In-chamber metrology of substrates for process characterization and improvement

Abstract

The method includes receiving, by a processing device, first data generated by a first sensor of a substrate processing system, the first data being generated in response to the first sensor receiving electromagnetic emissions from a substrate held by a robot arm of a transfer chamber in the substrate processing system. The method further includes processing the first data to obtain second data, the second data including a first indication of performance of the substrate processing system. The method further includes performing corrective action associated with the substrate processing system in consideration of the second data.

Description

[Technical Field]

[0001] Embodiments of the present disclosure relate to process control and process learning for substrate processing systems. Embodiments of the present disclosure also relate to metrology systems capable of generating metrology measurements of substrates as they are held by a robotic arm (e.g., in a transfer chamber). [Background technology]

[0002] Chambers are used in many types of processing systems. Examples of chambers include etch chambers, deposition chambers, anneal chambers, and other similar chambers. Typically, a substrate, such as a semiconductor wafer, is placed on a substrate support within the chamber, and conditions within the chamber are set and maintained to process the substrate. A detailed understanding of the processing conditions, their effects on the substrate, and the evolution of these parameters over time allows for tight control of product characteristics. Measuring one or more characteristics of the substrate (e.g., performing metrology operations) can inform decisions or actions related to updating or maintaining the processing conditions of the substrate. Summary of the Invention

[0003] The following is a simplified summary of the present disclosure to provide a basic understanding of some aspects of the disclosure. This summary is not an extensive overview of the disclosure. It is not intended to identify key or critical elements of the disclosure, nor is it intended to limit the scope of particular embodiments of the disclosure or the claims. Its sole purpose is to present some concepts of the disclosure in a simplified form as a prelude to the more detailed description that is presented later.

[0004] The method includes receiving, by a processing device, first data generated by a first sensor of a substrate processing system, the first data being generated in response to the first sensor receiving electromagnetic emissions from a substrate held by a robot arm of a transfer chamber in the substrate processing system. The method further includes processing the first data to obtain second data, the second data including a first indication of performance of the substrate processing system. The method further includes performing corrective action associated with the substrate processing system in consideration of the second data.

[0005] The substrate processing system includes a process chamber, a transfer chamber, a robot arm in the transfer chamber, a sensor, and a processing device. The transfer chamber is coupled to the process chamber. The robot arm is configured to transfer a substrate between the process chamber and the transfer chamber. The sensor has a field of view. At least a portion of the field of view includes at least a portion of the substrate when the substrate is transferred to or from the process chamber by the robot arm. The sensor is configured to receive electromagnetic radiation that has interacted with the substrate. The processing device is configured to process data. The data is generated in response to the sensor receiving the electromagnetic radiation. The processing device is configured to determine at least one of (i) one or more indications of performance of the substrate processing system or (ii) one or more characteristics of the substrate.

[0006] The non-transitory machine-readable storage medium stores instructions that, when executed, cause a processing device to perform an operation. The operation includes receiving first data generated by a first sensor of a substrate processing system. The first data is generated in response to the first sensor receiving electromagnetic radiation from a substrate held by a robot arm of a transfer chamber of the substrate processing system. The operation further includes processing the first data to obtain second data. The second data includes values ​​of a first one or more characteristics of the substrate. The operation further includes performing corrective action associated with the substrate processing system in consideration of the second data.

[0007] Numerous other features are provided in accordance with these and other aspects of the present disclosure. Other features and aspects of the present disclosure will become more fully apparent from the following detailed description, the claims and the accompanying drawings.

[0008] In the figures of the accompanying drawings, the present disclosure is illustrated by way of example, and not by way of limitation. In these figures, like reference numerals refer to like elements. It should be noted that different references to "one" or "an" embodiment in the present disclosure do not necessarily refer to the same embodiment, but rather that such references mean at least one. [Brief explanation of the drawings]

[0009] [Figure 1] 1 is a schematic top view of an exemplary processing system according to some embodiments. [Figure 2] 1 illustrates a simplified side view of a substrate processing system including an electromagnetic radiation probe, according to some embodiments. [Figure 3A] 1 is a schematic diagram of a measurement system including an array of probes, according to some embodiments. [Figure 3B] 2 is a schematic bottom view of a blade of a substrate processing system, according to some embodiments. [Figure 3C]FIG. 1 is a schematic diagram of a metrology system including an array of probes and a collection trajectory, according to some embodiments. [Figure 3D] FIG. 1 is a schematic diagram of a metrology system including an array of probes and a collection trajectory, according to some embodiments. [Figure 4] FIG. 4 is a block diagram illustrating an example system architecture 400 according to some embodiments. [Figure 5] FIG. 1 is a block diagram of an example dataset generator used to generate datasets for one or more models, according to some embodiments. [Figure 6] FIG. 1 is a block diagram of a system for generating output data for performing corrective actions, according to some embodiments. [Figure 7A] 1 is a flow diagram of a method for generating a dataset for a model for generating output data, according to some embodiments. [Figure 7B] 1 is a flow diagram of a method for generating and utilizing data based on radiation received by a probe while a substrate is in a processing system, according to some embodiments. [Figure 7C] 7 is a flow diagram of a method 700C for predicting properties of a substrate from probe data collected from within a processing system, according to some embodiments. [Figure 8] FIG. 1 is a block diagram of an exemplary computing device, according to some embodiments. DETAILED DESCRIPTION OF THE INVENTION

[0010] Described herein are methods, systems, apparatus, etc. related to performing metrology of substrates in controlled environments. For example, the disclosed methods, systems, and apparatus may relate to performing metrology measurements of substrates in a vacuum, throughput-transparent metrology measurements of substrates in processing, measuring substrates in environmentally controlled chambers that indicate processing conditions related to the substrates, etc.

[0011] In some systems, substrates are processed and / or fabricated in one or more processing chambers. The processing chamber may distinguish a processing environment (e.g., a region of space in which a substrate is processed) from ambient conditions. For example, substrate processing may be performed under controlled gas pressure, a controlled gas mixture, under vacuum, etc.

[0012] In some systems, multiple processing chambers may be coupled to a single processing system, for example, several processing chambers may each be coupled to a central transfer chamber. The environments within the processing chambers, the transfer chamber, and various other chambers (e.g., loading chamber) may be controlled environments.

[0013] In some systems, a substrate processing system may be configured to fabricate multi-layer substrates. One or more operations may be performed to produce one layer of the substrate, a second one or more operations may be performed to produce an additional layer of the substrate, etc. In some systems, a substrate may be transferred between processing chambers during processing, e.g., a first processing chamber coupled to a transfer chamber may be configured to perform a first operation for processing the substrate, a second processing chamber coupled to the transfer chamber may be configured to perform a second operation for processing the substrate, etc. A substrate may be processed in multiple stages, in multiple processing chambers, without interacting with ambient atmospheric conditions, without leaving the controlled environment of the processing system, etc.

[0014] In some systems, the substrate may be sensitive to atmospheric conditions. When introduced to atmospheric conditions, an uncontrolled atmosphere, or other similar environment, the substrate (e.g., the substrate material, the substrate film, the substrate shape or dimensions, etc.) may be damaged, changed, or altered. In some systems, the substrate may be protected (e.g., the substrate may have a final layer deposited that is resistant to atmospheric contaminants) before it is removed from the processing system.

[0015] In some systems, substrate measurements (e.g., metrology measurements) are performed on fabricated substrates. The substrate measurements may provide an indication of substrate performance (e.g., a prediction of satisfactory / unsatisfactory performance of the substrate), substrate processing performance (e.g., the quality of the substrate processing procedure), etc. The measurements may be performed in a stand-alone metrology system (e.g., a room, facility, area, or the like, separate from the substrate processing system), or in an in-line metrology system (e.g., a metrology system coupled to the substrate processing system), or other such metrology system.

[0016] In some systems, metrology operations are performed outside of the processing system, e.g., the substrate leaves a controlled environment before measuring the substrate, the substrate comes into contact with atmospheric / ambient conditions before measuring the substrate, the substrate has metrology operations performed under ambient conditions, etc. In some systems, metrology operations may be performed separate in time from processing operations, e.g., an in-line metrology system may receive a substrate and perform measurements on the substrate minutes after the last processing operation has been performed, etc. (e.g., because of this delay, information from the substrate (e.g., for use in controlling or adjusting future processes) may not be available for some time.) A stand-alone metrology system may receive a substrate several hours after the last substrate processing operation has been performed, etc. Many other substrates may be processed between the processing of the substrate and the generation of metrology data by the stand-alone metrology system.

[0017] In some systems, metrology operations may be performed on a subset of the substrates produced by the substrate processing system. For example, randomly selected substrates may be flagged for additional metrology, a substrate may be flagged for additional metrology periodically (e.g., one substrate for every n substrates produced), selected substrates may be flagged for additional metrology based on performance metrics (e.g., sensor data associated with the processing of those substrates), etc. In some embodiments, a subset of the substrates may be flagged for the metrology operation, the metrology operation may be performed on a subset of the substrates, etc. Including more substrates in the metrology operation may reduce the processing throughput of the substrate processing system.

[0018] The present disclosure may address one or more of the shortcomings of conventional substrate processing and / or metrology systems. In some embodiments, a metrology system that generates measurements based on receiving electromagnetic (e.g., optical) radiation from a substrate is included as part of a processing system. The metrology system may be configured to receive radiation emitted by the substrate, radiation transmitted through the substrate, radiation reflected by the substrate, or other similar types of radiation. The metrology system may include a probe, e.g., a device for collecting electromagnetic radiation from the substrate. The probe may be located within the processing system; for example, the probe may be in a processing chamber, a transfer chamber, a load lock, a mainframe or chamber body, a factory interface, a controlled environment, etc. The probe may receive electromagnetic radiation from the substrate while the substrate is in a controlled environment (e.g., a processing environment, a vacuum, an inert environment, etc.).

[0019] A probe that receives electromagnetic radiation from a substrate may be positioned along a path along which the substrate travels during or between processing operations. For example, the probe may be positioned above a region of a chamber or component configured to accommodate one or more substrates (e.g., the probe may be suspended from the lid or top wall of a transfer chamber, process chamber, load lock, slit valve assembly, or factory interface). The probe may have a field of view from which the probe may receive electromagnetic radiation. The probe's field of view may be positioned to include one or more regions of a chamber configured to accommodate a substrate. For example, a probe near the top of a chamber may have a field of view directed downward. The probe's field of view may intersect with a region traversed by a substrate as it is being transferred, for example, from one chamber to another. For example, the probe may be positioned near an interface between a processing chamber and a transfer chamber. A portion of a substrate may pass through the probe's field of view as a robot arm (e.g., a robot arm associated with a transfer chamber, a robot arm configured to transfer one or more substrates within a processing system, etc.) transfers the substrate into and / or out of a processing chamber. In some embodiments, the probe may be positioned differently. For example, the probe may be within the transfer chamber but not associated with a particular processing chamber coupled to the transfer chamber. In another example, the probe may be located within a slit valve assembly that separates the processing chamber from the transfer chamber. In some embodiments, a first process chamber may be utilized for a first set of operations related to processing a substrate, and the substrate may be moved to a second chamber for a second set of operations. The probe may be positioned such that the field of view of the probe intersects with the substrate being moved from the first processing chamber to the second processing chamber through the transfer chamber.

[0020] In some embodiments, one or more measurements are performed as the substrate passes through the field of view of the probe. In some embodiments, the probe may provide radiation received from the substrate (e.g., via one or more optics, fiber optic cables, etc.) to a spectrometer, one or more detectors, one or more cameras or imagers, etc. In some embodiments, one or more measurements may be performed while the substrate is being transferred between multiple positions, e.g., the substrate may be continuously moving while radiation is being collected by the probe. In some embodiments, the probe may be positioned so that the substrate naturally passes through the field of view of the probe, e.g., receiving electromagnetic radiation from the substrate may not increase the processing time, completion time, production time, or the like of the substrate. In some embodiments, several data points are collected while the substrate is being transferred through the field of view of the probe. For example, several data points along the diameter of the substrate may be collected. In some embodiments, an electromagnetic source may be provided, and radiation reflected by or transmitted through the substrate may be collected by the probe. In some embodiments, the probe may include a pyrometer and may collect emitted radiation from the substrate.

[0021] In some embodiments, multiple probes may be used (e.g., to map larger portions of the substrate). In some embodiments, additional detection systems may be used, such as object detection systems to synchronize data collection with the presence of a substrate in the field of view, synchronize the operation of the robot with the operation of a spectrometer, etc. In some embodiments, a robot for transferring a substrate may be instructed to adjust the path of the substrate through a processing system to facilitate measurement by the probe. In some embodiments, for example, the robot may cause the substrate to pass through the field of view of the probe multiple times, or the substrate may remain in the field of view of the probe for an extended period of time, or the robot may transfer the substrate along a path such that a larger portion of the substrate passes through the field of view of the probe than would pass if the substrate were passed directly from one chamber to another. In some embodiments, one or more portions of the substrate may pass through the field of view of one or more probes multiple times during processing. For example, a substrate may be transferred into a processing chamber for film deposition. As the substrate is being transferred to the processing chamber, the substrate may pass through the field of view of the probe, and radiation may be provided to the probe. After processing, the substrate may be transferred to another chamber and the substrate may again pass through the field of view of the probe.

[0022] Data from the spectrometer or other sensors may be provided to a processing device (e.g., a computer processor). The processing device may infer one or more metrics associated with the measured substrate. In some embodiments, the processing device may generate an indication of one or more properties of the substrate. The processing device may determine and / or predict one or more properties of the substrate using data based on electromagnetic radiation received from the substrate. The processing device may determine the thickness of a film (e.g., a film deposited on a substrate), the composition of the film, or the like. In some embodiments, data from the spectrometer may be provided to a model based on physical phenomena to determine one or more properties of the substrate. In some embodiments, data from the spectrometer and / or other sensors may be provided to a machine learning model trained to receive the spectrometer data and generate an indication of the substrate property as an output.

[0023] In some embodiments, the processing equipment may generate an indication of the performance (e.g., quality of performance) of one or more processing operations associated with the substrate. The processing equipment may receive spectral data from the spectrometer. The processing equipment may provide this spectral data to a trained machine learning model configured to generate as output a prediction of the processing conditions under which the substrate was previously processed. In some embodiments, the processing equipment may provide the spectral data to a machine learning model configured to perform corrective action (e.g., scheduling maintenance, updating a process recipe, alerting a user, etc.).

[0024] Aspects of the present disclosure provide one or more technical advantages over conventional systems. The present disclosure enables measurements to be performed in a controlled environment (e.g., in a processing system, a transfer chamber, a process chamber, etc.) to guide the processing of a substrate. Conventional in-line or stand-alone metrology systems may not measure substrates that are sensitive to atmosphere. For example, a protective layer may be formed on the substrate before removal from the controlled environment, which may interfere with measurements of other layers of the substrate.

[0025] Aspects of the present disclosure allow for measuring a substrate multiple times throughout processing. For example, as the substrate is moved into or out of a process chamber, measurements may be taken by a probe that may be indicative of one or more characteristics of the substrate, one or more processing conditions of the substrate, etc. In some embodiments, the substrate may be measured several times throughout processing, for example, between successive processing operations. In this manner, layers that may be obscured by performing metrology of the complete stack of substrate layers may be measured more accurately.

[0026] Aspects of the present disclosure enable measuring a substrate shortly after performing one or more processing operations on the substrate (e.g., seconds after processing, minutes after processing, etc.). In some systems, metrology of one or more produced substrates may be utilized when taking corrective action, such as updating a process recipe, scheduling maintenance, alerting a user, etc. In some conventional systems, there may be a significant delay (e.g., in terms of time, additional substrates processed before changes are made, or the like) between processing a substrate and performing a metrology operation associated with the substrate. According to the present disclosure, a substrate may be measured (e.g., a probe may receive electromagnetic radiation from the substrate) immediately after processing. Updates to processing conditions, process recipes, or the like may be initiated and / or performed immediately, which reduces wasted material, energy, time, etc., of operating a substrate processing system at less-than-optimal processing conditions.

[0027] Aspects of the present disclosure may enable metrology measurements to be performed on a substrate while the substrate is being transferred through a processing system. For example, a transfer robot may operate normally, an electromagnetic probe may receive radiation from the substrate, a processing device may determine one or more characteristics of the substrate, etc., while the transfer robot operates normally (e.g., when the robot is operating without a probe included in the processing system). Aspects of the present disclosure may enable measurements to be performed without an impact on throughput of the substrate processing system (e.g., the metrology system may be throughput-transparent), without a significant impact on throughput (e.g., a metrology operation according to the present disclosure may increase throughput for a short time below a threshold), etc. Aspects of the present disclosure may enable more substrates of a substrate processing system to be measured, e.g., each substrate processed by the processed system may be measured by the metrology system, or a portion of the substrates processed by the processing system (e.g., each substrate processed in a first process chamber of the substrate processing system) may be measured by the metrology system. In some embodiments, feedback control, recipe updates, periodic maintenance, etc. may be based on substrate metrology. Enabling measurements of a larger portion of a substrate may increase the certainty of corrective action predictions (e.g., additional data may increase the statistical likelihood of accurate corrective action). Measuring a larger portion of substrates processed by a substrate processing system may facilitate the processing system avoiding erroneous changes to one or more process recipes, (e.g., by increasing the certainty of detected deviations from ideal conditions in the process chamber) enabling the processing system to implement appropriate corrective action (e.g., increasing the likelihood that produced substrates will meet target performance metrics), etc. This may provide a technical advantage to the substrate processing system by reducing the energy, time, and materials devoted to performing less-than-optimal processing operations, and may reduce energy, time, materials, and defective product disposal costs associated with the increased likelihood of producing defective products.

[0028] In one aspect of the present disclosure, a method includes receiving, by a processing device, first data generated by a first sensor of a substrate processing system. The first data is generated in response to the first sensor receiving electromagnetic emissions from a substrate held by a robot arm of a transfer chamber in the substrate processing system. The method further includes processing the first data to obtain second data. The second data includes a first indication of performance of the substrate processing system. The method further includes performing corrective action associated with the substrate processing system in consideration of the second data.

[0029] In another aspect of the present disclosure, a substrate processing system includes a process chamber, a transfer chamber, a robot arm for the transfer chamber, a sensor, and a processing device. The transfer chamber is coupled to the process chamber. The robot arm is configured to transfer a substrate between the process chamber and the transfer chamber. The sensor has a field of view. At least a portion of the field of view includes at least a portion of the substrate when the substrate is transferred to or from the process chamber by the robot arm. The sensor is configured to receive electromagnetic radiation that has interacted with the substrate. The processing device is configured to process data. The data is generated in response to the sensor receiving the electromagnetic radiation. The processing device is configured to determine at least one of (i) one or more indications of performance of the substrate processing system or (ii) one or more characteristics of the substrate.

[0030] In another aspect of the present disclosure, a non-transitory machine-readable storage medium stores instructions that, when executed, cause a processing device to perform an operation. The operation includes receiving first data generated by a first sensor of a substrate processing system. The first data is generated in response to the first sensor receiving electromagnetic radiation from a substrate held by a robot arm of a transfer chamber of the substrate processing system. The operation further includes processing the first data to obtain second data. The second data includes values ​​of a first one or more characteristics of the substrate. The operation further includes performing corrective action associated with the substrate processing system in consideration of the second data.

[0031] FIG. 1 is a schematic top view of an exemplary processing system 100 in accordance with some embodiments. The processing system 100 may be a substrate processing system. The processing system 100 includes a transfer chamber robot 101 and a factory interface robot 121, each adapted to pick and place a substrate 110 (sometimes also referred to as a “wafer” or “semiconductor wafer”) from or to a destination within an electronic device processing system, such as the processing system 100 shown in FIG. 1. However, any type of electronic device substrate, mask, or other silica-containing substrate (generally referred to herein as a “substrate”) may be transported and transferred by the disclosed robot. For example, the destination of the substrate 110 may be one or more of one or more processing chambers 103 and / or load lock apparatuses 107A, 107B, which may be distributed around and coupled to the transfer chamber 114. As shown, substrate transfer may be performed, for example, through a slit valve 111.

[0032] The processing system 100 may further include a main frame 102 that includes a transfer chamber 114 and several processing chambers 103. The housing of the main frame 102 contains the transfer chamber 114 therein. The transfer chamber 114 may include a top wall (not shown), a bottom wall (floor) 139, and side walls, and may include a controlled environment. The controlled environment may include a vacuum condition, a controlled pressure (e.g., different from ambient atmospheric pressure), a controlled gas environment (e.g., an inert gas such as argon or nitrogen gas or a gas mixture), or the like. In the illustrated embodiment, a transfer chamber robot 101 is mounted to the bottom wall (floor) 139. However, the transfer chamber robot 101 may also be mounted elsewhere, such as to the top wall.

[0033] In various embodiments, the processing chamber 103 may be adapted to perform any number of processes on the substrate 110. The processes may include deposition, oxidation, nitridation, etching, polishing, cleaning, lithography, metrology (e.g., integrated metrology), or the like. Other processes may also be performed. The load lock apparatus 107A, 107B may be adapted to interface with a factory interface 117 or other system components, which may receive the substrate 110 from, for example, a substrate carrier 119 (e.g., a front-opening unified pod (FOUP)) that may be docked to a load port of the factory interface 117. A factory interface robot 121 (shown in dotted lines) may be used to transfer the substrate 110 between the substrate carrier 119 and the respective load lock apparatus 107A, 107B. The transfer of the substrate 110 may be performed in any order or in any direction. In some embodiments, factory interface robot 121 may be the same as (or similar to) transfer chamber robot 101, but may further include a mechanism that allows the factory interface robot to move laterally as indicated by arrow 123. Other suitable robots may be used as factory interface robot 121. In some embodiments, system 100 may be coupled to (e.g., interfaced with) a metrology system, such as an integrated metrology system, an in-line metrology system, or the like.

[0034] In an embodiment, as an illustrated description of a robot, a transfer chamber robot 101 includes at least one arm 113 (e.g., a robot arm) and at least one end effector 115 coupled to the arm 113. The end effector 115 is controllable by the transfer chamber robot 101 to pick up a substrate 110 from a load lock apparatus 107A or 107B, guide the substrate 110 through one of the slit valves 111 of the processing chambers 103, and precisely place the substrate 110 on a substrate support of the processing chamber 103. In some embodiments, the end effector 115 may include a blade for supporting the substrate 110. In some embodiments, the end effector 115 may support a first portion of the substrate 110 and may be, for example, ring-shaped, allowing a portion of the substrate 110 to be visible from underneath while the substrate 110 is supported by the end effector 115.

[0035] In various embodiments, one or more of the processing chambers 103 may include a probe 120 (e.g., a device for collecting electromagnetic radiation), at least a portion of which is within the chamber of the processing system 100. In some embodiments, the probe 120 may be within the processing chamber 103 (as shown). In some embodiments, the probe 120 may be within the transfer chamber 114. In some embodiments, the probe 120 may be within a slit valve assembly including the slit valve 111. In some embodiments, the probe 120 may be within a load lock 107A-B. In some embodiments, the probe 120 may be within the factory interface 117. In some embodiments, the probe 120 may include optics designed to couple collected radiation from within the chamber of the processing system 100 and couple the radiation to one or more fiber optic cables / cores. In some embodiments, the probe 120 may be positioned above the path of the substrate (e.g., suspended from or embedded in the top wall of the processing system 100). In some embodiments, a probe 120 may be positioned below the path of the substrate. In some embodiments, multiple probes may be included within the processing system 100, processing chamber 103, transfer chamber 114, etc.

[0036] In some embodiments, for example, the field of view of the probe (e.g., the region of space from which the probe is configured to receive electromagnetic radiation) may intersect with the path of the substrate 110, the path of the end effector 115, or may be near a connector between chambers (e.g., near the slit valve 111). In some embodiments, the probe 120 may be configured to receive radiation reflected from the surface of the substrate. In some embodiments, the probe 120 may be configured to receive radiation transmitted through the substrate. In some embodiments, the probe 120 may be configured to receive radiation emitted by the substrate. In some embodiments, the probe 120 may include an array of devices, such as devices with different fields of view (overlapping, non-overlapping, or other similar types), devices with different functions (e.g., devices that receive radiation in different portions of the electromagnetic spectrum), devices configured to receive reflected, emitted, or transmitted light, etc.

[0037] In some embodiments, the probe 120 may be configured to receive radiation from the substrate 110 while the substrate 110 is being transferred from a first portion of the processing system 100 to a second portion of the processing system 100. For example, the probe 120 may be positioned such that the field of view of the probe 120 may intersect a path through which the transfer arm 113 may send the substrate 110. The probe 120 may receive radiation from the substrate 110 while the substrate 110 is being transferred from the transfer chamber 114 to the processing chamber 103. The probe 120 may receive radiation from the substrate 110 while the substrate 110 is being transferred from the processing chamber 103 to the transfer chamber 114.

[0038] In some embodiments, light received by the probe 120 may be directed to a spectrometer 125 for analysis. The light received by the probe 120 may be focused into a fiber optic cable coupled to the spectrometer 125 for analysis, e.g., spectroscopic analysis. The spectrometer 125 may perform an operation to determine one or more spectra of the light, which may be used to determine at least one optical property of the substrate (e.g., the spectrometer 125 may include or be coupled to a processing device). In some embodiments, the light is received from the substrate before substrate processing. In some embodiments, the light is received from the substrate after substrate processing. In some embodiments, the light is received from the substrate between processing operations. The probe 120 and the spectrometer 125 may comprise optical sensors associated with the processing chamber 103, the transfer chamber 114, the processing system 100, etc. The probe 120 and the spectrometer 125 may comprise metrology systems operating in a controlled environment. The probe 120 and the spectrometer 125 may comprise in-vacuum metrology systems. The optical sensor may detect at least one property of the material of the substrate 110. In some embodiments, via thin film optical calculations by a processing device (e.g., processing device 130), the optical sensor may provide sensor data indicative of the thickness of the material of the substrate 110. For example, the optical sensor may be used to determine the thickness of the latest film formed on the substrate 110, the latest exposed film of the substrate 110 (e.g., after an etching operation), etc. The optical sensor may be used to determine the total thickness of the substrate. The optical sensor may be used to determine additional properties of the substrate. The optical sensor may be used to determine the geometry of the substrate (e.g., a measurement of one or more dimensions of the substrate). The optical sensor may be used to determine the chemical or physical composition of the substrate or a portion of the substrate. The optical sensor may be used to classify, analyze, and / or characterize the patterns of a patterned substrate.

[0039] In some embodiments, the processing system 100 may further include an electromagnetic radiation source, e.g., an electromagnetic radiation source optically coupled to the lobe 120 (e.g., the probe 120 is configured to receive radiation generated by this source). In some embodiments, the light received by the probe 120 may be reflected from the substrate 110. In some embodiments, radiation may be provided to the substrate 110 from a location near the probe 120, a fiber core bundled with the fiber cord of the probe 120, the same side of the substrate 110 as the probe 120 (e.g., the top side, the bottom side, etc.), etc. Radiation may be provided to the substrate 110 by a radiation coupling device embedded in a wall of the chamber, a radiation coupling device supported by a bottom wall of the chamber (e.g., bottom wall 139), a radiation coupling device suspended from the top wall of the chamber, etc. Light received by the probe 120 may be transmitted through the substrate 110. Radiation may be provided to the substrate 110 from a location disposed on the opposite side of the probe 120 (e.g., with respect to the substrate 110). For example, the probe 120 may be suspended from and / or near the top wall of the processing system 100, and radiation may be provided from near the bottom wall 139. In some embodiments, for example, the system may include multiple probes, may include one or more arrays of probes, or may include probes positioned below and / or above the substrate.

[0040] A controller 109 (e.g., a tool and equipment controller) may control various aspects of the processing system 100, such as the gas pressures in the processing chamber 103, the individual gas flow rates, spatial flow ratios, the temperatures of various chamber components, and the radio frequency (RF) or electrical conditions of the processing chamber 103. The controller 109 may receive signals from, and send commands to, the factory interface robot 121, the transfer chamber robot 101, one or more sensors, and / or other processing components of the processing system 100. Thus, the controller 109 may control the start and stop of processing, adjust the deposition rate, type, or mix of deposition compositions, and the like. The controller 109 may also receive and process sensing data from various sensors, such as sensors associated with the processing system 100.

[0041] In various embodiments, the controller 109 includes (or is coupled to) a processing unit 130 and is coupled to the spectrometer 125. The processing unit 130 may be configured to receive and process sensor data, including the results of spectroscopic analysis performed by the spectrometer 125. The processing unit 130 may be configured to accept multiple measurements and perform signal processing, such as signal processing configured to separate thin film effects (e.g., from optical effects of plasma generation) by receiving data collected with and without an external light source. The processing unit 130 may calculate a spectral signal (e.g., indicative of one or more properties of the substrate 110) by subtracting a second measurement from a first measurement, e.g., by subtracting the second spectrum from the first spectrum. The processing unit 130 may then fit the reflectometry signal to a thin film optical model to determine information including one or more optical thin film properties of the process film layer. From the thin film optical model, the processing unit 130 may determine one or more properties of the process product included in the substrate 110. The processing device 130 may calculate properties including thickness, reflectivity, chemistry, etc. Depending on the results of analyzing one or more optical film properties, the processing device 130 (e.g., controller 109) may perform one or more actions. For example, the processing device 130 may guide the processing chamber 103 in a process modification or adjustment. For example, the controller 109 may adjust process parameters or settings, such as the deposition rate, the type or mix of deposition compositions, the timing of performing a cleaning process in the processing chamber, and other actions discussed in more detail with reference to FIGS. 7A-C. Additionally, in some embodiments, the controller 109 may cause the following to occur: calibrate one or more sensors in the processing chamber 103 based on the results of analyzing one or more optical film properties; perform chamber matching based on the results of analyzing one or more optical film properties; identify chamber drift based on the results of analyzing one or more optical film properties; etc.

[0042] In some embodiments, the processing device 130 may perform analyses that are not intended to determine one or more characteristics of the substrate 110. The processing device 130 may receive spectral data from the spectrometer 125 and utilize the spectral data to determine one or more indications of the performance of the processing system 100 (e.g., performance quality, processing quality, etc.). In some embodiments, the processing device 130 may provide data from the spectrometer 125 (e.g., spectral data, processed spectral data, input vectors based on the spectral data, etc.) to a trained machine learning model. The trained machine learning model may be configured to generate as output one or more indications of processing conditions associated with the substrate (e.g., processing conditions under which the substrate was processed). The trained machine learning model may be configured to generate as output one or more corrective actions associated with the processing system 100 (e.g., prescribing a corrective action, causing a corrective action to be implemented, etc.). Corrective action may include updating a process recipe, updating a non-process recipe (e.g., updating a cleaning recipe, updating a seasoning recipe, updating a calibration recipe, updating a qualification recipe, etc.), updating equipment constants, scheduling preventative maintenance (e.g., cleaning, calibration, part replacement, etc.), scheduling corrective maintenance, issuing an alert to a user, or the like.

[0043] The controller 109 and / or the processing unit 130 may be and / or include a computing device such as a personal computer, a server computer, a programmable logic controller (PLC), a microcontroller, etc. The controller 109 and / or the processing unit 130 may include (or be) one or more processing units, which may be general-purpose processing units such as a microprocessor, a central processing unit, or the like. More specifically, the processing units may be complex instruction set computing (CISC) microprocessors, reduced instruction set computing (RISC) microprocessors, very long instruction word (VLIW) microprocessors, or processors implementing other instruction sets or combinations of instruction sets. The processing units may also be one or more special-purpose processing units such as an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), a network processor, or the like. The controller 109 and / or processing unit 130 may include data storage devices (e.g., one or more disk drives and / or solid state drives), main memory, static memory, a network interface, and / or other components. The processing unit 130 may execute instructions to perform one or more of the methodologies and / or embodiments described herein. The instructions may be stored on a computer-readable storage medium, which may include main memory, static memory, secondary storage, and / or the processing unit (during execution of the instructions).

[0044] 2 shows a simplified side view of a substrate processing system 200 including an electromagnetic radiation probe 202, according to some embodiments. The substrate processing system includes a transfer chamber 204 and a process chamber 206 (e.g., transfer chamber 114 and process chamber 103 of FIG. 1). The transfer chamber 114 may include a transfer robot 208. The transfer robot 208 may be configured to transfer one or more substrates within the transfer chamber 204, between the transfer chamber 204 and the process chamber 206, between the process chamber 206 and another chamber (not shown) of the processing system 200, or other similar locations.

[0045] The transfer robot 208 may be coupled to or may include a robot arm 212. The robot arm 212 may be configured to support the substrate 210, to transfer the substrate 210 between portions of the substrate processing system 200, or the like. As will be described in more detail with respect to FIG. 3, the robot arm 212 may include a blade and / or an end effector.

[0046] The transfer robot 208 may be configured to extend the robot arm 212 (as indicated by the arrow shown in FIG. 2 ) to carry the substrate 210 from the transfer chamber 204 to the process chamber 206. One or more portions of the substrate 210 may pass through the field of view 214 of the probe 202. In some embodiments, the probe 202 may be positioned near an interface between chambers of the substrate processing system 200, for example, near a slit valve 216 (e.g., slit valve 111 in FIG. 1 ). In some embodiments (e.g., embodiments including the probe 202 positioned near an interface between chambers of the substrate processing system 200), a portion of the substrate 210 may be within the field of view 214 while the substrate 210 is being transferred from one chamber to another (as shown in FIG. 2 ). In some embodiments, the probe 202 is positioned at a height higher than the height of the slit valve, and the field of view 214 of the probe 202 faces downward. Alternatively, the probe 202 may be positioned at a height lower than the height of the slit valve, with the field of view 214 of the probe 202 facing upward. In some embodiments, the probe 202 may be located in a different position, for example, the probe 202 may be configured such that a portion of the substrate 210 is within the field of view 214 while the substrate 210 is in the transfer chamber 204, or the probe 202 may be configured such that a portion of the substrate 210 is within the field of view 214 while the substrate 210 is in the process chamber 206.

[0047] In some embodiments, the probe 202 may be coupled to other components associated with performing metrology, obtaining measurements of the substrate, etc. In some embodiments, the probe 202 may include a detector (e.g., an electromagnetic radiation detector). In some embodiments, the probe 202 may include a radiation collection device. In some embodiments, the probe 202 may be coupled to a spectrometer, e.g., the probe 202 may be coupled to a fiber optic cable that may direct radiation received by the probe 202 from the substrate 210 to the spectrometer. In some embodiments, the probe 202 may be coupled to a detector. In some embodiments, one or more detectors associated with the probe 202 may be communicatively coupled to a processing device.

[0048] In some embodiments, the probe 202 may receive radiation emitted by the substrate 210. For example, the probe 202 may receive infrared radiation emitted by the substrate 210. A processing device associated with the probe 202 may determine one or more characteristics of the substrate 210 based on the radiation emitted by the substrate 210. The processing device may determine the temperature of the substrate 210 based on the emitted radiation, the chemical composition of the substrate 210, etc. In some embodiments, the probe 202 may comprise part of a pyrometer system.

[0049] In some embodiments, the probe 202 may be configured to receive light reflected by the substrate 210. The substrate processing system 200 may include an electromagnetic radiation source configured to illuminate at least a portion of the substrate 210. In some embodiments, illumination of the substrate 210 may be performed from above, e.g., to allow reflected radiation to be collected by the probe 202. In some embodiments, the illumination may be broadband illumination, e.g., provided by a white light source. In some embodiments, the illumination may be narrowband illumination, e.g., providing a limited range of wavelengths to the substrate 210. In some embodiments, the illumination may be pulsed illumination. A spectrometer associated with the probe 202 may operate in a pulsed mode, which may correspond to (e.g., be synchronized in time with) the pulsed mode of the illumination source. The illumination source may be a pulsed flash lamp, e.g., a xenon flash lamp. In some embodiments, illumination may be provided via components of the probe 202; for example, the probe 202 may include two optical fiber cores: one optical fiber core for providing radiation from a radiation source to the substrate 210, and one optical fiber core for receiving radiation reflected from the substrate 210 and providing the reflected radiation to an associated spectrometer.

[0050] In some embodiments, the radiation used to illuminate the substrate 210 may be collimated or substantially collimated. Collimated illumination may be substantially insensitive to the distance of the substrate from the probe. In some embodiments, the radiation used to illuminate the substrate 210 may be focused radiation (e.g., convergent radiation). Convergent illumination may make the metrology system substantially insensitive to the substrate angle (e.g., the angle of the substrate's surface relative to the optical axis of the metrology system). In some embodiments, the radiation used to illuminate the substrate 210 may be divergent illumination. In some embodiments, the collected radiation (e.g., spectral data) may be compared to a reference, e.g., a known silicon substrate reference. Comparison to the reference substrate may be used for background subtraction, calibration, etc.

[0051] In some embodiments, differences in substrate placement may be compensated for. For example, additional equipment may be included to perform measurements of the angle of the surface of the substrate 210. Data may be provided to the processing equipment indicating the angle at which the substrate was positioned, the angle of a portion of the substrate (e.g., due to bending or warping), or other similar angles to correct conclusions drawn from the in-chamber metrology system.

[0052] In some embodiments, the probe 202 may be configured to receive light that has been transmitted through the substrate 210. An electromagnetic radiation source may be located on the opposite side of the probe 202 (with respect to the substrate 210), for example, below the path taken by the substrate 210. The radiation may pass through the substrate 210 (e.g., through one or more openings in the bottom of the robot arm 212) and be collected by the probe 202.

[0053] In some embodiments, the probe 202 may be one probe in an array of probes for collecting electromagnetic radiation from the substrate 210. In some embodiments, several similar probes may be arranged within the substrate processing system 200. For example, several probes may be arranged so that different portions of the substrate 210 are captured by the probe's field of view. This may achieve, for example, generating a more complete map of the substrate's properties, improving noise reduction, and / or reducing errors in conclusions or analyses based on data from the probes. In some embodiments, probes measuring different quantities may be utilized. For example, the substrate processing system 200 may include probes configured to operate in a combination of emission, reflection, and transmission modes. In some embodiments, multiple probes operating in different wavelength ranges (e.g., targeting different portions of the electromagnetic spectrum) may be included.

[0054] In some embodiments, the substrate processing system 200 may further include components configured to detect the presence and / or position of the substrate 210. For example, a radiation source may be activated when the substrate enters the field of view 214 (e.g., an object detection system may provide an indication that the substrate has entered the field of view 214, e.g., by having a field of view that overlaps with that of the field of view 214, by having a field of view in which the substrate 210 enters earlier, and providing a delayed signal commanding the radiation source to activate, etc.), or a spectrometer may be activated when the substrate enters the field of view 214. In some embodiments, the object detection component may facilitate synchronization of the operation of the spectrometer with the operation of the robot, for example, synchronizing a pulsed light source (e.g., utilizing an object detection system that determines the rate of movement of the substrate 210 in cooperation with the control system of the robot 208) to enable data collection from a target portion of the substrate 210, e.g., a portion near the outer edge, a portion at a given distance from the outer edge, etc. In some embodiments, data points may be collected at several intervals as the substrate 210 passes through the field of view 214. In some embodiments, pulses of the pulsed light source may correspond to data points. In some embodiments, the integration time (together with the robot speed) of a system utilizing a continuous radiation source may define the size of the area of ​​the substrate 210 that corresponds to a data point.

[0055] In some embodiments, the robot arm 212 may deviate from a straight path as the robot arm 212 is extended. For example, the placement and / or orientation of the substrate 210 relative to the probe 202 may change as the robot arm 212 extends from a position where the first portion of the substrate 210 (e.g., the leading edge) enters the field of view 214 to a position where the last portion of the substrate 210 (e.g., the trailing edge) exits the field of view 214. In some embodiments, the distance between the substrate 210 and the probe 202 may change as the robot arm 212 is extended and / or retracted. In some embodiments, the angle between the substrate 210 (e.g., the angle of a ray normal to the surface of the substrate 210) and the field of view 214 (e.g., a ray passing through the center of the field of view 214, the optical axis of the probe, etc.) may change as the robot arm 212 is extended and / or retracted. In some embodiments, the strain (e.g., the angular droop of the robot arm 212, the vertical droop of the robot arm 212) may have different values ​​during the extension and retraction operations. In some embodiments, the behavior of the robot arm 212 during extension / retraction operations may be repeatable, predictable, accounted for by one or more processing devices associated with performing measurements of the substrate in a controlled environment, etc.

[0056] 3A is a schematic diagram of a metrology system 300A including an array of probes 304, according to some embodiments. Probe 302 is configured to receive electromagnetic radiation from a substrate 310 (e.g., substrate 110 of FIG. 1). In some embodiments, the array of probes 304 receives electromagnetic radiation emitted by the substrate 310. In some embodiments, the array of probes 304 receives radiation transmitted through the substrate 310. In some embodiments, the array of probes 304 receives radiation reflected from the substrate 310. In some embodiments, probes 302, 306, and 308 may comprise an array of probes. Each probe may have an associated field of view (e.g., indicated by a dashed cone).

[0057] In some embodiments, for example, each of the probes 304 in the array may be substantially similar, e.g., configured to perform a similar operation or receive electromagnetic radiation in a similar wavelength range. In some embodiments, one or more probes may perform different functions. For example, probe 302 may be configured to receive emitted radiation in a first frequency range (e.g., to receive infrared radiation to perform temperature measurements of the substrate, operating as a pyrometer), probe 306 may be configured to receive reflected radiation in a second frequency range (e.g., to receive radiation in the visible wavelength range provided by a source and reflected from one or more surfaces of substrate 310), and probe 308 may be configured to receive transmitted radiation in a third frequency range (e.g., to receive radiation with a wavelength of approximately 1 μm provided by a source from the opposite side of substrate 310).

[0058] The substrate 310 may be supported by a blade 312 (e.g., comprising the end effector 115 of FIG. 1 ). The blade 312 may be coupled to a robot arm, a transfer robot, or any mechanism configured to transfer the substrate 310 from one location (e.g., a first portion of the substrate processing system) to a second location (e.g., a second portion of the substrate processing system). The array of probes 304 may be stationary (e.g., fixed to a wall of the substrate processing system) and may take measurements of the substrate 310 as it is transferred (e.g., through the field of view of one or more probes of the array of probes 304).

[0059] In some embodiments, a probe (e.g., probe 302) may receive radiation continuously. For example, a radiation source optically coupled to probe 302 may be a continuous source. In some embodiments, a detector, sensor, spectrometer, etc. associated with the probe may be configured to process radiation non-continuously. For example, a spectrometer may integrate received data over a range of time (e.g., as substrate 310 is moved by blade 312), pause (e.g., separate a first region of substrate 310 from a second region of substrate 310 in the field of view of probe 302 during data integration), and then integrate for a range of time. In some embodiments, the processing system may include a pulsed source. For example, the processing system may include a pulsed flash lamp source. A probe (e.g., probe 302) may receive radiation from a source according to which the source is generating radiation, e.g., alternately measuring and not measuring regions of substrate 310 along a path across substrate 310 as substrate 310 moves through the field of view of probe 302. In some embodiments, the radiation received by the probe 302 while the source is not providing radiation may be utilized in correction, calibration, baselining operations, etc., associated with the analysis of the received radiation.

[0060] FIG. 3A also illustrates an exemplary data collection path / trajectory 314. In some embodiments, a transfer robot is configured to move the substrate 310 along one or more trajectories (e.g., straight trajectory segments, curved trajectory segments, etc.) to intersect one or more fields of view of the array of probes 304. The path 314 indicates an area of ​​the substrate 310 from which the array of probes 304 may receive radiation. The path 314 may be associated, for example, with transferring the substrate 310 from one location to another (e.g., between a transfer chamber and a process chamber). The trajectory of the substrate 310 resulting in measurements along the path 314 may be a normal part of the operation of the processing system; for example, data may be collected along the path 314 without affecting substrate processing time, system throughput, etc. Performing in-vacuum metrology may be throughput-transparent. For example, data may be acquired along the path 314 while the robot arm supporting the substrate 310 is extended or retracted. Data may be acquired along a path similar to path 314 using a single probe, an array of probes, multiple arrays of probes (e.g., probes both above and below substrate 310), etc. Data (e.g., spectral data, process data, substrate property data, etc.) associated with radiation received from substrate 310 may be synchronized with the movement of blade 312 (e.g., to generate a map of data across a portion of the surface of substrate 310, generate a map of properties across a portion of the surface of substrate 310, extrapolate a map of properties across a larger portion of substrate 310 based on data associated with path 314, etc.). Additional components (e.g., object detection systems, object detection devices, etc.) may be utilized to synchronize the movement of blade 312 with the collection / analysis of data associated with probe 308 receiving radiation from substrate 310.

[0061] In some embodiments, the substrate 310 may be patterned, e.g., a patterned wafer. Processing of data received from a patterned wafer may include additional operations. For example, spectral data received from a patterned wafer may be compared to a database of such data. Data (e.g., spectra) substantially similar to data received from a wafer may be used to characterize the substrate, characterize the pattern, etc. Models, e.g., models based on physical phenomena, machine learning models, statistical models, etc., may be used to bin similar spectra, compare spectra, extract feature values, etc.

[0062] 3B is a schematic bottom view (as viewed from below) of a blade 320 of a processing system 300B, according to some embodiments. The blade 320 may support a substrate 322. The blade 320 may be coupled to a robot arm of a transfer robot of the processing system. The metrology system 300B may include metrology components 324 and 326. In some embodiments, one or more of the metrology components 324 and 326 may be probes. One or more of the metrology components 324 and 326 may be a source, e.g., configured to direct radiation at the substrate 322.

[0063] In some embodiments, a portion of the bottom of the substrate 322 (e.g., the bottom surface of the substrate 322 supported by the blade 320) may be optically accessible while the substrate 322 is on the blade 320. As shown in FIG. 3B, the blade 320 may allow a portion of the substrate 322 to be visible from below (e.g., the blade 320 may be configured to contact the substrate 322 along the outer edge of the substrate 322). The metrology system 300B may include components both above and below the path along which the substrate 322 travels (e.g., while being transferred between chambers of a processing system). In some embodiments, one or both of the components 326 and 324 may represent arrays, such as an array of probes and an array of radiation sources, or the like. In an exemplary embodiment, the component 324 may be configured to transmit light to the substrate 322, and the component 326 may be a probe configured to receive light transmitted through the substrate 322. In additional exemplary embodiments, component 324 may perform operations including providing electromagnetic radiation to substrate 322 and receiving radiation reflected from substrate 322, and component 326 may receive radiation emitted from substrate 322 (e.g., to generate an indication of the temperature of substrate 322).

[0064] In some embodiments, additional data may be collected from additional probes. For example, probes positioned above or below the plane of substrate movement may be used to detect substrate geometry, such as substrate shape, substrate distortion, and other substrate dimensions. Data received from these probes may be used to correct or adjust data received by other probes. For example, to perform reflectometry, probes may be used to measure the angle of the top surface of the substrate, e.g., the angle of the top surface of the substrate relative to an illumination system. Data generated by receiving reflected light corresponding to the measured angle of a portion of the top surface of the substrate may be corrected. Wafer shape, distortion, bow, bending, and the like may be corrected in a similar manner.

[0065] 3C is a schematic diagram of a metrology system 300C including an array of probes 304 and a path 318 (e.g., a collection trajectory), according to some embodiments. Metrology system 300C may share one or more features with metrology system 300A of FIG. 3A, e.g., metrology system 300C includes an array of probes (e.g., probes 302 and 308) having associated fields of view, a substrate 310, a blade 312, etc. In some embodiments, a transfer robot is configured to move substrate 310 along one or more trajectories (e.g., straight trajectory segments, curved trajectory segments, etc.) to intersect one or more fields of view of the array of probes 304.

[0066] The path 318 includes several trajectories. The trajectory of the path 318 may be, for example, straight (as shown), curved, and may have a variable or uniform length. The robot arm may be configured to move the substrate 310 on a path that is not limited to moving the substrate 310 closer to a destination; for example, the robot arm may extend and retract multiple times repeatedly along the same path, or may extend and retract multiple times along two or more different (e.g., parallel) paths. In some embodiments, the robot arm moves laterally between extensions / retractions, for example, to provide a larger area of ​​the substrate 310 for one or more fields of view of the probes in the array of probes 304. In some embodiments, the robot arm does not move laterally between extensions / retractions, for example, to provide more data (e.g., repeated measurements) of a first region of the substrate 310, to provide more statistical certainty, to fill gaps (e.g., due to an integration window, pulse source, etc.), etc.

[0067] 3D is a schematic diagram of a metrology system 300D including an array of probes 304 and a collection trajectory 316, according to some embodiments. Metrology system 300D may share one or more features with metrology system 300A of FIG. 3A, for example, metrology system 300D includes probes 302 and 306, a substrate 310, a blade 312, etc. A transfer robot may be configured to move substrate 310 along the trajectory shown as collection trajectory 316.

[0068] Collection trajectory 316 includes several trajectories, e.g., several segments of a path along which a robotic arm moves the substrate, several segments of a path along which a field of view traces the surface of substrate 310, etc. The trajectories may be, for example, straight (as shown), curved, or of varying or uniform length. In some embodiments, an indirect path (e.g., a path that does not move substrate 310 to its destination with high efficiency) may be used. If the field of view of the array of probes 304 intersects with substrate 310 in a pattern similar to collection trajectory 316, e.g., a pattern different from a straight path (e.g., path 314 in FIG. 3A), a probe (e.g., a probe in array of probes 304) may receive radiation from a larger portion of substrate 310. Radiation may be received along one or more collection trajectories (e.g., a trajectory including paths / trajectories 314, 316, 318, etc.) by a single probe, an array of probes, multiple arrays of probes, etc.

[0069] FIG. 4 is a block diagram illustrating an exemplary system architecture 400 in which embodiments of the present disclosure may function. As shown in FIG. 4 , system architecture 400 includes a manufacturing system 402, a metrology system 410, a client device 450, a data store 440, a server 420, and a modeling system 470. Metrology system 410 may include in-system tools 415 (e.g., metrology tools located within an area where one or more chambers of the manufacturing system provide a controlled environment, a vacuum environment, etc.). In some embodiments, metrology system 410 may include metrology tools 414, which may include components described with respect to FIGS. 1-3 , probes, spectrometers, radiation sources, etc. Modeling system 470 may be part of server 420. In some embodiments, one or more components of modeling system 470 may be fully or partially integrated into client device 450. The manufacturing system 402, the metrology system 410, the client device 450, the data store 440, the server 420, and the modeling system 470 may each be hosted on one or more computing devices, including a server computer, a desktop computer, a laptop computer, a tablet computer, a notebook computer, a personal digital assistant (PDA), a mobile communication device, a mobile phone, a handheld computer, or a similar computing device.

[0070] The manufacturing system 402, the metrology system 410, the client device 450, the data store 440, the server 420, and the modeling system 470 may be coupled to each other (e.g., to perform the methodologies described herein) via a network 460. In some embodiments, the network 460 is a private network that provides each element of the system architecture 400 with access to each other and to other privately available computing devices. The network 460 may include one or more wide area networks (WANs), local area networks (LANs), wired networks (e.g., Ethernet networks), wireless networks (e.g., 802.11 networks or Wi-Fi networks), cellular networks (e.g., Long Term Evolution (LTE) networks), routers, hubs, switches, server computers, and / or any combination thereof. In some embodiments, processing and / or data storage operations may be performed by cloud-based remote devices, virtual devices, etc., and the network 460 may enable access to and communication with these devices. Alternatively, or in addition, any of these elements of system architecture 400 may be integrated together or otherwise coupled without the use of network 460 .

[0071] Client device 450 may be or include any personal computer (PC), laptop, mobile phone, tablet computer, netbook computer, network-connected television ("smart TV"), network-connected media player (e.g., Blu-ray player), set-top box, over-the-top (OTT) streaming device, operator box, etc. Client device 450 may include a browser 452, an application 454, and / or other tools described and executed by other systems in system architecture 400. In some embodiments, client device 450 may be able to access manufacturing system 402, metrology system 410, data store 440, server 420, and / or modeling system 470, and may be able to communicate (e.g., send and / or receive) sensor data, processed data, data classification, process result data, process update data instructions, and / or inputs and outputs of various process tools (e.g., metrology tool 414, data preparation tool 416, chamber matching tool 422, sensor matching tool 424, recipe development tool 426, process control tool 428, and / or prediction component 494) at various stages of processing of system architecture 400, as described herein.

[0072] 4, the manufacturing system 402 includes process tools 404, process steps 406, and a process controller 408. The process controller 408 may coordinate the operation of the process tools 404 to perform one or more process steps 406. For example, the various process tools 404 may include specialized chambers such as etch chambers, deposition chambers (including chambers for atomic layer deposition, chemical vapor deposition, plasma-enhanced atomic layer deposition, or plasma-enhanced chemical vapor deposition), anneal chambers, and / or other similar chambers. In another example, a machine may include a sample transport system (e.g., a selective compliance assembly robot arm (SCARA) robot, a transfer chamber, a front opening pod (FOUP), a side storage pod (SSP), and / or other similar) for transporting samples between machines and process operations.

[0073] In some embodiments, one or more of the metrology tools 414 may be utilized while a substrate is being transported, for example, between machines, process operations, in or out of the fabrication system 402. For example, a radiation source of the metrology tool 414 may provide electromagnetic radiation that is directed at a substrate being transferred from a transfer chamber to a process chamber of the fabrication system 402. The radiation source may be coupled to a component of the in-system tool 415 that directs the radiation at the substrate. A probe of the in-system tool 415 may receive radiation reflected from the substrate and provide the radiation to a spectrometer in the metrology system 410 for further analysis.

[0074] The process steps 406, which may also be referred to as process recipes, process operations, or process operations, may include various specifications for performing operations with the process tool 404. For example, the process steps 406 may include process specifications such as the run-up duration of the process operation, the process tool to use for that operation, machine (e.g., chamber) temperatures, flow rates, pressures, etc., deposition sequence, and the like. In another example, the process steps may include transfer instructions to transport a sample to additional process operations or to transport a sample for measurement by the metrology system 410. One or more process steps 406 may include preventative maintenance procedures and cleaning procedures.

[0075] The process controller 408 may include devices designed to manage and / or regulate the operation of the process tool 404. In some embodiments, the process controller 408 is associated with a process recipe, or a set of process step instructions, that when applied as designed, results in a desired process outcome for the substrate process. For example, a process recipe may be associated with processing a substrate to produce a target process result (e.g., critical dimension, thickness, uniformity criteria, etc.). The process controller 408 may coordinate the execution of preventive maintenance procedures (e.g., chamber cleans), chamber recovery procedures (e.g., chamber seasonings), etc. to return the chamber to a production state, and / or updates to process operations associated with the processing chamber. Additionally, the process controller 408 may cause updates to the calibration of sensors associated with the process tool 404 or the metrology system 410.

[0076] As shown in FIG. 4 , the metrology system 410 includes a metrology tool 414 and a data preparation tool 416. The metrology tool 414 can include various sensors for measuring process results (e.g., critical dimensions, thickness, uniformity, etc.) and detecting parameters within the fabrication system 402. In some embodiments, the metrology tool 414 includes one or more components for generating, directing, and / or receiving electromagnetic radiation, as described in FIGS. 2-3 . In some embodiments, the metrology tool 414 includes sensors for measuring one or more characteristics of a substrate within a processing system (e.g., a fabrication system). The metrology tool 414 may include tools for performing in-line metrology, integrated metrology, stand-alone metrology, sub-vacuum metrology, etc. The metrology tool 414 may include multiple sensors for detecting conditions of the environment of a processing chamber. For example, sensors disposed within one or more processing chambers can be used to measure conditions within the processing chambers, such as temperature, pressure, vacuum conditions, deposition layer thickness, and the like. In another example, a metrology tool (e.g., a probe, which may be coupled to one or more spectrometers) located within a processing chamber may be used to measure one or more properties of a substrate in the fabrication system 402. In some embodiments, one or more sensors in the system 400 (e.g., in-system tools 415) may be utilized to collect measurements (e.g., spectral data, data from reflected, transmitted, or emitted radiation, etc.) from a substrate as it is being transported by a robot in the fabrication system. In some embodiments, one or more sensors in the system 400 may be utilized to determine the chemical properties (e.g., chemical composition) of a substrate, a layer of a substrate, etc. In some embodiments, one or more sensors in the system 400 may be utilized to determine processing conditions associated with a substrate, recommend corrective actions associated with the metrology system 410 and / or the fabrication system 402, etc. In some embodiments, a sensor in the system 400 may be utilized to determine the thickness of a substrate. In some embodiments, a metrology tool may be utilized to measure the thickness of a process product deposited on a surface of a substrate.The metrology tools 414 may also include devices for measuring process results of substrates processed using the manufacturing system. For example, process results such as critical dimensions, thickness measurements (e.g., film layers from etching, deposition, etc.) of substrates processed according to process recipes and / or operations executed by the process controller 408 may be evaluated. In another example, these various sensors may measure chamber conditions throughout a process operation.

[0077] The data preparation tool 416 may include process methodologies for extracting features and / or generating synthetic / engineered data related to the data measured by the metrology tool 414. In some embodiments, the data preparation tool 416 may identify correlations, patterns, and / or anomalies in the metrology and / or process execution data. For example, the data preparation tool 416 may perform feature extraction, in which the data preparation tool 416 uses a combination of measurement data to determine whether a decision criterion is met. For example, the data preparation tool 416 may analyze multiple data points of relevant parameters (e.g., temperature, pressure, deposition thickness, vacuum conditions, plasma conditions, process product thickness, spectral data of radiation at various wavelengths, etc.) to determine whether rapid changes occurred during iterations of a process operation. In some embodiments, the data preparation tool 416 performs normalization across various sensor data associated with various process chamber conditions. Normalization may include processing the incoming sensor data so that it appears similar across the various chambers and sensors used to acquire the data. The data preparation tool 415 may prepare the data to be fed to another component, such as the model 490, for further processing.

[0078] In some embodiments, the data preparation tools 416 may perform one or more of the following analyses on the metrology data (e.g., acquired by the metrology tools 414): a process control analysis, a univariate limit violation analysis, or a multivariate limit violation analysis. For example, the data preparation tools 416 may perform statistical process control (SPC) by using a statistical-based methodology to monitor and control the process controller 408. For example, SPC may increase the efficiency and accuracy of process operations (e.g., by identifying data points that fall within and / or outside of control limits).

[0079] In some embodiments, the performance (e.g., quality of performance) of one or more processing chambers may be measured throughout a processing procedure (e.g., a recipe for processing a substrate). For example, the properties (e.g., thickness) of a process product deposited on a substrate in a processing chamber during a process operation (e.g., during a process recipe operation, during a cleaning operation, etc.) may be measured. For example, a substrate may be processed in multiple chambers. The substrate may enter a process chamber and be measured by one or more probes (e.g., by in-system tools 415, spectrometer 125 in FIG. 1, etc.) as the substrate is transferred into the chamber. The substrate may then undergo a process in that chamber, e.g., a deposition process, an etch process, etc. The substrate may then be removed from the processing chamber and be measured again by one or more probes while the substrate is transferred out of the processing chamber. In some embodiments, an increasing amount of metrology data is acquired during / between a given processing operation. For example, additional sensors may be activated immediately after processing a wafer, and / or currently activated sensors may acquire additional data. In some embodiments, the process controller 408 may trigger measurements by the metrology tool 414 based on the operation performed by the process tool 404. For example, the process controller 408 may trigger activation of one or more sensors and / or one or more radiation sources (e.g., of the metrology tool 414) in response to process-related updates, such as updates related to the completion of a processing operation, the transfer of a substrate from one location to another within the processing system, etc.

[0080] In some embodiments, the extracted features, generated synthetic / engineered data, and statistical analysis can be used in connection with the modeling system 470 (e.g., to train, validate, and / or test the model 490). Additionally and / or alternatively, the data preparation tool 416 can output data to the server 420 for use by any of the chamber matching tool 422, the sensor matching tool 424, the recipe development tool 426, and / or the process control tool 428.

[0081] The data store 440 may be a memory (e.g., random access memory), a drive (e.g., a hard drive, a flash drive), a database system, or another type of component or device capable of storing data. The data store 440 may store one or more pieces of historical data 442, including historical sensor data 444 of previous chamber conditions, historical metrology data 446, and / or historical chamber parameter data 447, and process results of substrates processed at related chamber conditions. In some embodiments, the historical data 442 may be provided to train the model 490. The trained model 490 may be provided with sensor data to obtain an output. In some embodiments, the historical sensor data 444 includes historical data from the metrology system 410, for example, spectral data generated from radiation received from a substrate by a probe. The historical sensor data 444 may include optical spectral data indicative of the thickness of a layer deposited on the substrate. The historical sensor data 444 may be indicative of another characteristic of the substrate, such as the processing conditions of the substrate.

[0082] In some embodiments, historical chamber parameters 447 may include historical updates to one or more process operations associated with the processing chamber. For example, historical chamber update data 447 may include historical updates (e.g., cleaning recipe updates, etch recipe updates, deposition recipe updates, etc.) made to one or more process operations associated with the processing chamber based on historical sensor data 444. In some embodiments, historical chamber parameters data 447 may be mapped to historical sensor data 444, historical metrology data 446, etc. In some embodiments, historical data 442 may be used to train, validate, and / or test models 490 of modeling system 470 (see, e.g., FIG. 6 for an exemplary methodology).

[0083] The server 420 may include one or more computing devices such as a rack-mounted server, a router computer, a server computer, a personal computer, a mainframe computer, a laptop computer, a tablet computer, a desktop computer, etc. The server 420 may include a chamber matching tool 422, a sensor matching tool 424, a recipe development tool 426, and a process control tool 428.

[0084] Each of the chamber matching tool 422, the sensor matching tool 424, the recipe development tool 426, and the process control tool 428 may receive data from the data preparation tool 416. Each of the chamber matching tool 422, the sensor matching tool 424, the recipe development tool 426, and the process control tool 428 may receive data from the model 490. Each component of the server 420 may receive data from the prediction component 494.

[0085] In some embodiments, the chamber matching tool 422, the sensor matching tool 424, the recipe development tool 426, and the process control tool 428 each receive chamber data from the process tool 404 and / or processed data output from the data preparation tool 416. In some embodiments, the data preparation tool 416 receives raw sensor data from the metrology tool 414. This raw sensor data may be combined with engineered synthetic data from the data preparation tool 416. The chamber matching tool 422, the sensor matching tool 424, the recipe development tool 426, and the process control tool 428 each may generate updates to one or more process steps 406 and / or generate reports / notifications indicating chamber matching, chamber drift, anomalies, etc.

[0086] In one example, the chamber matching tool 422 may determine updates to the process operation of the first process chamber based on the received data to more closely match the process operation performed on the first process chamber with a corresponding process operation performed on the second process chamber.

[0087] In another example, the sensor matching tool 424 may determine a calibration for one or more sensors associated with the process chamber based on the received value data. Additionally, the sensor matching tool 424 may determine a sensor calibration that accounts for sensor drift based on the received value data.

[0088] In an additional example, the recipe development tool 426 may determine updates to a process recipe based on the received data. Specifically, the recipe development tool 426 may determine updates to one or more process parameters (e.g., run time, clean time, process knobs, etc.) based on the received data to optimize the process recipe (e.g., process operation). In one embodiment, the recipe development tool 42 determines updates to a cleaning recipe that is periodically executed to clean a process chamber based on the received data.

[0089] As a further example, the process control tool 428 may determine updates to the endpoint of a process operation (eg, the endpoint of a chamber clean operation) based on the received data.

[0090] These updates to the process step may change the way the process tool 404 performs the process step 406. For example, the updates may include extending or shortening the process time of one or more process operations in the process step 406. For example, the updates may extend or shorten the duration of a chamber clean operation. In some embodiments, the chamber matching tool 422, the sensor matching tool 424, the recipe development tool 426, and the process control tool 428 may each use a model (e.g., a model in the modeling system 470) that receives input data, such as chamber data from the process tool 404, the metrology tools 414, including the in-system tools 415, and / or processed data output from the data preparation tool 416, and outputs one or more updates to the process step 406. The model 490 may be a machine learning model, including, among others, a bootstrap aggregation model, a random forest tree decision tree model, a partial least squares regression (PLS) model, a least absolute shrinkage and selection operator (LASSO) regression model, and / or a Ridge regression model. The model may include ensemble modeling, which includes multiple models (e.g., stacked modeling) and utilizes a more reliable model for final classification of received data. The model may be unsupervised, semi-supervised, or supervised. Additionally, the model may include linear models, dynamic models, digital models, mechanistic models, statistical models, models based on physical phenomena, digital twin models, etc.

[0091] In some embodiments, one or more of the chamber matching tool 422, the sensor matching tool 424, the recipe development tool 426, and / or the process control tool 428 may implement their described methodologies using machine learning models. In some embodiments, one or more of the chamber matching tool 422, the sensor matching tool 424, the recipe development tool 426, and / or the process control tool 428 may implement their described methodologies using traditional models and / or physics-based models (e.g., digital twins). These relevant models may be generated (e.g., trained, validated, and / or tested) using the modeling system 470.

[0092] In some embodiments, modeling system 470 may include server machine 472 and server machine 480. Server machine 472 includes a dataset generator 474 that can generate one or more datasets (e.g., a set of data inputs and a set of target outputs) for training, validating, and / or testing model 190. Some operations of dataset generator 474 are described in detail with respect to FIGS. 5 and 7A . In some embodiments, dataset generator 172 may divide historical data (e.g., historical sensor data 444, historical measurement data 446, chamber parameter data 447, etc.) into a training set (e.g., 60 percent of the data), a validation set (e.g., 20 percent of the data), and a test set (e.g., 20 percent of the data). In some embodiments, modeling system 470 generates (e.g., via prediction component 494) sets of attributes (e.g., feature vectors, vectors, etc.). For example, the first set of attributes may correspond to a first set of types of sensor data (e.g., from a first set of sensors, a first combination of values ​​from the first set of sensors, a first pattern of values ​​from the first set of sensors) corresponding to each of the data sets (e.g., a training set, a validation set, and a test set), and the second set of attributes may correspond to a second set of types of sensor data (e.g., from a second set of sensors different from the first set of sensors, a second combination of values ​​different from the first combination, a second pattern different from the first pattern) corresponding to each of the data sets.

[0093] In some embodiments, server machine 480 includes a training engine 482, a verification engine 484, a selection engine, and / or a test engine 486. An engine (e.g., training engine 482, verification engine 484, test engine 486) may refer to hardware (e.g., circuitry, dedicated logic circuitry, programmable logic circuitry, microcode, processing unit, etc.), software (e.g., instructions executing on a processing unit, general-purpose computer system, or dedicated machine), firmware, microcode, or a combination thereof. Training engine 482 may generate multiple trained models 490. Each trained model in the multiple trained models may correspond to a different set of attributes of the training set (e.g., sensor data from a different set of sensors). For example, a first trained machine learning model may be trained using all attributes (e.g., X1-X5), a second trained machine learning model may be trained using a first subset of attributes (e.g., X1, X2, and X4), and a third trained machine learning model may be trained using a second subset of attributes (e.g., X1, X3, X4, and X5), where the second subset of attributes may overlap with the first subset of attributes. A dataset generator 474 may receive the output of a trained model (e.g., 490), assemble the data into training, validation, and test datasets, and use those datasets to train a second model. Some or all of the operations of server machine 480 may be used to train various types of models, including models based on physical phenomena, supervised machine learning models, unsupervised machine learning models, etc.

[0094] The validation engine 484 may be capable of validating the trained models 490 using a corresponding set of features in the validation set from the dataset generator 474. For example, a first trained model 490 trained using a first set of attributes in the training set may be validated using the first set of attributes in the validation set. The validation engine 484 may determine the accuracy of each of the trained models 490 based on the corresponding set of features in the validation set. The validation engine 484 may discard trained models 490 with an accuracy that does not meet a threshold accuracy. In some embodiments, a selection engine (not shown) may be capable of selecting one or more trained models 490 with an accuracy that meets the threshold accuracy. In some embodiments, the selection engine may be capable of selecting the trained model 490 with the highest accuracy among the trained models 490.

[0095] The testing engine 486 may be capable of testing the trained model 490 using a corresponding set of attributes of a test set from the dataset generator 474. For example, a first trained model 190 trained using a first set of attributes of a training set may be tested using a first set of attributes of a test set. The testing engine 486 may determine the trained model 490 that has the highest accuracy of all of the trained models based on the test set.

[0096] Model 490 may refer to a machine learning model, which may be a model artifact generated by training engine 482 using a training set including data inputs and corresponding target outputs (correct answers for each training input). Additionally or alternatively, model 490 may refer to a statistical model or a model based on physical phenomena. Patterns in a data set that map data inputs to target outputs (correct answers) can be found, and model 490 is provided with a mapping that captures these patterns. In some embodiments, model 490 may predict characteristics of a substrate. In some embodiments, model 490 may predict failure modes of a manufacturing chamber component.

[0097] One type of machine learning model that may be used to perform some or all of the above tasks is an artificial neural network, such as a deep neural network. Artificial neural networks generally include a feature representation component with a classifier or recurrent layer that maps features to a desired output space. For example, a convolutional neural network (CNN) hosts multiple layers of convolutional filters. Pooling may be performed and nonlinearities may be addressed in lower layers, and above the lower layers, a multilayer perceptron is typically added to map upper layer features extracted by the convolutional layers to a decision (e.g., a classification output).

[0098] A recurrent neural network (RNN) is another type of machine learning model. Recurrent neural network models are designed to interpret a series of inputs that are intrinsically related to each other, such as time trace data, sequential data, etc. The output of a perceptron in an RNN is fed back as input to that perceptron to generate the next output.

[0099] Deep learning describes a class of machine learning algorithms that use a cascade of multiple layers of nonlinear processing units to perform feature extraction and transformation. Each successive layer uses the output from the previous layer as input. Deep neural networks may learn in a supervised (e.g., classification) and / or unsupervised manner (e.g., pattern analysis). Deep neural networks include a hierarchy of layers, with different layers learning different levels of representation corresponding to different levels of abstraction. In deep learning, each level learns to transform its input data into a slightly more abstract and complex representation. For example, in an image recognition application, the raw input may be a matrix of pixels; a first representation layer may extract the pixels and encode edges; a second layer may construct and encode the edge configuration; a third layer may encode higher-order shapes (e.g., recognizing substrate structures such as gates and masks); and a fourth layer may generate a classification output. Notably, the deep learning process can independently learn which features are optimally placed at which levels. The "deep" in "deep learning" refers to the number of layers through which data is transformed. More precisely, deep learning systems have significant credit assignment path (CAP) depth. A CAP is a chain of transformations from input to output. A CAP describes the potentially causal connections between input and output. For feedforward neural networks, the CAP depth may be the depth of the network or the number of hidden layers + 1. For recurrent neural networks, where signals may propagate through layers more than once, the CAP depth is potentially infinite.

[0100] Training a neural network may be accomplished with supervised learning, which involves feeding a training dataset consisting of labeled inputs through the network, observing its outputs, defining an error (by measuring the difference between the output and the label value), and using techniques such as deep gradient descent and backpropagation to tune the network's weights across all its layers and nodes so that the error is minimized. In many applications, repeating this process across many labeled inputs in the training dataset provides a network that can generate correct outputs when presented with inputs that differ from those present in the training dataset.

[0101] Model 490 may refer to a trained physical-phenomenon-based model. A trained physical-phenomenon-based model may be configured to find solutions to one or more equations describing physical quantities of a processing chamber, such as mass flow rate (e.g., gas flow rate), heat transfer equations, fluid dynamics equations, or the like. In some embodiments, the assumptions used to generate the physical-phenomenon-based model may not be completely accurate (e.g., due to inaccurate measurements, manufacturing or material defects, inconsistent component fabrication tolerances, component aging, drifting or behavior that differs from expected, or the like). Training a physical-phenomenon-based model may correct one or more of these assumptions that introduce error into the physical-phenomenon-based model, for example, by allowing one or more parameters of the model to be altered to better fit the training data.

[0102] The prediction component 494 may provide input data to the trained model 490 and may run the model 490 on the input to obtain one or more outputs. The prediction component 494 may be able to determine (e.g., extract) predicted data 468 from the output of the model 490, and from that output, determine (e.g., extract) confidence data indicating the confidence that the predicted data 468 is an accurate predictor of a process related to the input data or a component of the manufacturing system 402 for a product that has been or will be produced. The prediction component 494 may be able to determine the predicted data 468, including predictions for finished substrate properties and predictions of the useful life of components of the manufacturing system 402, sensors, or metrology system 410, based on the output of the model 490. The prediction component 494 or a corrective action component associated with the client device 450 (e.g., a component of the application 454) may use this confidence data to determine whether to perform a corrective action associated with the manufacturing system 402 based on the predicted data 468.

[0103] The confidence data may include or indicate a confidence level. As an example, the prediction data 468 may indicate predicted manufacturing conditions (e.g., conditions in a processing chamber while a substrate is being processed) based on in-vacuum measurements of the substrate. The confidence data may indicate that the prediction data 468 is an accurate prediction for a product associated with at least a portion of the input data. In one example, the confidence level is a real number between 0 and 1, inclusive, where 0 indicates no confidence that the prediction data 468 is an accurate prediction for a product processed according to the input data, and 1 indicates absolute confidence that the prediction data 468 accurately predicted the characteristics of a product processed according to the input data. In response to the confidence data indicating a confidence level below a threshold level for a predetermined number of instances (e.g., a percentage of instances, a frequency of instances, a total number of instances, etc.), the prediction component 494 may cause the model 490 to be retrained (e.g., by updating the training, validation, and / or test sets to include additional data, more recent data, different data, etc.).

[0104] For purposes of illustration and not limitation, aspects of the present disclosure describe training and using models using information about historical data 442. In other embodiments, heuristic or rule-based models may be used to determine process operation updates.

[0105] In some embodiments, the functionality of client device 450, server 420, data store 440, and modeling system 470 may be provided by fewer machines than shown in FIG. 4. For example, in some embodiments, server machines 472 and 480 may be combined into a single machine, and in other embodiments, server machines 472, 480, and 492 may be combined into a single machine. In some embodiments, modeling system 470 may be provided completely or partially by server 420. In some embodiments, the functionality of client device 450, server 420, data store 440, and modeling system 470 may be provided by more machines than shown in FIG. 4.

[0106] In general, functions described in one embodiment as being performed by client device 450, data store 440, metrology system 410, manufacturing system 402, and modeling system 470 may in other embodiments be performed, where appropriate, on server 420. Furthermore, functions attributed to a particular component may also be performed by different or multiple components operating together.

[0107] In some embodiments, a "user" may be expressed as a single individual. However, other embodiments of the present disclosure encompass a "user" being an entity controlled by multiple users and / or automated sources. For example, a collection of individual users united as a group of administrators may be considered a "user."

[0108] FIG. 5 is a block diagram of an example dataset generator 572 (e.g., dataset generator 474 of FIG. 4 ) used to generate a dataset for a model (e.g., model 490 of FIG. 3 ), according to some embodiments. Dataset generator 572 may be part of server machine 472 of FIG. 4 . In some embodiments, modeling system 470 of FIG. 4 may include multiple models. Each model may have a separate dataset generator, or multiple models may share one or more dataset generators. For example, different models (e.g., machine learning models, models based on physical phenomena, statistical models) may be utilized to accept different inputs (e.g., manufacturing parameters, sensor data, metrology data, spectral data, etc.), generate different outputs (e.g., predicted measurements, predicted manufacturing conditions, recommended corrective actions), engage in different types of analysis, etc.

[0109] In some embodiments, a model may be configured to accept as input data (e.g., spectral data) regarding electromagnetic radiation received from the substrate by the probe and generate as output one or more indications of a property of the substrate (e.g., thickness, geometry, e.g., critical dimension, etch depth or sidewall angle, pattern uniformity, optical properties, chemical properties, etc.). Such a model may include machine learning operations, operations based on physics, etc. In some embodiments, a model is configured to accept as input one or more indications of a property of the substrate (e.g., indications of one or more properties of the substrate at one or more locations on the substrate coincident with the field of view of one or more probes) and generate as output a predictive map of properties of other regions of the substrate (e.g., for a system having one probe, properties at locations not directly on the path within the field of view of the probe as the substrate is transferred through the field of view of the probe from one part of the processing system to another). In some embodiments, a model is configured to accept as input one or more indications of a property of the substrate and generate output indicative of processing conditions for the substrate (e.g., indicative of processing conditions that result in properties of the substrate). In some embodiments, the model receives data related to radiation received from the substrate (e.g., spectral data) and generates as output a prediction of processing conditions for the substrate. In some embodiments, the model receives data (e.g., spectral data, characteristic data, etc.) and generates one or more indications of properties of a top layer of the substrate (e.g., the model may receive data collected from the substrate before and after a processing operation). In some embodiments, the model receives input data (e.g., spectral data, characteristic data, etc.) and generates as output one or more recommended corrective actions and / or causes one or more corrective actions to be implemented.

[0110] 5-6 will focus on systems and operations related to machine learning models configured to receive spectral data (e.g., collected by a probe from a substrate being transferred by a transfer robot from one location in a processing system to a second location in the processing system) and generate as output instructions for processing conditions associated with processing the substrate. Similar methods and systems may be utilized for different machine learning models (e.g., models that receive different data as input and / or generate different data as output), different types of models (e.g., models based on physical phenomena, statistical models), etc., and may be within the scope of this disclosure.

[0111] The system 500, including the dataset generator 572, generates a dataset for a model (e.g., model 490 in FIG. 4 ). The dataset generator 572 may generate the dataset using historical data. The dataset generator 572 may generate the dataset using data generated by a metrology tool, for example, data generated by receiving electromagnetic emissions from a substrate in a processing system by one or more probes. The dataset generator 572 may generate the dataset using data indicative of processing conditions of the substrate, for example, data from an apparatus configured to measure conditions at or near the location of the substrate, sensor data from a processing chamber, etc.

[0112] The dataset generator 572 generates the data input 510. The dataset generator 572 may further generate the target output 520. In some embodiments, the dataset generator may not generate the target output 520, for example, to train an unsupervised machine learning model. The dataset generator 572 may generate the data input 510 using substrate emission data. The substrate emission data may relate to radiation emitted by a substrate. The substrate emission data may relate to radiation reflected by one or more substrates. The substrate emission data may relate to radiation transmitted through a substrate. The substrate emission data may relate to radiation received by one or more probes of a processing system (e.g., a substrate processing system) while one or more substrates are moving within the processing system (e.g., data collected while one or more substrates are being transferred between portions of the processing system). The substrate emission data may be spectral data. The substrate emission data may include a combination of data, for example, data collected by multiple probes, data associated with multiple radiation sources (e.g., reflected and emitted light), etc. The target output 520 may include process condition data. The processing condition data may be generated by equipment (e.g., designed to be located at or near the processing location). The processing condition data may be generated by sensors associated with the processing system (e.g., substrate processing system).

[0113] The dataset may include pre-processed data, such as smoothed data, cleaned data (e.g., data with outliers removed), combined data, data aggregated into feature vectors or attributes, etc.

[0114] The dataset generator 572 may generate a dataset (e.g., a training set, a validation set, a test set) that includes one or more data inputs 510 (e.g., training inputs, validation inputs, test inputs) and may include one or more target outputs 520 corresponding to the data inputs 510. The dataset may further include mapping data that maps the data inputs 510 to the target outputs 520. The data inputs 510 may also be referred to as "features," "attributes," or "information." In some embodiments, the dataset generator 572 provides one or more datasets to the training engine, the validation engine, and / or the test engine of FIG. 4. The data may be used by the engine to train, validate, or test the model 490 of FIG. 4. Some embodiments of generating a training set, a validation set, or a test set may be further described with respect to FIG. 7A.

[0115] The data set generator 474 may generate a first data input corresponding to the first set of substrate emission data 544A to train, validate, or test a first model. The data set generator 572 may generate a second data input corresponding to the second set of sensor data 544B to train, validate, or test a second model.

[0116] In some embodiments, a dataset generator 474 performs operations on one or more of the data inputs 510 and target outputs 520. The dataset generator 572 may extract patterns from the data (slope, curvature, etc.), combine the data (average, feature generation, etc.), or split the data into groups (e.g., train a model on a subset of the substrate emission data) and use those groups to train separate models.

[0117] The data inputs 510 and target outputs 520 for training, validating, or testing a model may include information for a particular substrate processing recipe (e.g., a particular substrate design). The data inputs 510 and target outputs 520 may include information for a particular substrate processing system (e.g., for a particular set of manufacturing equipment). The data inputs 510 and target outputs 520 may include information for a particular type of process, target substrate characteristics, type of processing equipment, processing facility, or may be grouped together in another manner.

[0118] In some embodiments, a dataset generator 572 generates a set of target outputs 520 that includes the set of process condition data 530. The target outputs 520 may be divided into sets that correspond to the sets of input data. Different sets of target outputs 520 may be used for similarly defined sets of data inputs 510, including training different models, using different sets for training, validation, and testing, etc.

[0119] In some embodiments, a model is trained (e.g., an unsupervised or semi-supervised model) without a target output 520. A model trained without being provided with a target output may, for example, be trained to recognize significant (e.g., outside an error threshold) differences between predicted and measured data (e.g., between process condition setpoints and process conditions implied by in-vacuum measurements of substrates being transferred by a robotic arm of a processing system).

[0120] In some embodiments, the information used to train the model may be from a particular type of manufacturing equipment, e.g., a particular type of manufacturing facility. The types of manufacturing equipment contributing to the dataset of dataset generator 572 may share one or more characteristics, such as the equipment's structure and / or model. The trained model may be able to generate outputs associated with a group of manufacturing equipment tools based on the data inputs 510 and target outputs 520 used to train the model. In some embodiments, the data used to train the model may come from more than one manufacturing facility.

[0121] In some embodiments, following generating a dataset and using the dataset to train, validate, or test a model, the model may be further trained, validated, or tested, or may be adjusted. For example, additional data may be provided to the model from substrates processed after the model has been trained, validated, and tested, as retraining data, revalidation data, retest data, or the like.

[0122] The dataset generator 572 may generate datasets for training, validating, and / or testing the model. Training the model may include generating a model mapping that the model uses to connect input data to output data.

[0123] A dataset generator performing functions similar to dataset generator 572 may be utilized to train a physical-phenomenon-based model. A physical-phenomenon-based model may be configured to generate outputs based on a physical understanding of a system, physical assumptions of the system's operation, one or more numerical solutions of one or more physical equations (e.g., heat transfer equations, mass balance equations, fluid dynamics equations, etc.), or the like. A physical-phenomenon-based model may be trained in a manner similar to a machine learning model. A physical-phenomenon-based model may be provided with training inputs and target outputs, and the physical-phenomenon-based model may adjust one or more parameters, weights, biases, or the like to better align the model output with the target output.

[0124] In some embodiments, a physical-phenomenon-based model receives a set of inputs (e.g., indicative of radiation received from a substrate by a probe). The physical-phenomenon-based model may generate an output based on the inputs (e.g., predicted processing conditions of the substrate). A target output (e.g., measured processing conditions of the substrate) may be provided to the physical-phenomenon-based model. The physical-phenomenon-based model may adjust one or more parameters of the model to produce an output that is more similar to the target output than before the adjustments were made.

[0125] 6 is a block diagram illustrating a system 600 for generating output data for performing corrective actions, according to some embodiments. System 600 may be used to train a model (e.g., a machine learning model) and utilize the output of the model. Some or all of the operations of system 600 may be used to generate output data for the machine learning model. Some or all of the operations of system 600 may be used to generate output data for models based on physical phenomena, statistical models, etc.

[0126] At block 610, system 600 (e.g., a component of modeling system 470 of FIG. 4 ) performs data partitioning of historical data 664 (e.g., via data set generator 474 of FIG. 4 ) to generate training set 602, validation set 604, and test set 606. The historical data may include data used for training input (e.g., spectral data generated in a vacuum from radiation received from one or more substrates), target output (e.g., data indicative of processing conditions for one or more substrates), etc. For example, the training set may be 60% of the historical data, the validation set may be 20% of the historical data, and the test set may be 20% of the historical data.

[0127] At block 612, the system 600 performs model training (e.g., via the training engine 484 of FIG. 4 ) using the training set 602. The system 600 may train one model or may train multiple models using multiple sets of features (e.g., attributes, feature vectors) from the training set 602 (e.g., a first set of attributes including a subset of the historical data from the training set 602, a second set of attributes including a different subset of the historical data from the training set 602, etc.). For example, the system 300 may train machine learning models to generate a first trained machine learning model using the first set of attributes in the training set and to generate a second trained machine learning model using the second set of attributes in the training set (e.g., data different from the data used to train the first machine learning model). In some embodiments, the first trained machine learning model and the second trained machine learning model may be combined to generate a third trained machine learning model (e.g., which may, alone, be a better predictor than the first or second trained machine learning models). In some embodiments, the sets of attributes used in comparing the models may overlap (e.g., one model may be trained using performance data indicative of film thickness, another model may be trained using performance data indicative of both film thickness and film stress, a different model may be trained using data from different locations on the substrate, or a model may be trained that includes input from different sets of sensors or manufacturing parameters). In some embodiments, hundreds of models may be generated, including models with various permutations of attributes and combinations of models.

[0128] At block 614, the system 600 performs model validation (e.g., via validation engine 484 of FIG. 4 ) using the validation set 604. The system 600 may validate each of the trained models using the corresponding set of features in the validation set 604. For example, the validation set 604 may use the same subset of historical data types (e.g., associated with the same substrate characteristics, the same sensors, the same input parameters, etc.) used for the models in the training set 602, but with different input data values. The system 600 may validate hundreds of models (e.g., models with various permutations of features or attributes, combinations of models, etc.) generated at block 612. At block 614, the system 600 may determine the accuracy of each of the one or more trained models (e.g., via model validation) and may determine whether one or more of the trained models have an accuracy that meets a threshold accuracy. In response to determining that none of the trained models have an accuracy that meets the threshold accuracy, flow may return to block 612, where the system 600 performs model training using a different set of attributes from the training set. In response to determining that one or more of the trained models have an accuracy that meets the threshold accuracy, flow may proceed to block 616. The system 600 may discard trained machine learning models that have an accuracy that is lower than the threshold accuracy (e.g., based on a validation set).

[0129] At block 616, the system 600 may perform model selection (e.g., via a selection engine) to determine which model of the one or more trained models that meet the threshold accuracy has the highest accuracy (e.g., selected model 608 based on the validation of block 614). If only a single model was trained (e.g., in the validation, only a single model had an accuracy higher than the accuracy threshold), the operations of block 616 may be skipped. In response to determining that two or more models of the trained models that meet the threshold accuracy have the same accuracy (e.g., within the threshold), flow may return to block 612, where the system 600 performs model training to determine the trained model with the highest accuracy using a further refined training set (e.g., corresponding to a further refined set of attributes).

[0130] At block 618, the system 600 performs model testing (e.g., via the testing engine 486 of FIG. 4 ) using the test set 606 to test the selected model 608. The system 600 may test a first model (e.g., a trained machine learning model) using a first set of features in the test set and determine that the first trained model meets a threshold accuracy (e.g., based on the first set of features of the test set 606). In response to the accuracy of the selected model 608 not meeting the threshold accuracy (e.g., the selected model 608 is too well-fitted to the training set 602 and / or the validation set 604 and cannot be applied to other datasets, such as the test set 606), flow may proceed to block 612, where the system 600 performs model training (e.g., retraining) using a different training set, perhaps corresponding to a different set of features or attributes, or a reorganization of the dataset divided into training, validation, and test sets. In response to determining, based on the test set 606, that the selected model 608 has an accuracy that meets the threshold accuracy, flow proceeds to block 620. At least in block 612, the model may learn patterns in the input data to make predictions, and in block 618, the system 600 may apply the model to the remaining data (e.g., the test set 606) to test the predictions.

[0131] In block 620, the system 600 uses the trained model (e.g., selected model 608) to receive current radiation data 654 (e.g., data based on radiation received from a substrate being processed) and, via the trained model, determine (e.g., predict) the process conditions (e.g., process condition data 669) to use to process the current substrate. The process condition data 669 may be utilized in taking corrective action to improve the process conditions of the process chamber, for example, by updating a recipe, scheduling maintenance, alerting a user, etc.

[0132] In some embodiments, the model is retrained by providing additional data to further train the model. Current radiation data 654 may be provided for additional model training (e.g., provided to the training engine for use as training inputs). Current condition data 646 (e.g., measurements of substrate processing conditions) may be provided to the model training engine for use as target outputs, for example. Based on this data, the selected model 608 may be retrained. One or more of operations 610-620 may be performed in various orders and / or with other operations not presented and described herein. In some embodiments, one or more of operations 610-620 may not be performed. For example, in some embodiments, one or more of data partitioning of block 610, model validation of block 614, model selection of block 616, or model testing of block 618 may not be performed. When training a model based on physical phenomena, a subset of the described operations may be performed.

[0133] 7A-C are flow diagrams of methods 700A-C associated with utilizing electromagnetic radiation-based in-chamber metrology, according to some embodiments. Methods 700A-C may be performed by processing logic, which may include hardware (e.g., circuits, dedicated logic circuitry, programmable logic circuitry, microcode, a processing unit, etc.), software (e.g., instructions executing on a processing unit, a general-purpose computer system, or a dedicated machine), firmware, microcode, or a combination thereof. In some embodiments, methods 700A-C may be performed in part by modeling system 470 of FIG. 4. Method 700A may be performed in part by modeling system 470 of FIG. 4 (e.g., dataset generator 474 of FIG. 4). Modeling system 470 may use method 700A to generate a dataset for at least one of training, validating, or testing a model according to embodiments of the present disclosure. The model may be a physical phenomenon-based model, a machine learning model, a statistical model, or another model trained to receive inputs and generate outputs related to substrate manufacturing or processing. Methods 700B-C may be performed by processing device 130 of Figure 1, measurement system 410 of Figure 4, etc. In some embodiments, a non-transitory storage medium stores instructions that, when executed by a processing device (e.g., a processing device of modeling system 470, a processing device of system 400, a processing device of server machine 480, etc.), cause the processing device to perform one or more of methods 700A-C.

[0134] For ease of explanation, methods 700A-C are shown and described as a series of operations. However, operations in accordance with the present disclosure may be performed in various orders and / or simultaneously, and with other operations not shown and described herein. Moreover, not all illustrated operations may be performed to implement methods 700A-C in accordance with the disclosed subject matter. Furthermore, those skilled in the art will understand and appreciate that methods 700A-C may also be represented as a series of interrelated states via a state diagram of events.

[0135] 7A is a flow diagram of a method 700A for generating a dataset for a model for generating output data, according to some embodiments. At block 701, processing logic performing method 700A initializes a training set T to be an empty set.

[0136] At block 702, processing logic generates a first data input (e.g., a first training input, a first validation input) that may include sensor data, manufacturing parameter data, measured substrate performance data, substrate metrology data (e.g., film properties such as thickness, material composition, optical properties, roughness, etc.), etc. In some embodiments, this first data input may include a first set of attributes for a type of data, and the second data input may include a second set of attributes for the type of data (e.g., as described with respect to FIG. 6 ).

[0137] At block 703, processing logic generates a first target output for one or more of these data inputs (e.g., the first data input). In some embodiments, the first target output is performance data for the substrate. In some embodiments, the first target output is data that indicates corrective action. In some embodiments, no target output is generated (e.g., to train an unsupervised machine learning model).

[0138] At block 704, processing logic may generate mapping data indicating an input / output mapping. This input / output mapping (or mapping data) may relate to a data input (e.g., one or more of the data inputs described herein), a target output for the data input, and an association between the data input and the target output. In some embodiments (e.g., embodiments in which target output data is not generated), these operations may not be performed.

[0139] In some embodiments, at block 705, processing logic adds the mapping data generated at block 404 to the dataset T.

[0140] At block 706, processing logic branches based on whether dataset T is sufficient for at least one of training, validating, and / or testing a model (e.g., model 490 in FIG. 4 ). If T is sufficient, execution proceeds to block 707; if not, execution returns to block 702. It should be noted that in some embodiments, whether dataset T is sufficient may be determined simply based on the number of inputs in the dataset, and in some embodiments based on the number of inputs in the dataset that are mapped to outputs, and in some other implementations, whether dataset T is sufficient may be determined based on one or more other criteria in addition to or instead of the number of inputs (e.g., measures such as the diversity of data examples, accuracy, span of the input and / or output space, etc.).

[0141] At block 707, processing logic provides dataset T (e.g., to server machine 480 of FIG. 4 ) for training, validating, and / or testing a model (e.g., model 190). In some embodiments, dataset T is a training set, and dataset T is provided to training engine 482 of server machine 4780 to perform training. In some embodiments, dataset T is a validation set, and dataset T is provided to validation engine 484 of server machine 480 to perform validation. In some embodiments, dataset T is a test set, and dataset T is provided to test engine 486 of server machine 480 to perform testing.

[0142] 7B is a flow diagram of a method 700B of generating and utilizing data based on radiation received by a probe while the substrate is in a processing system, according to some embodiments. Method 700B may include receiving radiation from the substrate while the substrate is in a controlled environment (e.g., in a vacuum). Method 700B may also include receiving radiation from the substrate while the substrate is in motion (e.g., while a robot is transferring the substrate from one location to another in a substrate processing system).

[0143] At block 710, processing logic receives first data. The first data is generated by a sensor of the substrate processing system (e.g., via a probe, a spectrometer, and a processing device / processor of the substrate processing system). The first data may include spectral data. The first data is generated in response to a sensor receiving electromagnetic radiation. The electromagnetic radiation is received from a substrate. The substrate is held by a robot arm in a transfer chamber within the substrate processing system. The first data may be generated in response to the electromagnetic radiation being received (e.g., by a probe, a sensor, a spectrometer, etc.) while the robot arm is transferring the substrate from a first region of the substrate processing system to a second region of the substrate processing system. In some embodiments, radiation may be directed to and received by one or more sensors while the substrate is being transferred, e.g., in a throughput-transparent manner. In some embodiments, the robot arm may be configured to change / adjust the path of the substrate to increase, change, or improve the data generated by the sensors. In some embodiments, the robot may transfer the substrate through multiple trajectories, e.g., to increase the portion of the substrate intersecting the field of view of one or more probes.

[0144] In some embodiments, data for a substrate may be recorded multiple times. For example, as a portion of the substrate passes through a sensor's field of view, electromagnetic radiation received from the substrate may be divided into several data points, e.g., several data points each corresponding to a different spatial location on the substrate. One or more of these regions (e.g., one region associated with first data, other locations associated with other data, etc.) may have associated data analyzed. By receiving electromagnetic radiation from multiple locations on the substrate, a map or sub-map of a measured property, processing condition, or the like may be generated. In some embodiments, spatially distinct (e.g., separated) data may be used to generate a profile of the substrate (e.g., a thickness profile, a temperature profile, etc.).

[0145] In some embodiments, the operation of one or more sensors may be triggered. For example, an object detection system (e.g., a laser object detection system, an object detection device, etc.) may detect the presence of a substrate at a location in a substrate processing system. In response to detecting the substrate, one or more sensor operations may be initiated, modified, etc. In some embodiments, the velocity of a robot arm supporting the substrate may be used in determining the timing of the sensor operations. In some embodiments, a detection device may be activated at a target time in response to object detection. In some embodiments, an illumination device (e.g., an electromagnetic radiation source) may be activated in response to object detection. In some embodiments, the illumination device may be operated in a pulsed mode, and its timing may be modified in response to object detection, for example, to target illumination to a particular region of the substrate (e.g., edge, distance from edge, center, etc.).

[0146] In some embodiments, additional data is generated by additional sensors, e.g., additional probes. For example, an array of probes may generate data associated with additional positions on the substrate (e.g., to improve statistical certainty, improve knowledge of the complete substrate profile, etc.). Multiple probes may report on different properties (e.g., substrate thickness and substrate temperature), different radiation wavelength ranges, different illumination conditions (e.g., reflection, transmission, emission, etc.), or the like. The data may be processed into different forms of data, e.g., temperature data, thickness data, corrective action data, process condition data, etc. Data from the additional probes and / or sensors may be utilized in the same manner as the first data.

[0147] In some embodiments, collected data (e.g., data provided by sensors) may be corrected to account for, for example, the operation of the substrate processing system. For example, a robot arm may differ from ideal behavior as a function of position, velocity, direction of movement, percent elongation, or the like. In some embodiments, the robot arm may hold a substrate at different angles depending on the position and movement of the robot arm. The data may be corrected for this effect. In some embodiments, the robot arm may hold a substrate at different heights depending on the position and movement of the robot arm. The data may be corrected for this effect.

[0148] At block 712, processing logic processes the first data to obtain second data. The second data includes an indication of the performance (e.g., quality of performance) of the substrate processing system. The second data may include one or more indications of a characteristic of the substrate. The second data may include one or more characteristics of the substrate, such as thickness, chemical composition, optical properties, surface pattern, etc. The second data may include optical data, such as an absorption spectrum, a reflectance spectrum, or the like. The second data may include substrate geometry, such as etch depth, sidewall angle, critical dimension, etc. The second data may include substrate shape, substrate distortion, bow, etc. The second data may include one or more indications of the state of the substrate processing system or the quality of performance of the substrate processing system. The second data may include one or more indications of the conditions under which the substrate processing system processed the substrate. The second data may include potential problems with the processing system, such as unexpected processing conditions, incorrect or insufficient calibration, or the like.

[0149] In some embodiments, processing the first data may include providing the first data to a model. The first data may be provided to a machine learning model, a physically based model, a statistical model, or other similar model. The model may perform one or more of many different functions. The model may receive data generated by one or more sensors and generate data indicative of substrate properties (e.g., thickness, chemistry, temperature, etc.). The model may receive data indicative of properties associated with one or more spatial regions of the substrate (e.g., regions arranged linearly across the substrate based on radiation collected as the substrate moves past the field of view of the probe) and generate a map of the substrate properties, e.g., a map of the substrate properties including one or more regions where radiation was not collected. The model may receive data indicative of the one or more substrate properties and generate one or more indications of processing conditions for the substrate. The model may receive data indicative of the one or more substrate properties and generate and / or implement recommended corrective actions. The model may receive data related to radiation received from the substrate (e.g., spectral data) and generate one or more indications of processing conditions for the substrate. The model may receive data related to emissions received from the substrate and generate and / or enact one or more recommended corrective actions. The model may receive data and generate an indication of a property of a single layer of the substrate (e.g., emissions from the substrate may be collected before and after processing operations are performed, e.g., by a probe positioned near a coupling between a processing chamber and a transfer chamber of a substrate processing system). In some embodiments, multiple models may be utilized. In some embodiments, models may be utilized sequentially (e.g., spectral data may be provided to a model to generate property data, and the property data may be provided to a model to generate a corrective action recommendation). In some embodiments, an ensemble model (e.g., a model that performs operations of multiple models) may be utilized.

[0150] A model (e.g., a machine learning model) may be trained by providing historical data. In some embodiments, this historical data may relate to substrates generated to train the model. For example, many substrates may be generated using processing conditions that are specifically and / or purposefully varied to build data for the input and / or output spaces of the model.

[0151] At block 714, processing logic may cause corrective action associated with the substrate processing system to be performed in consideration of the second data. The corrective action may include, for example, updating a process recipe, scheduling maintenance, issuing an alert to a user, etc.

[0152] FIG. 7C is a flowchart of a method 700C for performing corrective action taking into account one or more characteristics of a substrate, according to some embodiments. At block 720, processing logic receives first data. The first data may share one or more characteristics with the first data of FIG. 7B. The first data is generated by a sensor of a substrate processing system. The sensor may be a spectrometer. The first data is generated in response to the sensor receiving electromagnetic radiation from the substrate. The substrate is held by (e.g., supported by, supported on) a robot arm. The robot arm is within (e.g., coupled to a transfer chamber of) the substrate processing system. The radiation may be received by the sensor while the robot arm is moving the substrate, e.g., measurements may be taken while the substrate is being transferred. The measurements may be taken while the substrate is being transferred from one chamber (e.g., a transfer chamber, a processing chamber, a load lock, etc.) to another chamber. The sensor may include a probe (e.g., for receiving the radiation). The probe may be at least partially located within the processing system (e.g., embedded in a wall of the system). The processing system may include one or more probes, one or more arrays of probes, one or more sensors, and the like.

[0153] At block 722, processing logic provides the first data to a model. The model may be a physics-based model, a machine learning model, etc. The model may be configured to receive optical data, radiative data, spectral data, etc. as input and generate output based on the input data.

[0154] At block 724, processing logic obtains second data from the model. The second data indicates one or more properties of the substrate. The properties may include one or more of physical properties, optical properties, electrical properties, current conditions (e.g., temperature), etc. The properties may include a selected layer or group of layers of the substrate. The properties may include thickness, refractive index, extinction coefficient, temperature, etc. Probes and / or sensors may be used to determine the target properties, for example, a pyrometer probe may be utilized to determine temperature, or a broadband reflectometer (e.g., sensitive to white light) may be utilized to determine film thickness. In some embodiments, the model may receive historical measurements (e.g., measurements from before one or more processing operations) as additional inputs (e.g., to isolate properties of deposited or etched layers or substrates).

[0155] At block 726, processing logic causes corrective action to be performed. The corrective action may include process / feedback control (e.g., updating one or more process recipes to account for the characteristics determined by the model). The corrective action may include scheduling metrology (e.g., of substrates potentially outside target specifications). The corrective action may include scheduling maintenance (e.g., corrective or preventive maintenance). The corrective action may include initiating maintenance (e.g., a chamber clean or chamber seasoning procedure).

[0156] 8 shows a block diagram of an exemplary computing device that operates in accordance with one or more aspects of the present disclosure. In various illustrative examples, various components of computing device 800 may represent various components of client device 450, measurement system 410, server 420, data store 140, and modeling system 470 shown in FIG.

[0157] The exemplary computing device 800 may be connected to other computer devices on a LAN, an intranet, an extranet, and / or the Internet. The computing device 800 may operate in the capacity of a server in a client-server network environment. The computing device 800 may be a personal computer (PC), a set-top box (STB), a server, a network router, switch, or bridge, or any device capable of executing (sequentially or otherwise) a set of instructions that specify actions to be taken by the device. Furthermore, while only an exemplary single computing device is shown, the term "computer" should be taken to include any collection of computers that individually or jointly execute a set (or sets) of instructions to perform any one or more of the methodologies discussed herein.

[0158] The exemplary computing device 800 may include a processing unit 802 (also referred to as a processor or CPU), a main memory 804 (e.g., read-only memory (ROM), flash memory, dynamic random access memory (DRAM), such as synchronous DRAM (SDRAM)), a static memory 806 (e.g., flash memory, static random access memory (SRAM)), and a secondary memory (e.g., a data storage device 818), which may communicate with each other via a bus 830.

[0159] Processing unit 802 represents one or more general-purpose processing devices, such as a microprocessor, a central processing unit, or the like. More specifically, processing unit 802 may be a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a processor implementing other instruction sets, or a processor implementing a combination of instruction sets. Processing unit 802 may also be one or more special-purpose processing devices, such as an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), a network processor, or the like. According to one or more aspects of the present disclosure, processing unit 802 may be configured to execute instructions to implement methods 700A-C illustrated in FIGS. 7A-C and / or the process illustrated in FIG. 6.

[0160] The exemplary computing device 800 may further include a network interface device 808, which may be communicatively coupled to a network 820. The exemplary computing device 800 may further include a video display 810 (e.g., a liquid crystal display (LCD), touch screen, or cathode ray tube (CRT)), an alphanumeric input device 812 (e.g., a keyboard), a cursor control device 814 (e.g., a mouse), and an audio signal generating device 816 (e.g., a speaker).

[0161] The data storage device 818 may include a computer-readable storage medium (or more specifically, a non-transitory machine-readable storage medium) 828 having stored thereon one or more sets of executable instructions 822. According to one or more aspects of the present disclosure, the executable instructions 822 may include executable instructions related to any of the methods disclosed herein, e.g., instructions for performing any of the methods disclosed herein. The executable instructions 822 may be related to performing the method illustrated in FIG. 6 and / or methods 700A-C illustrated in FIGS. 7A-C.

[0162] The executable instructions 822 may also reside, completely or at least partially, within the main memory 804 and / or within the processing unit 802 during execution of the executable instructions 822 by the exemplary computing device 800, with the main memory 804 and processing unit 802 also constituting computer-readable storage media. The executable instructions 822 may also be transmitted or received across a network via the network interface device 808.

[0163] 9 illustrates computer-readable storage medium 828 as a single medium, the term "computer-readable storage medium" should be interpreted to include a single medium or multiple media (e.g., centralized or distributed databases and / or associated caches and servers) that store one or more sets of operating instructions. The term "computer-readable storage medium" should also be interpreted to include any medium that can store or encode a set of instructions for execution by a machine, causing the machine to perform any one or more of the methodologies described herein. Accordingly, the term "computer-readable storage medium" should be interpreted to include, but is not limited to, solid-state memory and optical and magnetic media.

[0164] Some portions of the above detailed descriptions are presented in terms of algorithms and symbolic representations of operations on data bits within a computer memory. These algorithmic descriptions and representations are the means used by those skilled in the data processing arts to most effectively convey the substance of their work to others skilled in the art. An algorithm is here, and generally, conceived to be a self-consistent sequence of steps leading to a desired result. These steps require physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of electrical or magnetic signals capable of being stored, transferred, combined, compared, and otherwise manipulated. It is convenient at times, principally for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, or the like.

[0165] It should be borne in mind, however, that all of these and similar terms are associated with the appropriate physical quantities and are merely convenient labels applied to those quantities. Unless otherwise indicated, as will be apparent from the discussion that follows, throughout this description, discussions utilizing terms such as "identify," "determine," "store," "adjust," "produce," "return," "compare," "produce," "stop," "load," "copy," "cast," "exchange," "execute," "receive," "process," "generate," "trigger," "train," or other similar terms will be understood to refer to operations and processes of a computer system or similar electronic computing device that manipulate data in the computer system's registers and memory, which are represented as physical (electronic) quantities, and transform such data into other data, also represented as physical quantities, in the computer system's memory or registers or other such information storage, transmission, or display devices.

[0166] Examples of the present disclosure further relate to apparatus for performing the methods described herein. This apparatus may be specially constructed for the required purposes, or it may be a general-purpose computer system selectively programmed by a computer program stored in the computer system. Such computer program may be stored on a computer-readable storage medium, such as, but not limited to, any type of disk, including optical disks, compact disk read-only memories (CD-ROMs) and magneto-optical disks, read-only memories (ROMs), random access memories (RAMs), erasable programmable read-only memories (EPROMs), electrically erasable programmable read-only memories (EEPROMs), magnetic disk storage media, optical storage media, flash memory devices, other types of machine-accessible storage media, or any type of media suitable for storing electronic instructions, each coupled to a computer system bus.

[0167] The foregoing description sets forth numerous specific details, such as examples of particular systems, components, methods, and the like, to provide a thorough understanding of some embodiments of the present disclosure. However, it will be apparent to one skilled in the art that at least some embodiments of the present disclosure can be practiced without these specific details. In other instances, well-known components or methods have not been described in detail or have been shown in simple block diagram form to avoid unnecessarily obscuring the present disclosure. Thus, the specific details described are merely examples. It is contemplated that particular implementations may vary from these illustrative details and still be within the scope of the present disclosure.

[0168] Throughout this specification, a reference to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described with respect to that embodiment is included in at least one embodiment. Thus, appearances of the phrase "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Additionally, the term "or" is intended to mean an inclusive "or" rather than an exclusive "or." When the term "about" or "approximately" is used herein, this term is intended to mean that the stated nominal value is accurate to within + / - 10%.

[0169] Although the operations of the methods herein are shown and described in a particular order, the order of the operations of each method may be changed so that certain operations are performed in reverse order, or so that certain operations are performed, at least in part, concurrently with other operations. In alternative embodiments, instructions or sub-operations of different operations may be performed intermittently and / or alternately.

[0170] It is understood that the foregoing description is intended to be illustrative, and not limiting. Many other embodiments will be apparent to those skilled in the art upon reading and understanding the foregoing description. The scope of the present disclosure should, therefore, be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled.

Claims

1. receiving, by a processing device, first data generated by a first sensor of a substrate processing system, the first data being generated in response to the first sensor receiving electromagnetic radiation from a substrate held by a robot arm of a transfer chamber in the substrate processing system; processing the first data to obtain second data, the second data including a first indication of performance of the substrate processing system; and performing corrective action associated with the substrate processing system in consideration of the second data. A method comprising:

2. The method of claim 1 , wherein the first data comprises spectral data and the second data comprises an indication of a thickness of the substrate.

3. The method of claim 1 , wherein the second data comprises an indication of conditions under which the substrate was processed by the substrate processing system.

4. 10. The method of claim 1, wherein the first sensor receives the electromagnetic radiation from the substrate while the robot arm is transferring the substrate from a first region of the substrate processing system to a second region of the substrate processing system.

5. receiving third data from the first sensor, the first data relating to a first portion of the substrate and the third data relating to a second portion of the substrate; processing the third data to obtain fourth data, the fourth data including a second indication of performance of the substrate processing system; and generating a profile of the substrate taking into account the third data and the fourth data; The method of claim 1 further comprising:

6. receiving a signal from an object detection device, the signal indicating the substrate is present at a first location of the substrate processing system; and triggering operation of the first sensor in response to receiving the signal from the object detection device, wherein triggering operation of the first sensor includes causing the first sensor to detect electromagnetic emissions from the substrate at a target time, the target time being based on (i) a time when the object detection device indicates that the substrate is present at the first location, and (ii) a speed at which the substrate is being transported by the robot arm. The method of claim 1 further comprising:

7. receiving third data generated by one or more additional sensors of the substrate processing system, the first sensor and the one or more additional sensors forming an array of sensors, the third data generated in response to the one or more additional sensors receiving electromagnetic radiation from one or more regions of the substrate different from the region of the substrate from which the first sensor received electromagnetic radiation; and processing the third data to obtain fourth data, the fourth data including a second indication of performance of the substrate processing system. The method of claim 1 further comprising:

8. receiving third data generated by a second sensor of the substrate processing system; and processing the third data to obtain fourth data, the fourth data including a second indication of performance of the substrate processing system, the first sensor receiving electromagnetic radiation in a first wavelength range, and the second sensor receiving electromagnetic radiation in a second wavelength range. The method of claim 1 further comprising:

9. The method of claim 8 , wherein the fourth data includes temperature data of the substrate, and the first sensor is configured to receive radiation reflected from the substrate.

10. receiving a plurality of sensor data associated with a first plurality of substrates after the first plurality of substrates have been processed by an etching process or a deposition process; receiving a plurality of data indicative of processing conditions for the first plurality of substrates; and training the machine learning model by providing the plurality of sensor data as training inputs to the machine learning model and providing the plurality of data indicating processing conditions as target outputs to the machine learning model; The method of claim 1 further comprising:

11. 10. The method of claim 1, wherein processing the first data to obtain second data comprises providing the first data to a trained machine learning model, the trained machine learning model configured to receive sensor data as input and generate an indication of a processing condition for the substrate as output.

12. a process chamber; a transfer chamber coupled to the process chamber; a robot arm of the transfer chamber configured to transfer substrates between the process chamber and the transfer chamber; a sensor having a first field of view, at least a portion of the first field of view including at least a portion of the substrate when the substrate is transferred into or out of the process chamber by the robot arm, the sensor configured to receive electromagnetic radiation that has interacted with the substrate; a processing device for processing data generated in response to the sensor receiving electromagnetic radiation to determine at least one of: (i) one or more indications of performance of the substrate processing system; or (ii) one or more characteristics of the substrate; A substrate processing system comprising:

13. 13. The substrate processing system of claim 12, further comprising an electromagnetic radiation source, wherein the sensor comprises an electromagnetic probe, the electromagnetic probe positioned to receive electromagnetic radiation generated by the source and that has interacted with the substrate.

14. The substrate processing system of claim 13 , wherein the electromagnetic radiation that interacts with the substrate comprises electromagnetic radiation that is reflected from a surface of the substrate.

15. The substrate processing system of claim 13 , wherein the electromagnetic radiation that interacts with the substrate comprises electromagnetic radiation that is transmitted through the substrate.

16. The substrate processing system of claim 12 , wherein the sensor receives infrared radiation emitted by the substrate.

17. 13. The substrate processing system of claim 12, further comprising one or more additional sensors, wherein the sensor and the one or more additional sensors form an array of sensors, a second sensor in the array of sensors having a second field of view, the second field of view including a second portion of the substrate, the second portion of the substrate being different from the at least one portion of the substrate included in the first field of view.

18. The processing device first data related to electromagnetic radiation received by the sensor at a first time from a first region of the substrate; second data related to electromagnetic radiation received by the sensor at a second time from a second region of the substrate; and 13. The substrate processing system of claim 12, wherein the substrate processing system is configured to process data comprising:

19. 13. The substrate processing system of claim 12, further comprising a pyrometer, wherein the sensor is adapted to receive electromagnetic radiation reflected from the substrate and the pyrometer is adapted to receive electromagnetic radiation emitted from the substrate.

20. A non-transitory machine-readable storage medium having stored thereon instructions that, when executed, cause a processing device to perform operations, the operations including: receiving first data generated by a first sensor of a substrate processing system, the first data being generated in response to the first sensor receiving electromagnetic radiation from a substrate held by a robot arm of a transfer chamber in the substrate processing system; processing the first data to obtain second data, the second data including values ​​of a first one or more characteristics of the substrate; and performing corrective action associated with the substrate processing system in consideration of the second data.

1. A non-transitory machine-readable storage medium comprising:

21. The operation is receiving third data generated by the first sensor of the substrate processing system; processing the third data to obtain a second one or more property values ​​of the substrate; receiving fourth data generated by the first sensor of the substrate processing system; and processing the fourth data to obtain a third one or more property values ​​of the substrate; further comprising the first data, the second data, and the third data are each generated in response to the first sensor receiving electromagnetic radiation from the substrate while the substrate is being transferred by the robot arm; the first data corresponds to a first spatial region of the substrate, the third data corresponds to a second spatial region of the substrate, and the fourth data corresponds to a third spatial region of the substrate; the first spatial region is different from the second spatial region, and the third spatial region is different from the first and second spatial regions; 21. The non-transitory machine-readable storage medium of claim 20.

22. the robot arm is configured to transfer the substrate according to a first trajectory and a second trajectory; the first trajectory causing a first portion of the substrate to intersect a field of view of the first sensor; the second trajectory causes a second portion of the substrate to intersect the field of view of the first sensor; the first portion includes the first spatial region; the second portion includes the second spatial region; 22. The non-transitory machine-readable storage medium of claim 21.

23. processing the first data to obtain second data; the angle of the surface of the substrate relative to the optical axis of the first sensor; or a distance of the surface of the substrate from one or more components of the first sensor; 21. The non-transitory machine-readable storage medium of claim 20, wherein the non-transitory machine-readable storage medium is implemented with consideration of at least one of:

24. The operation is receiving third data generated by a second sensor of the substrate processing system, the third data being generated in response to the second sensor receiving electromagnetic radiation from the substrate; and processing the third data to obtain fourth data, the fourth data including an indication of a temperature of the substrate, the first sensor receiving electromagnetic radiation in a first wavelength range, and the second sensor receiving electromagnetic radiation in a second wavelength range.

21. The non-transitory machine-readable storage medium of claim 20, further comprising:

Images