Generation method, substrate processing method, substrate processing apparatus, information processing apparatus, article manufacturing method and program

Abstract

A method for generating a prediction model for predicting the processing result of a substrate processed by a substrate processing device comprises the step of generating a prediction model based on device data representing the state of the substrate processing device when processing up to the Nth substrate (N is a natural number) among the plurality of substrates, which is obtained by processing a plurality of substrates continuously under the same processing conditions by the substrate processing device, and measurement data obtained by measuring the processing result of the (N+1)th substrate among the plurality of substrates.

Description

Technology Field

[0001] The present disclosure relates to a method of generation, a method of processing a substrate, a substrate processing apparatus, an information processing apparatus, a method of manufacturing an article, and a program. Background Technology

[0002] Recently, requirements for various processes in the manufacturing process of semiconductor device articles have become strict. For example, in photolithography equipment, requirements for overlay are becoming strict, so a technology for reducing overlay error is proposed in Japanese Patent Publication No. 2023-158946. Japanese Patent Publication No. 2023-158946 discloses a technology that predicts substrate distortion using physical simulation from data regarding substrate holding by a substrate holding part, and calculates a control value to reduce overlay error from such distortion.

[0003] However, in conventional technology, distortion caused by substrate retention is dominant as a factor in overlay error (substrate processing result), but in order to further reduce overlay error, it is necessary to consider other factors.

[0004] For example, an exposure device has a drive unit that moves a substrate. Since this drive unit operates using electromagnetic force as a power source, it generates heat in response to the application of current, causing its operating state (such as resistance) to change. Although measures such as cooling the drive unit are taken to suppress these changes in its operating state, it is difficult to maintain a constant operating state. Consequently, deviations occur in the actual operation relative to the control value of the drive unit due to changes in its operating state. Furthermore, since the driving state of the drive unit changes internally over time, it is difficult to measure it directly.

[0005] At this time, it is difficult to ascertain the entire state of the exposure device, including not only the state of the drive unit but also the fluctuations of the air inside the device. Even if the entire state is ascertained, it is difficult to determine the extent to which they affect the overlay. The problem to be solved

[0006] The present disclosure provides a technique advantageous for generating a prediction model to predict the processing results of a substrate. means of solving the problem

[0007] According to one aspect of the present disclosure, a method for generating a prediction model for predicting the processing result of a substrate processed by a substrate processing device is provided, the method comprising the step of generating the prediction model based on device data representing the state of the substrate processing device when processing up to the Nth substrate (N is a natural number) among the plurality of substrates, obtained by processing a plurality of substrates continuously under the same processing conditions by the substrate processing device, and measurement data obtained by measuring the processing result of the (N+1)th substrate among the plurality of substrates.

[0008] Another feature of the present invention will become apparent from the description of the following embodiments with reference to the accompanying drawings. Brief explanation of the drawing

[0009] FIG. 1 is a schematic diagram illustrating the configuration of an exposure apparatus according to one aspect of the present disclosure. Figure 2 is a diagram illustrating the overlay error. Figure 3 is a diagram illustrating focus error. FIG. 4 is a diagram illustrating the learning of a prediction model and prediction using the prediction model according to the present embodiment. FIG. 5 is a diagram illustrating the learning of a prediction model and prediction using the prediction model according to the present embodiment. FIG. 6 is a diagram illustrating the learning of a prediction model and prediction using the prediction model according to the present embodiment. FIGS. 7A and FIGS. 7B are drawings illustrating an example of a two-dimensional map image of an overlay. Specific details for implementing the invention

[0010] Hereinafter, embodiments are described in detail with reference to the attached drawings. At this time, the following embodiments do not limit the invention with respect to the claims. Although multiple features are described in the embodiments, not all of these multiple features are essential to the invention, and multiple features may be combined at will. Furthermore, in the attached drawings, the same reference number is assigned to identical or similar components, and redundant descriptions are omitted.

[0011] FIG. 1 is a schematic diagram illustrating the configuration of an exposure apparatus (900) according to one aspect of the present disclosure. The exposure apparatus (900) is a lithography apparatus used in a lithography step for manufacturing an article including a device represented by a semiconductor element. The exposure apparatus (900) is a substrate processing apparatus that forms a pattern on a substrate by exposing a substrate (wafer or plate) through a plate. The exposure apparatus (900) projects the pattern of the plate onto the substrate through a projection optical system, thereby transferring the pattern of the plate to the substrate.

[0012] In this embodiment, an exposure device (900) is described as an example of a substrate processing device, but the substrate processing device broadly includes other processing devices such as an imprint device and a drawing device. The imprint device includes a device that forms an imprint material on a substrate using a mold to form a pattern of the imprint material on the substrate. At this time, the imprint device also includes a flattening device that flattens the composition on the substrate using a mold having a flat surface. The drawing device includes a device that draws a pattern on the substrate using a charged particle beam (electron beam, ion beam, etc.).

[0013] In this specification and the accompanying drawings, directions are indicated in an XYZ coordinate system in which a direction perpendicular to the surface of a substrate (vertical direction) is defined as the Z-axis, and two directions parallel to the plane perpendicular to the Z-axis and perpendicular to each other are defined as the X-axis and the Y-axis. Additionally, directions parallel to the X-axis, Y-axis, and Z-axis in the XYZ coordinate system are defined as the X-direction, Y-direction, and Z-direction, respectively.

[0014] In this embodiment, the exposure device (900) is a step-and-scan type exposure device (scanner) that exposes the substrate (302) while synchronously scanning the original plate stage (902) (original plate (406)) and the substrate stage (301) (substrate (302)). However, the exposure device (900) may also be a step-and-repeat type exposure device (stepper) that exposes the substrate (302) while the original plate stage (902) and the substrate stage (301) are stopped.

[0015] As shown in FIG. 1, the exposure device (900) has a light source unit (907), an illumination optical system (908), a disc stage (902), a projection optical system (404), a substrate stage (301), a substrate chuck (407), a control unit (916), and an output unit (917). Additionally, the exposure device (900) has an interferometer system (909) on the disc side, an interferometer system (307) on the substrate side, a focus measurement unit (405), a substrate transport unit (912), a disc transport unit (914), and an alignment scope (915).

[0016] The light source unit (907) includes, for example, a high-pressure mercury lamp, an ArF excimer laser, a KrF excimer laser, an EUV light source, etc. The light source unit (907) may be placed outside a chamber that accommodates components other than the light source unit (907) of the exposure device (900), or may be accommodated in one chamber together with components other than the light source unit (907) of the exposure device (900).

[0017] The illumination optical system (908) illuminates the disc (406) with light from the light source unit (907). The disc (406) is referred to as a reticle or mask. The disc (406) has a pattern to be transferred to a substrate (302) on which photoresist is placed, and is held by a disc stage (902). The pattern of the disc (406) typically includes a plurality of features (e.g., lines, holes, etc.). The disc stage (902) holds the disc (406) via a disc chuck and is driven by a disc driving mechanism, such as a linear motor.

[0018] The projection optical system (404) projects the pattern of the original plate (406) onto the substrate (302). Accordingly, the pattern of the original plate (406) is transferred to the photoresist placed on the substrate (302). The projection optical system (404), for example, projects the pattern of the original plate (406) onto the substrate (302) by reducing it according to the projection magnification ratio (for example, 1 / 4). The pattern of the original plate (406) is projected sequentially onto a plurality of shot areas of the substrate (302), and the plurality of shot areas are sequentially exposed.

[0019] The substrate stage (301) is driven by a substrate driving mechanism including a linear motor, etc., and is movable in the X direction and Y direction. The substrate chuck (407) is placed on the substrate stage (301) and holds the substrate (302). The substrate stage (301) may be configured to position the substrate chuck (407) with respect to the Z direction, θz direction, θx direction, and θy direction. The substrate (302) held by the substrate chuck (407) is positioned via the substrate stage (301) and the substrate chuck (407).

[0020] The interferometer system 909 includes a plurality of laser interferometers configured to measure the X-direction and Y-direction position of the disk stage (902) and the orientation (θx, θy, θz) of the disk stage (902). The interferometer system 307 includes a plurality of laser interferometers configured to measure the X-direction and Y-direction position of the substrate stage (301) holding the substrate (302) and the orientation (θx, θy, θz) of the substrate stage (301). The position and orientation of the disk stage (902) and the substrate stage (301) are controlled by a control unit (916) based on the position and orientation measured by the interferometer systems 909 and 307.

[0021] The focus measurement unit (405) includes a light-projecting unit (405a) that projects a plurality of beams onto a substrate (302) at an oblique incidence, and a light-receiving unit (405b) that receives a plurality of beams reflected from the substrate (302). The light-receiving unit (405b) includes a detection unit that detects a plurality of beams incident on the light-receiving unit (405b) and provides a signal corresponding to them to the control unit (916). The light-projecting unit (405a) and the light-receiving unit (405b) are arranged across the optical axis of the projection optical system (404). Based on the signal obtained from the focus measurement unit (405), the control unit (916) determines the position of the substrate (302) in the Z direction and controls the movement of the substrate (302) via the substrate stage (301).

[0022] The substrate transport unit (912) is a mechanism configured to transport a substrate (302). The substrate transport unit (912) has, for example, a function of transporting a substrate (302) from a substrate receiving container that receives the substrate (302) to a substrate stage (301), and a function of transporting a substrate (302) from the substrate stage (301) to a substrate receiving container, etc.

[0023] The disc return unit (914) is a mechanism for returning the disc (406). The disc return unit (914) has, for example, a function of returning the disc (406) from a disc receiving container that receives the disc (406) to a disc stage (902), and a function of returning the disc (406) from the disc stage (902) to a disc receiving container, etc.

[0024] The alignment scope (915) acquires a digital image signal by capturing a mark (alignment mark) installed on the substrate (302) in order to align the substrate (302) held by the substrate chuck (407). The alignment scope (915) includes an image sensor that outputs an image signal according to the light intensity distribution formed by reflected light from the substrate (302), and an A / D converter that converts the image signal output from the image sensor into a digital image signal. Based on the digital image signal acquired by the alignment scope (915), the control unit (916) determines the position of the mark on the substrate (302) and aligns the substrate (302) through the substrate stage (301).

[0025] The output unit (917) includes, for example, a display device such as a touch panel and a voice output device such as a speaker, and is configured to display various user interfaces (screens) and output various voices. The output unit (917) may be configured integrally with the exposure device (900) (in a common housing) or separately from the exposure device (900) (in a different housing).

[0026] The control unit (916) controls the exposure process of exposing the substrate (302) by comprehensively controlling each component of the exposure device (900). The control unit (916) is, for example, a programmable logic device (PLD) such as a field programmable gate array (FPGA), an application-specific integrated circuit (ASIC), a computer with a program installed, or an information processing device composed of all or part of these. The control unit (916) may be composed of processors such as multiple CPUs. Additionally, the control unit (916) may be placed inside the housing together with other components of the exposure device (900), or placed outside the housing separately from other components of the exposure device (900).

[0027] The control unit (916) is configured to control and operate the exposure device (900) according to control conditions obtained from a memory unit (not shown) to perform an exposure process (lithography process) that exposes (processes) the substrate (302) (i.e., functions as a processing unit). The control conditions applied to the exposure process include, for example, a transport condition that defines the transport speed, transport path, etc., when the substrate (302) is transported to the substrate stage (301) by the substrate transport unit (912). The control conditions applied to the exposure process include, for example, a positioning condition that defines the allowable error, etc., when the substrate (302) is positioned by the substrate stage (301). The control conditions applied to the exposure process include, for example, a measurement condition that defines the irradiation time and irradiation timing of a plurality of beams when the focus is measured by the focus measurement unit (405). The control conditions applied to the exposure process include, for example, an identification of the original plate (406) when exposing the substrate (302), a layout of a plurality of shot areas of the substrate (302), an exposure condition defining an illumination mode, etc.

[0028] Below, a more specific example is described regarding the case where the target to be evaluated as a result of processing by the exposure device (900) is an overlay error and a focus error.

[0029] Referring to FIG. 2, an overlay error is described. A substrate (302) is placed on a substrate stage (301), and for a target location (303) where a feature of the substrate (302) is to be transferred (formed), the actual transfer location (304) of the transferred feature includes an error (amount of misalignment) ΔX (306) and ΔY (305). If the target location (303) is a target location based on a feature of the lower layer, the error ΔX (306) and ΔY (305) are understood as alignment errors. If the target location (303) is a target location not based on a feature of the lower layer, the error ΔX (306) and ΔY (305) are understood as placement errors of the feature transferred to the substrate (302). The error ΔX (306) is an error in the X direction, and the error ΔY (305) is an error in the Y direction. If the overlay error exceeds the allowable range (or threshold), there is a possibility of a connection failure between the features of the layer (exposed layer) on which the pattern is transferred by the exposure device (900) and the features of the layer below.

[0030] The position of the substrate stage (301) is measured using the interferometer 307x and interferometer 307y of the interferometer system 307. Meanwhile, it is difficult to measure the transfer position (304) in the exposure device (900). This is because there are various error factors, such as the precision of the X-axis bar mirror (308) and Y-axis bar mirror (310) installed on the substrate stage (301), the rotation error of the substrate stage (301), and the change in the wavelength of light used in the interferometer system 307 caused by changes in ambient temperature.

[0031] With reference to FIG. 3, the focus error is explained. FIG. 3 schematically illustrates the components related to the measurement of focus in an exposure device (900). The distance between the upper surface of the projection optical system (404) and the surface (upper surface) of the substrate (302) corresponds to the focus error ΔZ (403). In the exposure device (900), the focus is controlled so that the focus error ΔZ (403) falls within the depth of focus. The focus error ΔZ (403) is evaluated by observing the substrate (302) using an SEM or the like after exposing the substrate (302).

[0032] The device data affecting the focus error ΔZ (403) includes at least one of the following various data. Examples of device data include data indicating measurement conditions in the focus measurement unit (405) and data indicating control deviation of the substrate stage (301). Other examples of device data include data indicating the environment in which the exposure device (900) is placed during focus measurement (e.g., data such as temperature, humidity, and atmospheric pressure) and data indicating aberrations of the projection optical system (404) during focus measurement.

[0033] In addition to changes in the environment in which the exposure device (900) is placed, it is difficult to measure the focus error ΔZ (403) due to the presence of various error factors.

[0034] These error factors include complex factors, but regarding each factor, when exposure is performed using the same pattern and the same layout (same processing conditions), such as with the exposure device (900), a general trend can be identified. For example, the amount of heat supplied to the driving mechanism of each stage of the exposure device (900) changes according to the time-dependent change in the torque required to drive the stage, that is, the current. This amount of heat can be considered to be nearly the same if the pattern and layout are identical. However, since the initial state of each factor changes and each factor interacts in a complex manner, direct addition is insufficient, and it is difficult to classify each factor.

[0035] Therefore, in the prior art, instead of solving each factor, a prediction model used to predict the result of the exposure treatment (the result of the treatment of the substrate (302)) is generated by machine learning. More specifically, the prediction model is generated based on device data recorded (stored) as a log in the exposure device (900) (the memory unit of the control unit (916)) and measurement data obtained by measuring the result of the exposure treatment corresponding thereto. Using the prediction model, the result of the exposure treatment corresponding thereto is predicted from other device data. However, since this is a prediction after the exposure treatment, it is not possible to improve the result of the exposure treatment for the current exposure treatment.

[0036] The present embodiment provides a technique advantageous for generating a prediction model to predict the result of the exposure treatment of a substrate (302) when the exposure device (900) continuously exposes (processes) a plurality of substrates (302) under the same processing conditions. In the present embodiment, regarding measurement data representing the result of the exposure treatment of a certain substrate, a prediction model is generated by learning, through machine learning, the relationship with the device data when a substrate prior to that substrate was exposed. Then, by inputting the device data when another substrate was exposed into the prediction model, the result of the exposure treatment of the next substrate is predicted. Furthermore, by using the predicted result of the exposure treatment to control the exposure treatment of the next substrate by the exposure device (900), the result of the exposure treatment for the current exposure treatment can be improved.

[0037] In this embodiment, a prediction model is generated based on device data indicating the state of the exposure device (900) when processing up to the Nth substrate among the plurality of substrates (302), and measurement data obtained by measuring the result of the exposure treatment of the (N+1)th substrate among the plurality of substrates (302). Next, using this prediction model, the processing result of the substrate (302) to be exposed by the exposure device (900) is predicted, and the predicted processing result is obtained. Then, based on this predicted processing result, the control conditions of the exposure device (900) when processing the substrate (302) by the exposure device (900) are determined, and the exposure treatment of the substrate (302) is performed while controlling the exposure device (900) according to the control conditions. At this time, N is a natural number.

[0038] In addition, in this embodiment, the control unit (916) performs the steps of generating a prediction model and acquiring a prediction processing result; that is, it is assumed that the control unit (916) functions as a generating unit configured to generate a prediction model and an acquiring unit configured to acquire a prediction processing result. However, the steps of generating a prediction model and acquiring a prediction processing result may be performed by an external information processing device having a generating unit and an acquiring unit.

[0039] Hereinafter, a specific example of improving the overlay, that is, reducing the overlay error, as a result of the exposure treatment of the substrate (302) will be described. With reference to FIG. 4, in this embodiment, the learning (creation) of a prediction model for predicting the overlay error and the prediction of the overlay error using the prediction model will be explained separately.

[0040] First, the training of the prediction model is described. As a target variable, measurement data obtained by measuring the result of the exposure treatment of the substrate (302) is considered. The measurement data includes, for example, data regarding at least one of overlay, focus, line width uniformity (CD uniformity), and edge placement error, and here, as described above, overlay is used. In addition, as an explanatory variable, device data indicating the state of the exposure device (900) when the substrate (302) is exposed is considered, that is, data regarding the device state that changes over time when the exposure device (900) performs exposure treatment of a plurality of substrates (302). The device data includes, for example, data regarding at least one of the position and deviation of the substrate stage (301), the position and deviation of the plate stage (902), alignment results, and the environment (temperature, humidity, atmospheric pressure, etc.) in which the exposure device (900) is installed.

[0041] As a combination of data sets, for example, a combination of measurement data of the third substrate SB3, i.e., an overlay, and device data of the second substrate SB2 or the first substrate SB1 is considered. Similarly, a combination of measurement data of the fourth substrate SB4, i.e., an overlay, and device data of the third substrate SB3 or the second substrate SB2 is considered. Based on these combinations of data sets, the relationship between device data and measurement data is regressively learned using machine learning observational learning to generate a prediction model (regression model). At this time, the prediction model may be a polynomial function, or an SVM, GBDT, CNN, or RNN.

[0042] In this way, during the training of the prediction model, a prediction model is generated for device data when processing up to the Nth substrate among a plurality of substrates, to predict the overlay resulting from the exposure treatment of the (N+1)th substrate.

[0043] Next, a prediction using a prediction model is described. A prediction using a prediction model is performed, for example, when the photolithography process for the N'th substrate included in a different lot is completed. In this embodiment, a prediction model is used to predict (calculate) the result of photolithography of the (N'+1)th substrate from the device data when the N'th substrate is photolithographed. More specifically, a prediction model is used to predict the overlay resulting from the photolithography of the 3rd substrate SB3' from the device data when the 2nd substrate SB2' or the 1st substrate SB1' is photolithographed. Likewise, a prediction model is used to predict the overlay resulting from the photolithography of the 4th substrate SB4' from the device data when the 3rd substrate SB3' or the 2nd substrate SB2' is photolithographed. In this way, the result of exposing the (N'+1)th substrate is predicted, and the predicted exposure result (predicted processing result) is obtained. When exposing the (N'+1)th substrate, the exposure device (900) is fedforward controlled based on the predicted exposure result to improve the result of exposing the (N'+1)th substrate. That is, the overlay error is reduced. At this time, N' is a natural number.

[0044] In this embodiment, the overlay of the (N'+1)th substrate is predicted, but it is also possible to predict the overlay of substrates after the (N'+1)th substrate. If processing of device data requires time and the control of the exposure treatment of the (N'+1)th substrate is delayed, for example, as shown in FIG. 5, the overlay of the (N'+2)th substrate may be predicted using a prediction model. In this case, as a combination of the data set for training the prediction model, the combination of the overlay of the 3rd substrate SB3 and the device data when the 1st substrate SB1 is exposed is considered. Additionally, as a combination of the data set for training the prediction model, the combination of the overlay of the 4th substrate SB4 and the device data when the 2nd substrate SB2 is exposed is considered. In the prediction using the prediction model, the overlay of the third substrate SB3' is predicted from the device data when the first substrate SB1' is exposed, and the overlay of the fourth substrate SB4' is predicted from the device data when the second substrate SB2' is exposed.

[0045] In addition, in FIG. 4, the combination of the data set for training the prediction model is a combination of multiple device data and one measurement data. In other words, the prediction model is generated based on multiple device data when each of the multiple substrates up to the Nth substrate is photolithographically processed and the measurement data of the (N+1)th substrate, but the present disclosure is not limited to this. For example, as shown in FIG. 6, the combination of the data set for training the prediction model may be a combination of one device data and one measurement data, so that the device data and the measurement data correspond one-to-one. In other words, the prediction model may be generated based on the device data when the Nth substrate is photolithographically processed and the measurement data of the (N+1)th substrate. In this case, as the combination of the data set for training the prediction model, the combination of the overlay of the 2nd substrate SB2 and the device data when the 1st substrate SB1 is photolithographically processed is considered. In addition, as a combination of data sets for training the prediction model, a combination of the overlay of the third substrate SB3 and the device data when the second substrate SB2 is exposed is considered. When predicting using the prediction model, the overlay of the second substrate SB2' is predicted from the device data when the first substrate SB1' is exposed, and the overlay of the third substrate SB3' is predicted from the device data when the second substrate SB2' is exposed.

[0046] Additionally, device data indicating the state of the exposure device (900) may include data outside the exposure device, that is, data regarding the state of the device outside the control target in the exposure device (900). This device data may include, for example, data regarding the environment of the cleanroom where the exposure device (900) is installed. Data regarding the cleanroom environment is data such as the temperature, airflow, air pressure, and humidity of the air conditioning in the cleanroom. These data can be used as device data indicating the future state of the exposure device (900). Therefore, it is possible to consider data regarding the cleanroom environment as an explanatory variable (device data) for the learning of the prediction model. As information added to the explanatory variable to increase prediction accuracy, one specific example was described above. In this regard, for example, an extended idea can be taken that the data of the measurement data of substrate SB3 before the formation of the measurement target is influencing the measurement target of substrate SB3. Therefore, the explanatory variable may include data of the processing performed on substrate SB3 prior to the current exposure. For example, when multiple exposures are performed on substrate SB3, this data may include device data of the exposure device during the exposure performed prior to the current exposure, and device data of the coating device during the coating of the resist placed on substrate SB3 prior to the exposure.

[0047] In addition, in this embodiment, when the result of the exposure treatment of a substrate is predicted using a prediction model and the predicted exposure treatment result is obtained, the amount of correction (control condition) that can be corrected by the exposure device (900) is determined from the predicted exposure treatment result. Therefore, by measuring the actual result of the exposure treatment, it is possible to determine whether the result of the exposure treatment can be accurately corrected (whether the result of the exposure treatment is improved). If the result of the exposure treatment is not accurately corrected, or if there is almost no effect, it is desirable to retrain (regenerate) the prediction model or stop the prediction using the prediction model.

[0048] In the process of predicting the results of the exposure treatment of a substrate, it is presupposed that correction is performed by the exposure device (900). Accordingly, a correction amount (control condition) that can be corrected by the exposure device (900) may be obtained by a general affine transformation, and a prediction model may be generated such that the precision of the prediction model for that correction amount is optimized. This is advantageous for suppressing the decrease in the precision of the prediction model caused by components that cannot be corrected by the exposure device (900).

[0049] In this embodiment, the result of the exposure treatment of the first substrate cannot be predicted. For this reason, the first substrate is defined as a dummy substrate, and the result of the exposure treatment of the second and subsequent substrates may be predicted based on device data and measurement data obtained using such a dummy substrate.

[0050] In addition, in this embodiment, the result of the exposure processing predicted using a prediction model may be displayed on the output unit (917). For example, as shown in FIG. 7a, the result of the exposure processing predicted using a prediction model is displayed as a two-dimensional map image on a substrate unit. In FIG. 7a, the overlay is indicated by an arrow for each shot area of ​​the substrate. The starting point of the arrow indicates the predicted location (prediction point) of the overlay, the direction of the arrow indicates the direction of the overlay, and the length of the arrow indicates the amount (size) of the overlay. In addition to this map of the overlay, the average, standard deviation, maximum, and minimum values ​​of the overlay may be displayed as statistical numbers. The image of the two-dimensional map of the overlay shown in FIG. 7a may be recorded in the memory unit of the control unit (916), for example.

[0051] Additionally, the result of the exposure processing obtained by feedforwarding to the exposure device (900) based on the image of the two-dimensional map of the overlay shown in FIG. 7a may be displayed in the output unit (917). For example, as shown in FIG. 7b, the overlay is displayed on a substrate unit as a two-dimensional map image as the result of the exposure processing obtained by feedforwarding to the exposure device (900). In FIG. 7b, the overlay is indicated by an arrow for each shot area of ​​the substrate. The starting point of the arrow indicates the location (measurement point) where the overlay was measured, the direction of the arrow indicates the direction of the overlay, and the length of the arrow indicates the amount (size) of the overlay. In addition to this map of the overlay, the average, standard deviation, maximum, and minimum values ​​of the overlay may be displayed as statistical numbers. The image of the two-dimensional map of the overlay shown in FIG. 7b may be recorded in the memory unit of the control unit (916), for example.

[0052] The method for manufacturing an article according to an embodiment of the present disclosure is suitable for manufacturing articles such as devices (semiconductor devices, magnetic memory media, liquid crystal display devices, etc.). This manufacturing method includes the steps of: exposing a substrate to light using an exposure device (900) to form a pattern on the substrate; processing the substrate on which the pattern is formed; and manufacturing an article from the processed substrate. In addition, this manufacturing method may include other well-known steps (oxidation, film formation, deposition, doping, planarization, etching, resist stripping, dicing, bonding, packaging, etc.). The method for manufacturing an article according to the present embodiment is advantageous in at least one of the performance, quality, productivity, and production cost of the article compared to the prior art.

[0053] Other embodiments

[0054] Embodiments of the present invention may be implemented by a method performed by a computer of a system or device comprising one or more circuits (e.g., application-specific integrated circuits (ASICs)) that perform one or more functions of the aforementioned embodiment(s) of the present invention, by reading and executing computer-executable instructions (e.g., one or more programs) recorded in a storage medium (which may be referred to more specifically as a 'non-transient computer-readable storage medium') to perform one or more functions of the aforementioned embodiment(s), or by reading and executing computer-executable instructions from a storage medium to perform one or more functions of the aforementioned embodiment(s). The computer may have one or more central processing units (CPUs), microprocessors (MPUs), or other circuits, and may have a network of separate computers or separate computer processors. Computer-executable instructions may be given to the computer, for example, from a network of storage media. The storage medium may be, for example, one or more hard disks, random access memory (RAM), read-only memory (ROM), or a distributed computing system. Storage, optical disc (Compact Disc (CD), Digital Multifunction Disc (DVD), or Blu-ray Disc (BD) TM You may also provide flash memory devices, memory cards, etc.), etc.

[0055] The present invention can be executed in a process in which a program realizing one or more functions of the above-described embodiment is supplied to a system or device via a network or storage medium, and one or more processors in a computer of the system or device read and execute the program. In addition, it can also be executed by a circuit (e.g., an ASIC) realizing one or more functions.

[0056] According to the present disclosure, a technique advantageous for generating a prediction model for predicting the processing results of a substrate can be provided.

[0057] Although the present invention has been described with reference to exemplary embodiments, it is obvious that the invention is not limited to these embodiments. The scope of protection of the following claims shall be interpreted as broadly as possible to encompass all variations, equivalent structures, and functions.

Claims

Claim 1 A generating method for generating a prediction model for predicting the processing result of a substrate processed by a substrate processing device, comprising the step of generating the prediction model based on device data representing the state of the substrate processing device when processing up to the Nth substrate (N is a natural number) among the plurality of substrates, obtained by processing a plurality of substrates continuously under the same processing conditions by the substrate processing device, and measurement data obtained by measuring the processing result of the (N+1)th substrate among the plurality of substrates. Claim 2 In claim 1, the generating step comprises generating the prediction model based on the measurement data and a plurality of device data representing the state of the substrate processing device when each of the plurality of substrates up to the Nth substrate is processed. Claim 3 In claim 1, the generating step comprises generating the prediction model based on device data indicating the state of the substrate processing device when the Nth substrate is processed, and the measurement data. Claim 4 A method for generating, wherein the device data includes data regarding a device state that changes over time when the substrate processing device processes the plurality of substrates. Claim 5 A method for generating, wherein the substrate processing device comprises an exposure device that exposes a substrate through a disc, and the device data comprises the position and deviation of a substrate stage holding the substrate, the position and deviation of a disc stage holding the disc, an alignment result, and at least one of the environment in which the exposure device is installed. Claim 6 In claim 1, the device data comprises a generation method including data regarding the state of a device other than the control target in the substrate processing device. Claim 7 A method for generating, wherein the device data includes data regarding the environment of a cleanroom in which the substrate processing device is installed, in accordance with claim 1. Claim 8 A method for generating, wherein the measurement data comprises data regarding at least one of overlay, focus, line width uniformity, and edge placement error. Claim 9 A substrate processing method for processing a substrate using a substrate processing device, comprising the steps of: using a prediction model generated by the generation method described in Claim 1 to predict the processing result of a substrate processed by the substrate processing device and obtaining the predicted processing result; and determining the control conditions of the substrate processing device when processing the substrate by the substrate processing device based on the predicted processing result, and processing the substrate while controlling the substrate processing device according to the control conditions. Claim 10 A substrate processing method according to claim 9, further comprising the step of displaying the prediction processing result as an image of a two-dimensional map. Claim 11 A substrate processing method according to claim 10, further comprising the step of recording the above image. Claim 12 A substrate processing method according to claim 9, further comprising the step of displaying the processing result of the substrate when the substrate is processed while controlling the substrate processing device according to the control conditions determined based on the prediction processing result. Claim 13 A substrate processing method according to claim 12, further comprising the step of recording the processing result of the substrate when the substrate is processed while controlling the substrate processing device according to the control conditions determined based on the prediction processing result. Claim 14 A substrate processing device for processing a substrate, comprising: an acquisition unit configured to obtain a predicted processing result by predicting the processing result of a substrate processed by the substrate processing device using a prediction model generated by the generation method described in Claim 1; and a processing unit configured to determine control conditions of the substrate processing device when processing the substrate by the substrate processing device based on the predicted processing result, and to process the substrate while controlling the substrate processing device according to the control conditions. Claim 15 An information processing device for generating a prediction model for predicting the processing result of a substrate processed by a substrate processing device, the information processing device having a generation unit configured to generate the prediction model based on device data representing the state of the substrate processing device when processing up to the Nth substrate (N is a natural number) among the plurality of substrates, obtained by processing a plurality of substrates continuously under the same processing conditions by the substrate processing device, and measurement data obtained by measuring the processing result of the (N+1)th substrate among the plurality of substrates. Claim 16 A method for manufacturing an article having the steps of: forming a pattern on a substrate using the substrate processing method described in claim 9; processing the substrate on which the pattern is formed in the forming step; and manufacturing an article from the processed substrate. Claim 17 A program stored on a medium to execute the generation method described in claim 1 on a computer.

Images