to a10) of the feature map data 1100. For example, each feature amount in the set of feature amounts 1200 may represent a value corresponding to an operation parameter such as microwave intensity, coil voltage, magnetron current, valve opening degree, pressure, or the like.

[0091] As described in this specification, in steps S315 and S325 of the state prediction result generation process 300 described with reference to FIG. 3, representative values of sensor values collected during plasma processing can be calculated, target feature amounts can be calculated from those representative values, and can be stored in corresponding cells of the feature map data 1100. In a specific embodiment, various operations can be performed to calculate the target feature amount from the representative value. In a specific embodiment, the representative value can be normalized using other feature amounts stored in the feature map data 1100. In a specific embodiment, the representative value can be directly used as a feature amount stored in the feature map data 1100. Although an example of two-dimensional feature map data 1100 and a related set of feature amounts 1200 have been described with reference to FIG. 12, it should be noted that the present disclosure is not limited thereto, and three-dimensional feature map data can also be used.

[0092] Next, with reference to FIG. 13, an example of feature submap extraction according to an embodiment of the present disclosure will be described.

[0093] FIG. 13 is a diagram showing an example of feature submap extraction according to an embodiment of the present disclosure. As described in this specification, the feature submap is a part of a feature map (for example, a feature map corresponding to a target chamber) extracted from the second feature map 311. This feature submap can be compared with the first feature map 321 using a template matching method in order to grasp an offset value indicating a difference in a specific target feature, and thus the corresponding operation parameters between the target chamber and the reference chamber.

[0094] As shown in FIG. 13, within the two-dimensional second feature map 311, the region indicated by the thick line can be extracted as the feature sub-map 1305. In this way, the feature sub-map 1305 becomes a 3×3 two-dimensional feature map. In this feature map, the horizontal axis represents the range of the target feature B, and the vertical axis represents the range of the target feature A. Accordingly, each cell of the feature sub-map 1305 is associated with a two-dimensional target feature set value (for example, (A0, B0)). As described in this specification, by using the feature sub-map 1305 and the first feature map 321 to execute the template matching process, an offset value indicating the difference in specific target features between the target chamber and the reference chamber can be determined. Although an example of the two-dimensional feature map has been described with reference to FIG. 13, it should be noted that the present disclosure is not limited to this specification, and the size and dimension of the feature map and the related feature sub-map can be set based on the nature of the application and the target feature. For example, a 3×3×3 three-dimensional sub-map, a 1×1×3 one-dimensional sub-map, etc. can also be used.

[0095] Next, with reference to FIG. 14, an example of the template matching result will be described.

[0096] FIG. 14 is a diagram showing an example of a template matching result 1400 according to an embodiment of the present disclosure. The template matching result 1400 is an example of a result obtained by performing template matching using the first feature map 321 (for example, the feature map corresponding to a reference chamber having an acceptable chamber state) and the feature sub-map 1305 extracted from the second feature map 311 (for example, the feature map corresponding to a target chamber having an unacceptable chamber state).

[0097] In FIG. 14, the vertical axis represents the matching index (MI), and the horizontal axis represents the sub-map scan position (SSP) of the feature sub-map 1305 on the first feature map 321. As described in this specification, when the template matching in step S910 of the state prediction result generation process 900 is executed, the state prediction unit 230 scans the feature sub-map 1305 extracted in step S905 on the first feature map 321, and calculates the matching index (MI) at each scan position (SSP) to generate the template matching result 1400.

[0098] In the template matching result 1400 shown in FIG. 14, the scan position 1401 where the matching index MI is the minimum represents the point where the value of the target feature in the feature sub-map 1305 and the value of the target feature in the first feature map 321 are closest (e.g., most similar) to each other. The scan position 1402 represents the scan position on the first feature map 321 corresponding to the region on the second feature map 311 from which the feature sub-map 1305 shown in FIG. 13 was extracted. If there is no performance difference between the target chamber and the reference chamber, the matching index MI should achieve the minimum value at the scan position 1402. Accordingly, the difference between the scan position 1401 and the scan position 1402 represents the offset value (e.g., degree of difference) of the target chamber with respect to the reference chamber.

[0099] In an embodiment, this offset value can be calculated based on the difference in the feature amounts of the feature sub-map 1305 at the scan position 1401 and the scan position 1402. For example, if the value of the target feature of the central cell of the feature sub-map 1305 at the scan position 1401 is represented as (A1), and the value of the target feature of the central cell of the feature sub-map 1305 at the scan position 1402 is represented as (A0), the offset value (VA) characterizing the performance difference of a specific target feature of the target chamber with respect to the reference chamber can be calculated according to the following Equation 3.

Equation

[0100] As described herein, since each target feature corresponds to an adjustable operating parameter of the processing chamber of the semiconductor management device, the offset value (VA) can be used to characterize the difference between the target chamber and the reference chamber for a specific operating parameter.

[0101] Here, the matching index can be calculated using, for example, the sum of absolute differences (SAD), the sum of squared differences (SSD), the normalized cross-correlation (NCC), the zero-mean normalized cross-correlation (ZNCC), etc. In addition, although an example has been described in which the scan position with the lowest matching index is identified as the point where the feature sub-map 1305 and the first feature map 321 are most similar, the present invention is not limited herein, and a method of determining the matching point between the feature sub-map 1305 and the first feature map 321 by performing fitting using a regression model in the vicinity of the minimum value of the matching index can also be used.

[0102] After the offset value of one feature sub-map (e.g., feature sub-map 1305) with respect to the first feature map is calculated, a new feature sub-map may be extracted from a different region of the first feature map, and template matching using this new feature sub-map may be repeated to identify its offset value. In this way, a plurality of offset values for different feature sub-maps can be calculated. Thereafter, the state prediction unit 230 may generate an aggregated offset value based on each individual offset value result. For example, the state prediction unit 230 may calculate an average value, a median value, a mode value, a weighted value, or other representative value as the aggregated offset value based on each offset value result. As described herein, this aggregated offset value can be used to generate an operating parameter revision recommendation indicating the amount by which a specific operating parameter (e.g., the operating parameter corresponding to the target feature for which the aggregated offset value is calculated) should be adjusted to reduce the performance difference between the target chamber and the reference chamber.

[0103] As described herein, the performance of semiconductor manufacturing devices (such as plasma processing devices) can change due to aging, component replacement, cleaning, or other factors. Such performance changes can manifest in the performance of semiconductor devices manufactured by those semiconductor manufacturing devices. To suppress variations in the performance of the manufactured semiconductor devices, it is desirable to reduce the performance differences between plasma processing chambers.

[0104] Aspects of the present disclosure relate to the recognition that features having a high level of uniformity between acceptable and unacceptable chamber states more accurately characterize the conditions of a semiconductor manufacturing device processing chamber. Accordingly, the present disclosure relates to a state prediction technique that ranks based on uniformity (as indicated by the width of those correlation ranges) and generates state prediction results using those target features that achieve a ranking threshold (e.g., a high degree of uniformity). This state prediction result can indicate a recommended change to a set of operating parameters (e.g., operating parameters corresponding to high-rank target features) of the target chamber so as to reduce the performance difference between the second processing chamber and the reference chamber.

[0105] In this way, variations in the performance of semiconductor devices manufactured by semiconductor manufacturing devices including the target chamber can be suppressed, and uniform performance of the semiconductor devices can be achieved. Further, note that since the state prediction results are generated based on those target features determined to have high uniformity between acceptable and unacceptable chamber states, noise and non-conforming features can be eliminated, enabling more reliable and accurate generation of state prediction results.

[0106] The present invention may be a system, method, and / or computer program product. This computer program product may include a computer-readable storage medium having computer-readable program instructions for causing a processor to implement aspects of the present invention.

[0107] A computer-readable storage medium can be a tangible device that holds and stores instructions for use by an instruction execution device. The computer-readable storage medium can be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of computer-readable storage media includes portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disk read-only memory (CD-ROM), digital versatile disk (DVD), memory stick, floppy disk, mechanically encoded devices such as punch cards or raised structures in grooves in which instructions are recorded, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium should not be construed to be a primary signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., an optical pulse passing through an optical fiber cable), or an electrical signal transmitted by an electrical wire.

[0108] Aspects of the present invention are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.

[0109] The computer-readable program instructions described above may be provided to a processor of a general purpose computer, a special purpose computer, or other programmable data processing apparatus to create means for causing the functions / acts specified in the blocks of the flowchart and / or block diagram to be performed via the processor of the computer or other programmable data processing apparatus. These computer-readable program instructions may further be stored in a computer-readable storage medium that stores instructions for a product comprising a computer, a programmable data processing apparatus, and / or other devices that can be caused to function in a particular manner so that the computer-readable storage medium stores instructions for implementing the functions / acts specified in the blocks of the flowchart and / or block diagram.

[0110] The computer-readable program instructions described above may further be loaded onto a computer, other programmable apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus, or other device so as to create a computer-implemented process for causing the functions / acts specified in the blocks of the flowchart and / or block diagram to be performed.

[0111] Flowcharts and block diagrams in the drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram can represent a module, segment, or portion of instructions that include one or more executable instructions for implementing a specialized logical function. In some alternative embodiments, the functions described in the blocks can occur in an order different from that shown in the drawings. For example, two blocks shown in succession can actually be executed substantially simultaneously, or the blocks can, in some cases, be executed in the reverse order depending on the relevant functions. Also, it will be recognized that each block of the block diagrams and / or flowchart diagrams and combinations of blocks of the block diagrams and / or flowchart diagrams can be implemented by a system based on specialized hardware that performs a specialized function or operation or a combination of dedicated hardware and computer instructions.

[0112] The foregoing relates to exemplary embodiments of the present invention, but other additional embodiments may be devised without departing from the basic scope of the present invention, which is determined by the claims set forth below. The descriptions of the various embodiments of the present disclosure are provided for illustrative purposes, but are not intended to be exhaustive or limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terms used herein are chosen to explain the principles of the embodiments, to facilitate the practical application or technical improvement of technologies existing in the market, or to enable others with ordinary skill in the art to understand the embodiments disclosed herein.

[0113] The terms used in this specification are for the purpose of describing particular embodiments only and are not intended to limit the various embodiments. As used in this specification, unless the context clearly dictates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms as well. Terms such as "a set of", "a group of", "a bundle of", etc. are intended to include one or more. Further, as used in this specification, the terms "comprising" and / or "comprises" specify the presence of the stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. In the above detailed description of exemplary embodiments of the various embodiments, reference has been made to the accompanying drawings (like numbers indicate like elements) forming a part thereof, which are shown by way of example and in which specific exemplary embodiments are shown, and various embodiments are practicable. The above embodiments have been described in sufficient detail for those skilled in the art to practice the embodiments, but other embodiments may be used and logical, mechanical, electrical, and other changes may be made without departing from the scope of the various embodiments. In order to achieve a sufficient understanding of the various embodiments, numerous specific details have been set forth in the above description. However, the various embodiments may be practiced without those specific details. In other instances, well-known circuits, structures, and techniques have not been shown in detail so as not to obscure the embodiments.

[0114] As described herein, aspects of the present disclosure relate to the following aspects.

[0115] (Aspect 1) A state prediction device for a semiconductor manufacturing device, wherein the state prediction device is a data acquisition unit, acquires a first set of operation data for a first processing chamber of a first semiconductor manufacturing device that reaches an operation threshold, and acquires a second set of operation data for a second processing chamber of a second semiconductor manufacturing device that does not reach the operation threshold A data acquisition unit configured as follows, A feature management unit, Generates a first feature map for a first target feature based on the first set of operation data, Generates a second feature map for the first target feature based on the second set of operation data A feature management unit configured as follows, A correlation calculation unit configured to calculate a normalized cross - correlation result indicating a level of uniformity of the first target feature between the first feature map and the second feature map, A ranking unit, Assigns a rank to the first target feature indicating the suitability of the set of target features for process chamber state prediction based on the normalized cross - correlation result, Selects a subset of target features that reach a ranking threshold from among the set of target features A ranking unit configured as follows, A state prediction unit configured to generate a state prediction result characterizing a performance difference of the second process chamber with respect to the first process chamber based on the subset of target features A state prediction device comprising.

[0116] (Aspect 2) The correlation calculation unit, Calculates a cross - correlation result between the first feature map and the second feature map, Calculates a first auto - correlation result for the first feature map, Calculates a second auto - correlation result for the second feature map, Generates the normalized cross - correlation result by normalizing the cross - correlation result based on a first maximum auto - correlation value from the first auto - correlation result and a second maximum auto - correlation value from the second auto - correlation result, Generates a cross - correlation graph of the normalized cross - correlation result The state prediction device according to aspect 1, configured as described above.

[0117] (Aspect 3) The ranking unit identifies the full width at half maximum characteristic from the cross-correlation graph, the Full width at half maximum characteristic assigns a rank to the first target feature based on the above. The state prediction device according to aspect 2, configured as described above.

[0118] (Aspect 4) The state prediction unit When the first target feature reaches the ranking threshold, extracts a feature sub-map from the second feature map, uses a template matching method to compare the feature sub-map extracted from the second feature map with the first feature map, and calculates an offset value indicating the difference in the first target feature between the first processing chamber and the second processing chamber, generates, as the state prediction result, an operation parameter revision recommendation indicating a change to the set of operation parameters of the second semiconductor manufacturing device based on the offset value so as to reduce the performance difference between the second processing chamber and the first processing chamber. The state prediction device according to any one of aspects 1 to 3, configured as described above.

[0119] (Aspect 5) The feature management unit calculates a first average value of the first target feature in the first feature map, calculates a second average value of the first target feature in the second feature map, normalizes the first feature map using the first average value, normalizes the second feature map using the second average value. The state prediction device according to any one of aspects 1 to 4, configured as described above.

[0120] (Aspect 6) The first semiconductor manufacturing device and the second semiconductor manufacturing device are the state prediction devices according to any one of Aspects 1 to 5, which are plasma etching devices.

Explanation of Signs

[0121] 200 State prediction system 211 Target chamber data 220 State prediction device 221 Reference chamber data 222 Data acquisition unit 224 Feature management unit 226 Correlation calculation unit 228 Ranking unit 230 State prediction unit 240 User terminal

Claims

1. A state prediction device for a semiconductor manufacturing device, wherein the state prediction device comprises: A data acquisition unit, acquiring a first set of operation data for a first processing chamber of a first semiconductor manufacturing device that reaches an operation threshold value, acquiring a second set of operation data for a second processing chamber of a second semiconductor manufacturing device that does not reach the operation threshold value configured as such, a data acquisition unit; A feature management unit, generating a first feature map for a first target feature based on the first set of operation data, generating a second feature map for the first target feature based on the second set of operation data configured as such, a feature management unit; A correlation calculation unit configured to calculate a normalized cross-correlation result indicating a level of uniformity of the first target feature between the first feature map and the second feature map; A ranking unit, assigning a rank to the first target feature indicating the suitability of the set of target features for the processing chamber state prediction based on the normalized cross-correlation result, selecting a subset of target features that reach a ranking threshold from among the set of target features configured as such, a ranking unit; A state prediction unit configured to generate a state prediction result characterizing a performance difference of the second processing chamber with respect to the first processing chamber based on the subset of target features A state prediction device comprising the above.

2. The correlation calculation unit, calculating a cross-correlation result between the first feature map and the second feature map, calculating a first auto-correlation result for the first feature map, calculating a second auto-correlation result for the second feature map, generating the normalized cross-correlation result by normalizing the cross-correlation result based on a first maximum auto-correlation value from the first auto-correlation result and a second maximum auto-correlation value from the second auto-correlation result, generating a cross-correlation graph of the normalized cross-correlation result configured as such, the state prediction device according to Claim 1.

3. The ranking unit, identifying a full width at half maximum characteristic from the cross-correlation graph, Assigning a rank to the first target feature based on the half-value full-width characteristic The state prediction device according to claim 2, configured as described above.

4. The state prediction unit When the first target feature reaches the ranking threshold, extracting a feature sub-map from the second feature map Using a template matching method to compare the feature sub-map extracted from the second feature map with the first feature map, calculating an offset value indicating the difference in the first target feature between the first processing chamber and the second processing chamber Generating, as the state prediction result, an operation parameter revision recommendation indicating a change to the set of operation parameters of the second semiconductor manufacturing device based on the offset value so as to reduce the performance difference between the second processing chamber and the first processing chamber The state prediction device according to claim 1, configured as described above.

5. The feature management unit Calculating a first average value of the first target feature in the first feature map Calculating a second average value of the first target feature in the second feature map Normalizing the first feature map using the first average value Normalizing the second feature map using the second average value The state prediction device according to claim 1, configured as described above.

6. The state prediction device according to claim 1, wherein the first semiconductor manufacturing device and the second semiconductor manufacturing device are plasma etching devices.

7. A state prediction method for a semiconductor manufacturing device, the state prediction method comprising Obtaining a first set of operation data for a first processing chamber of a first semiconductor manufacturing device that reaches an operation threshold Obtaining a second set of operation data for a second processing chamber of a second semiconductor manufacturing device that does not reach the operation threshold Generating a first feature map for a first target feature based on the first set of operation data Generating a second feature map for the first target feature based on the second set of operation data Calculating a first average value of the first target feature in the first feature map Calculating a second average value of the first target feature in the second feature map Normalizing the first feature map using the first average value; Normalizing the second feature map using the second average value; Calculating a cross-correlation result between the first feature map and the second feature map; Calculating a first autocorrelation result for the first feature map; Calculating a second autocorrelation result for the second feature map; Generating a normalized cross-correlation result indicating a level of uniformity of a first target feature between the first feature map and the second feature map, Normalizing the cross-correlation result based on a first maximum autocorrelation value from the first autocorrelation result and a second maximum autocorrelation value from the second autocorrelation result; Generating a correlation graph of the normalized cross-correlation result; Identifying a full width at half maximum parameter from the correlation graph; Assigning a rank to the first target feature indicating a suitability of the first target feature with respect to a set of target features for processing chamber state prediction based on the full width at half maximum parameter; Selecting a subset of target features that reach a ranking threshold from among the set of target features; When the first target feature reaches the ranking threshold, extracting a feature sub-map from the second feature map; Calculating an offset value indicating a difference in the first target feature between the first processing chamber and the second processing chamber using a template matching method to compare the feature sub-map extracted from the second feature map with the first feature map; Generating, as a state prediction result, an operation parameter revision recommendation indicating a change to a set of operation parameters of the second semiconductor manufacturing device based on the offset value so as to reduce a performance difference of the second processing chamber with respect to the first processing chamber; A method including the above.

8. A state prediction system for a semiconductor manufacturing device, the state prediction system comprising: A semiconductor manufacturing device for manufacturing a semiconductor device; A state prediction device for generating a state prediction result for the semiconductor manufacturing device; A user terminal for managing the semiconductor manufacturing device, and The state prediction device is A data acquisition unit, Obtain a first set of operation data for a first processing chamber that reaches an operation threshold, Obtain a second set of operation data for a second processing chamber that does not reach the operation threshold from the semiconductor manufacturing device A data acquisition unit configured as follows, A feature management unit, Generate a first feature map for a first target feature based on the first set of operation data, Generate a second feature map for the first target feature based on the second set of operation data A feature management unit configured as follows, A correlation calculation unit configured to calculate a normalized cross - correlation result indicating the level of uniformity of the first target feature between the first feature map and the second feature map, A ranking unit, Based on the normalized cross - correlation result, assign a rank to the first target feature indicating the suitability of the set of target features for predicting the processing chamber state, Select a subset of target features that reach a ranking threshold from among the set of target features A ranking unit configured as follows, A state prediction unit, Based on the subset of target features, generate a state prediction result characterizing the performance difference of the second processing chamber with respect to the first processing chamber, Output the state prediction result to the user terminal A state prediction unit configured as follows A state prediction system including

Images