Measurement method in the presence of array-based CMOS (CuA) structures using an effective medium model through physical modeling

Abstract

The system may include a controller, which includes one or more processors configured to execute program instructions, the program instructions causing one or more processors to execute a measurement recipe by receiving optical measurement data relating to a training sample including an under-array CMOS (CuA) device, which includes a complementary metal-oxide-semiconductor (CMOS) structure disposed beneath a periodic memory array structure; creating a first measurement model for determining measurements of the CuA device based on the optical measurement data, wherein the CMOS structure is modeled as a CMOS-effective medium within the first measurement model; receiving reference data relating to the training sample; updating the first measurement model to a second measurement model, which includes values ​​for variability parameters of the CMOS-effective medium; receiving optical measurement data relating to a test sample including the CuA device; and generating one or more measurement values ​​for the CuA device based on the second measurement model.

Description

Technical Field

[0001] The present disclosure generally relates to optical measurement methods, and more particularly to optical measurement methods for memory structures including embedded CMOS structures.

Background Art

[0002] One approach to meeting the demand for making memory devices (e.g., 3D memory devices) more performant while maintaining or reducing their physical size is to fabricate CMOS circuits (e.g., logic circuits) beneath the memory array structure. This approach is commonly referred to as under-array complementary metal-oxide-semiconductor (CMOS) (CuA) technology.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, since the presence of CMOS circuits below can affect the measurement values of the memory array structure, CuA technology has presented unique challenges with respect to measurement system used for process control. Therefore, there is a need to develop systems and methods to address the above problems.

Means for Solving the Problems

[0005] A system according to one or more exemplary embodiments of the present disclosure is disclosed. In the embodiments, the system includes a controller including one or more processors configured to execute program instructions, the program instructions receiving optical measurement data relating to one or more training samples including under-array CMOS (CuA) devices comprising complementary metal-oxide-semiconductor (CMOS) structures disposed beneath a periodic memory array structure, a first measurement model for determining one or more measurements of the CuA devices based on the optical measurement data relating to one or more training samples, the first measurement model being parameterized by geometric parameters of the memory array structure, variability parameters of the memory array structure, geometric parameters of the CMOS effective medium, and variability parameters of the CMOS effective medium The measurement recipe is performed by one or more processors by creating a first measurement model, receiving reference data for one or more training samples, determining the values ​​of geometric parameters of a memory array structure based on the reference data, determining the values ​​of variability parameters of the CMOS effective medium by fixing the values ​​of the geometric parameters of the memory array structure with optical measurement data for one or more training samples as input, updating the first measurement model to a second measurement model that includes the values ​​of the variability parameters of the CMOS effective medium, receiving optical measurement data for one or more test samples including CuA devices, and generating one or more measurement values ​​of CuA devices on one or more test samples based on the optical measurement data for one or more test samples using the second measurement model.

[0006] A system according to one or more exemplary embodiments of the present disclosure is disclosed. In an embodiment, the system includes an optical characterization system. In an embodiment, the system includes a reference characterization system. In an embodiment, the system includes a controller communicatively coupled to the optical characterization system and the reference characterization system, the controller including one or more processors configured to execute program instructions, the program instructions receiving optical measurement data from the optical characterization system for one or more training samples including under-array CMOS (CuA) devices comprising complementary metal-oxide-semiconductor (CMOS) structures disposed beneath a periodic memory array structure, a first measurement model for determining one or more measurements of the CuA devices based on the optical measurement data for one or more training samples, the model of the photomatter interaction with the CuA devices, the CMOS structure being modeled as a CMOS effective medium within the first measurement model, the geometric parameters of the memory array structure, the variability parameters of the memory array structure, the geometric parameters of the CMOS effective medium and the variability of the CMOS effective medium The measurement recipe is performed by one or more processors by creating a first measurement model parameterized by parameters, receiving reference data for one or more training samples from a reference characterization system, determining the values ​​of geometric parameters of a memory array structure based on the reference data, determining the values ​​of variability parameters of the CMOS effective medium by fixing the values ​​of the geometric parameters of the memory array structure with optical measurement data for one or more training samples as input, updating the first measurement model to a second measurement model that includes the values ​​of the variability parameters of the CMOS effective medium, receiving optical measurement data for one or more test samples including CuA devices from an optical characterization system, and generating one or more measurement values ​​of CuA devices on one or more test samples using optical measurement data for one or more test samples based on the second measurement model.

[0007] Methods according to one or more exemplary embodiments of the present disclosure are disclosed. In embodiments, the method includes generating optical measurement data for one or more training samples, including under-array CMOS (CuA) devices, which include complementary metal-oxide-semiconductor (CMOS) structures disposed beneath a periodic memory array structure. In embodiments, the method includes creating a first measurement model for determining one or more measurements of a CuA device based on optical measurement data for one or more training samples, which incorporates a model of photo-matter interaction with the CuA device, where the CMOS structure is modeled as a CMOS-effective medium within the first measurement model, and which is parameterized by geometric parameters of the memory array structure, variability parameters of the memory array structure, geometric parameters of the CMOS-effective medium, and variability parameters of the CMOS-effective medium. In embodiments, the method includes generating reference data for one or more training samples. In embodiments, the method includes determining values ​​for the geometric parameters of the memory array structure based on the reference data. In embodiments, this method includes determining the value of a variation parameter of the CMOS effective medium by fixing the value of the geometric parameter of the memory array structure with optical measurement data for one or more training samples as input. In embodiments, this method includes updating a first measurement model to a second measurement model that includes the value of the variation parameter of the CMOS effective medium. In embodiments, this method includes generating optical measurement data for one or more test samples that include CuA devices. In embodiments, this method includes generating one or more measurement values ​​for CuA devices on one or more test samples using optical measurement data for one or more test samples based on the second measurement model.

[0008] It should be understood that the general statements above and the detailed statements below are illustrative and explanatory only and do not necessarily limit the claimed invention. The accompanying drawings incorporated herein and forming part thereof illustrate embodiments of the invention and, together with the general statements, serve to illustrate the principles of the invention.

[0009] The numerous advantages of this disclosure can be better understood by those skilled in the art by referring to the attached figures. [Brief explanation of the drawing]

[0010] [Figure 1A] This is a block diagram of a measurement system according to one or more embodiments of the present disclosure. [Figure 1B] This is a simplified schematic diagram of a characterization subsystem configured as an optical characterization subsystem according to one or more embodiments of the present disclosure. [Figure 1C] This is a simplified schematic diagram of a characterization subsystem configured as an X-ray characterization subsystem according to one or more embodiments of the present disclosure. [Figure 1D] This is a simplified schematic diagram of a characterization subsystem configured as a particle beam characterization subsystem according to one or more embodiments of the present disclosure. [Figure 2] This is a simplified schematic diagram of a CuA device according to one or more embodiments of the present disclosure. [Figure 3] This flowchart illustrates the steps taken in a method for characterizing a CuA device according to one or more embodiments of the present disclosure. [Figure 3 (cont.)] This flowchart illustrates the steps taken in a method for characterizing a CuA device according to one or more embodiments of the present disclosure. [Modes for carrying out the invention]

[0011] Herein, we refer in detail to the disclosed subject matter shown in the accompanying drawings. This disclosure is shown and described in detail with respect to several embodiments and their particular features. The embodiments described herein should be considered illustrative rather than limiting. It will be readily apparent to those skilled in the art that various changes and modifications in shape and detail can be made without departing from the spirit and scope of this disclosure.

[0012] Embodiments of the present disclosure relate to systems and methods for optical measurement techniques based on creating a multi-step physical basis model for array-under complementary metal-oxide-semiconductor (CMOS) (CuA) devices that breaks undesirable interactions between constituent elements within the CuA device (e.g., photo-matter interactions with the CuA device).

[0013] A CuA structure (e.g., a CuA memory structure) may generally include logic circuits (e.g., CMOS logic circuits, etc.) physically placed (e.g., arranged) beneath a memory array structure (e.g., a three-dimensional (3D) memory stack, a 3D NAND structure, etc.). As used herein, the term CuA structure may encompass a wide variety of designs for logic array structures and memory array structures. Thus, this disclosure is not limited to any particular CuA architecture.

[0014] Optical measurement methods are widely used for semiconductor process control because they offer relatively high measurement throughput and are typically non-destructive. In optical measurement, a sample is illuminated with light, and measurements are generated based on the corresponding light emitted from the sample. Optical measurement of subsurface features typically requires that the light propagates through at least the top of the sample to reach the subsurface feature of interest. As a result, optical measurement systems typically use wavelengths of light selected to propagate through the target structure with relatively low absorbance.

[0015] However, in the case of CuA structures, incident light may interact with both the memory array structure and the embedded CMOS structure in a way that could impair the ability to isolate and generate measurements related to the memory array structure. This specification assumes that existing optical measurement techniques may be insufficient to accurately characterize CuA structures, especially as dimensions shrink and device complexity increases. For example, some techniques may rely on wavelengths of light that are within the transmission band of the memory array structure in question but can be at least partially absorbed by the underlying logic circuits. For example, some logic circuits may use polysilicon layers that absorb light with wavelengths greater than approximately 450 nanometers (nm). In this case, optical measurement techniques with wavelengths less than approximately 450 nm can isolate and generate measurements of the memory array structure. However, such techniques may be constrained to selecting CuA designs that incorporate such absorbing materials, may have limited sensitivity to deep structures, and may have limited value compared to broadband optical measurement methods that benefit from multi-wavelength data. As another example, such techniques rely on supervised training of artificial neural networks using optical measurement data with labels generated by additional measurement methods. However, this technique may have various limitations, including, but are not limited to, the need for high-quality sampling of the ground truth reference data, the considerable time required to generate enough labels for training, limited performance for deeply embedded structures, insensitivity to process changes, and general inability to apply to CuA structures outside the training dataset.

[0016] In some embodiments of this disclosure, a physical base model is created to associate optical measurement data of a CuA device (e.g., spectroscopic data, polarization analysis data, reflected light analysis data, scattered light analysis data, etc.) with the geometric parameters of a CuA device, for the purpose of generating measurements such as, but not limited to, superposition measurements, critical dimension (CD) measurements, shape measurements (e.g., height measurements, tilt measurements, sidewall angle measurements, etc.), stress measurements, composition measurements, bandgap measurements, electrical property measurements, process state (e.g., focus and / or dose state, resist state, partial pressure, temperature, focusing model, etc.), defect identification, and defect classification.

[0017] In some physical substrate models, such as but not limited to the exact coupled-wave analysis (RCWA) method, the structure under study may need to be periodic. In some embodiments, effective medium modeling techniques are used to model the CMOS structure as the effective medium. This method requires accurate determination of both the variability and thickness of the effective medium layer. However, the parameters of the effective medium approximation of the CMOS structure (e.g., thickness, variability) may be strongly correlated with the parameters of the memory array structure, potentially resulting in instability, interrelationships, and parameter interactions. For example, floating all parameters associated with the CMOS effective medium and memory array structure can lead to undesirable interrelationships between coexisting parameters, potentially resulting in inaccurate measurements.

[0018] In some embodiments of the present disclosure, this undesirable coupling is avoided by a multi-step modeling process in which the variation parameters of the CMOS effective medium are first determined using accurate reference data that describes the geometry of the memory array structure. Specifically, the geometric parameters are fixed based on accurate reference measurements in a first measurement model, while the variation parameters of the CMOS effective medium are floated. This variation parameter of the CMOS effective medium can then be fixed in a second measurement model that may be suitable for generating measurements of CuA devices with unknown characteristics (e.g., in a manufacturing environment) in a manner that also avoids the aforementioned undesirable correlations.

[0019] In some embodiments, the second measurement mode is further used to generate synthetic data for training a supervised machine learning model. Additional reference measurements can be used to supplement and / or cross-validate the supervised machine learning model.

[0020] Referring now to FIGS. 1A - 3, a system and method for characterizing a CuA structure are described in more detail in accordance with one or more embodiments of the present disclosure.

[0021] FIG. 1A is a block diagram of a measurement system 100 according to one or more embodiments of the present disclosure. In some embodiments, measurement system 100 includes a characterization subsystem 102 for generating measurement data of sample 104 using optical techniques, and a controller 106 for generating one or more measurements based on the measurement data.

[0022] Characterization subsystem 102 may include any component or combination of components suitable for generating measurement data regarding sample 104.

[0023] In some embodiments, the characterization subsystem 102 includes an optical characterization subsystem 102 for generating measurement data based on the interaction of the sample 104 with light. For example, the characterization subsystem 102 may include, but is not limited to, a spectroscopic ellipsometer (SE), an SE with multiple irradiation angles, an SE that measures Müller matrix components (e.g., using a rotational compensator), a single-wavelength ellipsometer, a beam profile ellipsometer (angle-resolved ellipsometer), a beam profile reflectometer (angle-resolved reflectometer), a broadband reflectometer (spectroscopic reflectometer), a single-wavelength reflectometer, an angle-resolved reflectometer, an imaging system, a scattermeter (e.g., a speckle analyzer), or any combination thereof.

[0024] In some embodiments, the characterization subsystem 102 includes an X-ray characterization subsystem 102 for generating measurement data based on the interaction of sample 104 with X-rays. For example, the characterization subsystem 102 may include, but is not limited to, a small-angle X-ray scattering (SAXS) system or an X-ray reflection scattering (SXR) system.

[0025] In some embodiments, the characterization subsystem 102 includes a particle beam characterization subsystem 102 for generating measurement data based on the interaction of the sample 104 with a particle beam, such as but not limited to an electron beam (e-beam), an ion beam, or a neutral particle beam.

[0026] In some embodiments, the characterization subsystem 102 provides multiple types of measurements. In some embodiments, the measurement system 100 includes multiple measurement subsystems 102, each providing a different combination of one or more measurements. Furthermore, the measurement system 100 may be provided as a single tool or as a group of tools. A single tool providing multiple measurement settings is generally described in U.S. Patent No. 7,933026, issued April 26, 2011, which is incorporated by reference in its entirety. Multiple tools and structural analyses are generally described in U.S. Patent No. 7,478019, issued January 13, 2009, which is incorporated by reference in its entirety.

[0027] Furthermore, U.S. Patent No. 1,0458912, published on October 29, 2019, with the title "Model based optical measurements of semiconductor structures with anisotropic dielectric permittivity"; U.S. Patent No. 1,1573077, published on February 7, 2023, with the title "Scatterometry based methods and systems for measurement of strain in semiconductor structures"; U.S. Patent No. 1,1036898, published on June 15, 2021, with the title "Measurement models of nanowire semiconductor structures based on re-useable sub-structures"; U.S. Patent No. 1,1555689, published on January 17, 2023, with the title "Measuring thin films on grating and bandgap on grating"; and U.S. Patent No. 1,1555689, published on October 26, 2021, with the title "Measurement methodology of advanced U.S. Patent No. 11156548, titled "nanostructures," and U.S. Patent No. 10794839, published on October 6, 2020, titled "Visualization of three-dimensional semiconductor structures," are both incorporated herein by reference in their entirety.

[0028] In some embodiments, the controller 106 includes one or more processors 108, each processor 108 configured to execute a set of program instructions held in memory 110 or a memory device, which can cause the processor 108 to perform various operations.

[0029] One or more processors 108 of the controller 106 may include any processor or processing element known in the art. For the purposes of this disclosure, the terms “processor” or “processing element” may be broadly defined to include any device having one or more processing elements or logic elements (e.g., one or more microprocessor devices, one or more application-specific integrated circuit (ASIC) devices, one or more field-programmable gate arrays (FPGAs), or one or more digital signal processors (DSPs)). In this sense, one or more processors 108 may include any device configured to execute algorithms and / or instructions (e.g., program instructions stored in memory). In some embodiments, one or more processors 108 may be implemented as a desktop computer, mainframe computer system, workstation, image computer, parallel processor, network computer, or any other computer system configured to execute a program configured to operate with the characterization subsystem 102 as described throughout this disclosure. Furthermore, different subsystems of the measurement system 100 may include processor elements or logic elements which are suitable for performing at least some of the steps described in this disclosure. Accordingly, the above description should be interpreted as illustrative only and not as a limitation to embodiments of the present disclosure. Furthermore, the steps described throughout the present disclosure can be performed by a single controller, or alternatively, by multiple controllers. In addition, controller 106 may include one or more controllers housed in a common housing or multiple housings. Thus, any controller or combination of controllers may be separately packaged as a module suitable for integration into the measurement system 100.

[0030] The memory 110 may include any storage medium known in the art that is suitable for storing program instructions executable by one or more associated processors 108. For example, the memory 110 may include a non-temporary memory medium. Another example of the memory 110 may include, but is not limited to, read-only memory (ROM), random access memory (RAM), magnetic or optical memory devices (e.g., disks), magnetic tape, solid-state drives, etc. It should be further noted that the memory 110 may be housed in a common controller housing together with one or more processors 108. In some embodiments, the memory 110 may be located remotely from the physical locations of one or more processors 108 and the controller 106. For example, one or more processors 108 of the controller 106 may have access to remote memory (e.g., a server) accessible via a network (e.g., the Internet, an intranet, etc.).

[0031] The controller 106 can be communicatively coupled to any component or combination of components of the measurement system 100. In some embodiments, the controller 106 can receive data (e.g., measurement data) from one or more components of the measurement system 100. In some embodiments, the controller 106 controls one or more components of the measurement system 100 by drive signals. More generally, the controller 106 can perform any of the steps described herein.

[0032] In some embodiments, the controller 106 generates one or more measurements of sample 104 based at least in part on measurement data generated by the characterization subsystem 102. Measurements of the parameters of interest may include multiple algorithms, which can be performed by the controller 106. For example, the optical interaction of incident light with sample 104 can be modeled using an EM (electromagnetic field) solver, and algorithms such as, but not limited to, exact coupled-wave analysis (RCWA), finite element method (FEM), moment method, surface integral, volume integral, or finite difference time-domain (FDTD) method can be used. Sample 104 can be modeled (e.g., parameterized) using a geometry engine, a process modeling engine, or a combination of both. The use of process modeling is generally described in U.S. Patent No. 1,0769320, issued September 8, 2020, which is incorporated herein by reference in its entirety. Geometry engines are implemented, for example, in AcuShape software from KLA Corporation.

[0033] Controller 106 can analyze the collected measurement data using any suitable combination of data fitting and / or optimization techniques, including but not limited to libraries, rapid model reduction, regression, statistical methods (see, for example, "Statistical model-based metrology" by S. Pandev et al., U.S. Patent No. 1,0101670), machine learning algorithms (e.g., neural networks, support vector machines (SVM), principal component analysis (PCA), independent component analysis (ICA), local linear embeddings (LLE), and more generally, dimensionality reduction techniques), sparse representation techniques, Fourier transform techniques, wavelet transform techniques, or Kalman filtering. Statistical model-based metrology is generally described in U.S. Patent No. 1,0101670, issued on October 16, 2018, which is incorporated herein by reference in its entirety. Controller 106 can analyze the collected measurement data using algorithms that do not involve modeling, optimization, and / or fitting. Patterned wafer characterization is generally described in U.S. Patent No. 1,0502,694, issued November 10, 2019, which is incorporated herein by reference in its entirety. In some embodiments, the controller 106 uses one or more algorithms to drive matching from the same or different tool types (e.g., different instances or configurations of the characterization subsystem 102).

[0034] The controller 106 can be designed to deliver efficient performance through any suitable technique, including but not limited to parallelization, computation distribution, load balancing, multi-service support, and dynamic load optimization. Furthermore, the controller 106 can perform any step using any type or combination of configurations, including but not limited to dedicated hardware (e.g., FPGA), software, or firmware.

[0035] The controller 106 can further generate any type of measurement of sample 104 (or a portion thereof) based at least partially on the measurement data from the characterization subsystem 102. In some embodiments, the controller 106 generates measurement method measurements, including but not limited to superposition measurements, critical dimension (CD) measurements, shape measurements (e.g., height measurements, tilt measurements, sidewall angle measurements, etc.), stress measurements, composition measurements, bandgap measurements, electrical property measurements, or process state measurements (e.g., focus and / or dose state, resist state, partial pressure, temperature, focusing model, etc.). In some embodiments, the controller 106 generates inspection measurements in which at least one of one or more defects on sample 104 is identified and classified.

[0036] Any of the measurement system 100 and its components (e.g., the characterization subsystem 102, the controller 106, etc.) can be configured to perform a recipe (e.g., a measurement recipe), which can define various setting parameters and / or steps to be performed in a given measurement or a series of measurements.

[0037] For example, the recipe may include various aspects of the design of sample 104 (e.g., the design of the CuA device 202 on sample 104), including but not limited to the layout of features on one or more sample layers, the size of the features, or the pitch of the features. Another example is that the recipe may include illumination parameters such as, but not limited to, the wavelength of the illumination light, the illumination pupil distribution (e.g., the distribution of illumination angles and the associated illumination intensity at those angles), the polarization of the incident illumination light, the spatial distribution of the illumination, or the height of the sample. Another example is that the recipe may include acquisition parameters such as, but not limited to, the acquisition pupil distribution (e.g., the desired distribution of angled light from the sample to be used for measurement and the associated filtered intensity at those angles), the acquisition field aperture setting for selecting the portion of the sample to be measured, the polarization of the acquired light, or the wavelength filter. Another example is that the recipe may include various processing steps (which can be performed, for example, by the controller 106) for generating measurements based on the measurement data generated according to the recipe.

[0038] Referring now to Figure 2, which is a simplified schematic diagram of a CuA device 202 according to one or more embodiments of the present disclosure. The CuA device 202 may include a memory array structure 204 and various CMOS structures 206 (e.g., logic structures) located below the memory array structure 204.

[0039] The memory array structure 204 may include any number or type of structures suitable for forming a memory array. For example, the memory array structure 204 may include, but is not limited to, a 3D NAND structure formed from patterned features 208 in the multilayer stack 210. Furthermore, such a memory array structure 204 is typically a periodic structure whose periodicity follows one or more dimensions.

[0040] The CMOS structure 206 may include any number or type of structures fabricated beneath the memory array structure 204. For example, the CMOS structure 206 may, but is not required, be suitable for controlling and / or powering the memory array structure 204. In this way, the combination of the CMOS structure 206 and the memory array structure 204 may form a memory device (e.g., a 3D memory device). Furthermore, the CMOS structure 206 may typically have a spatially varying distribution in which the number and / or design of its constituent features may not be periodic across the CuA device 202. In this way, the CMOS structure 206 can generally be described as aperiodic. However, it should be noted that the CMOS structure 206 may exhibit local periodicity in some regions.

[0041] Furthermore, the memory array structure 204 and / or CMOS structure 206 may generally have any design, such that the term CuA device 202 as used herein is not limited to any particular design. For example, the CuA device 202 may include an intervening layer between the memory array structure 204 and the CMOS structure 206, such as but not limited to a source layer 212 (e.g., a polysilicon source layer). As another example, although not shown, the CuA device 202 may include an intervening layer between the CMOS structure 206 and the substrate 214.

[0042] Referring now to Figure 3, techniques for characterizing the CuA device 202 or a portion thereof are described in more detail by one or more embodiments of this disclosure.

[0043] It may be desirable to generate measurements of the structures constituting the CuA device 202 (e.g., the CMOS structure 206 and / or the memory array structure 204) at various stages of the fabrication process. Such measurements may include, but are not limited to, measurement measurements or defect measurements (e.g., inspection measurements). Measurement measurements may include, but are not limited to, superposition measurements between structures fabricated by various lithographic exposures, critical dimension (CD) measurements of one or more features, heights of one or more features, or inclinations of one or more features. Inspection measurements may include, but are not limited to, identifying and / or characterizing defects in the fabrication process (e.g., undesirable features, missing features, improperly shaped or placed features, etc.). Furthermore, such measurements can be used for a wide variety of purposes, including, but not limited to, process control and placement, or to evaluate the performance of the fabricated CuA device 202.

[0044] Measurements can be generated after any process step for fabricating the CuA device 202. For example, measurements can be generated after fabricating the CMOS structure 206 and / or after fabricating the memory array structure 204 to form the complete CuA device 202. Exemplary, measurements of the complete CuA device 202, including both the memory array structure 204 and the underlying CMOS structure 206, are referred to herein as “full-loop” measurements.

[0045] Measurements taken at a given process step are generally assumed to provide information about any features fabricated on sample 104, depending on the interaction of the irradiation line 114 with sample 104. In this way, it can be difficult to isolate measurements of newly fabricated features. For example, full-loop measurements generally provide information about both the memory array structure 204 and the underlying CMOS structure 206, or may be influenced by both, which may limit or impair the ability to generate isolated measurements of the memory array structure 204.

[0046] In some embodiments, measurements of various test structures can be generated to assist in generating measurements of several features in isolation. For example, measurements of a test structure including a memory array structure 204 that does not have a corresponding buried CMOS structure 206 are referred to herein as “short-loop” measurements.

[0047] Furthermore, measurements at any process step can generally be generated using any suitable technique, including but not limited to optical techniques, X-ray techniques, and particle-based techniques. However, different measurement techniques may have different trade-offs. For example, optical measurement techniques can generally provide high measurement throughput for non-destructive measurements, but their resolution may be limited or they may be limited to certain types of structures (e.g., periodic structures) based on the corresponding analysis or modeling step. Therefore, optical measurements are typically used during runtime when throughput is particularly important. As another example, X-ray and / or particle-based techniques offer higher resolution than some optical techniques, but they have the disadvantage of relatively low throughput and / or can result in destructive measurements. Consequently, such techniques are typically used for reference measurements.

[0048] However, it is assumed herein that generating all possible types of measurements in every measurement step may not be feasible or desirable in all application examples. In such cases, various techniques can be used to generate measurements of a particular structure (e.g., isolated measurements of the memory array structure 204) depending on the available data.

[0049] Figure 3 is a flowchart illustrating the steps taken in a method 300 for characterizing a CuA device 202 according to one or more embodiments of the present disclosure. The applicant notes that embodiments and techniques described herein in relation to the measurement system 100 should be interpreted as being extended to method 300. For example, any of the steps associated with method 300 can be carried out by the controller 106 and / or the characterization subsystem 102 of the measurement system 100. However, it should be further noted that method 300 is not limited to the architecture of the measurement system 100.

[0050] Method 300 is suitable for, but not limited to, applications where it is desirable to isolate and generate measurements of the memory array structure 204 without using short-loop measurements. One goal of this method is to create an effective representation of a CMOS structure 206 that can be used in physical foundation models that rely on the presence of periodic structures, such as, but not limited to, RCWA technology. One obstacle associated with this method is that the CMOS structure 206 is not generally periodic and therefore does not directly fit into such physical foundation models that require periodic structures.

[0051] In some embodiments, effective medium modeling techniques are used to model the CMOS structure 206 as a one-dimensional (1D) effective medium layer (referred to herein as the CMOS effective medium). This technique requires accurate determination of both the variability and thickness of the 1D effective medium layer. However, both of these parameters may be highly interrelated with the parameters of the memory array structure 204, which can result in potential instability, interrelationships, and parameter interactions in the overall model of the CuA device 202. For example, floating all parameters associated with the CMOS effective medium and the memory array structure 204 can lead to undesirable interrelationships between coexisting parameters, potentially resulting in inaccurate measurements.

[0052] In some embodiments, the parameters of the effective medium approximation for the CMOS structure 206 are coupled and separated from the parameters of the memory array structure 204 using a multi-step process in which the variability parameters of the effective medium approximation are determined separately and then fixed to avoid undesirable interrelationships.

[0053] In some embodiments, Method 300 includes step 302 of generating optical measurement data for one or more training samples, including a CuA device 202. The optical measurement data may include any type of data generated by any type of optical system, such as, but not limited to, the optical characterization subsystem 102 shown in Figure 1B. For example, the optical measurement data may include, but not limited to, spectral data from an ellipsometer, reflectometer, or scattermeter.

[0054] Furthermore, in some embodiments, one or more training samples include known modifications of the CuA device 202 (e.g., known modifications related to process deviations). In this way, the effects of such modifications are identified and incorporated into the model of Method 300 described below. Such a procedure may be referred to as Design of Experiments (DOE) and can improve the robustness of Method 300.

[0055] In some embodiments, the method 300 includes step 304 of creating a first measurement model based on optical measurement data for determining one or more measurements (e.g., full-loop data) of a CuA device 202, wherein the first measurement model incorporates a model of the photo-matter interaction with the CuA device 202 (e.g., a physical substrate model), and the CMOS structure 206 is modeled as a CMOS-effective medium in the first measurement model.

[0056] The first measurement model can be parameterized by geometric and variability parameters of the memory array structure 204, as well as geometric and variability parameters of the CMOS effective medium (as described, for example, by the initial variability model). Geometric parameters may include, but are not limited to, any physical properties of any relevant structure, such as CD, feature height, feature inclination (e.g., sidewall angle), film thickness, superposition between features fabricated by different lithography exposures, or composition. Variability parameters may include, but are not limited to, any parameters related to the variability of any structure, including, but are not limited to, the refractive index and / or dielectric constant at any wavelength or modeling weight used to describe the variability using any modeling technique (e.g., harmonic oscillator model, Tauc-Lorentz model, Cauchy model, effective medium model, etc.).

[0057] Furthermore, the CMOS structure 206 can be modeled within the first measurement model using any preferred technique. As previously described herein, some physical substrate measurement models, such as but not limited to the RCWA model, require periodic structures. In some embodiments, the CMOS structure 206 may generally be aperiodic, but is modeled as an effective medium to suit such a physical substrate measurement model. The CMOS structure 206 may generally be modeled using any effective medium technique. For example, the CMOS structure 206 can be modeled as a single effective medium layer or film stack, in which case the variability characteristics of any such layer can be modeled using any variability modeling technique. Thus, the effective medium model for characterizing the CMOS structure 206 can be parameterized by parameters that represent thickness characteristics and / or variability characteristics (e.g., modeling weights in the variability model).

[0058] In some embodiments, method 300 includes step 306 of generating reference data for one or more training samples. In some embodiments, method 300 includes step 308 of determining the values ​​of geometric parameters of the memory array structure 204 and the CMOS effective medium based on the reference data.

[0059] Reference data can be generated using any preferred technique. In some embodiments, the reference data is generated by a high-resolution characterization subsystem 102, such as, but not limited to, an X-ray characterization subsystem 102 or a particle beam characterization subsystem 102. Non-limiting examples include, but are not limited to, transmission electron microscope (TEM) data, scanning electron microscope (SEM) data (e.g., critical dimension SEM (CD-SEM) data, electron beam SEM (EB-SEM) data, etc.), SAXS data (e.g., transmission SAXS (T-SAXS) data, CD-SAXS data, etc.), X-ray photoelectron spectroscopy (XPS) data, or X-ray diffraction (XRD) data.

[0060] In this way, accurate measurements of various geometric parameters related to the CuA device 202 can be provided directly from the reference data or by analyzing the reference data.

[0061] In some embodiments, Method 300 includes step 310 of determining the value of the variability parameter of the CMOS effective medium by fixing the value of the geometric parameter of the memory array structure 204 in a first measurement model with optical measurement data from one or more training samples as input. In other words, the value of the geometric parameter of the memory array structure 204 can be feedforward to the first measurement model as a fixed input. For example, the value of the variability parameter of the CMOS effective medium can be determined by fitting optical measurement data from one or more training samples to the first measurement model and determining the value of the floating variability parameter using any preferred technique, such as but not limited to regression.

[0062] As previously described herein, simply floating all parameters of both the memory array structure 204 and the CMOS active medium (e.g., the active medium model of the CMOS structure 206) may result in undesirable coupling between the parameters of the memory array structure 204 and the CMOS active medium, potentially leading to poor modeling performance and / or lack of robustness. However, this coupling is broken in step 310 by fixing the geometric parameters of the memory array structure 204 based on reference data (from step 308). As a result, the geometric parameters and / or variability parameters of the CMOS active medium can be accurately determined.

[0063] In some embodiments, method 300 includes step 312 of updating the first measurement model to a second measurement model that includes values ​​for the variation parameter of the CMOS effective medium.

[0064] This specification assumes that the values ​​of the variation parameters of the CMOS effective medium can be incorporated into the second measurement model in a variety of ways.

[0065] For example, the second measurement model may be substantially identical to the first measurement model, but the geometric parameter values ​​of the CMOS effective medium are no longer floating parameters, but rather fixed based on determined values ​​derived in step 308.

[0066] As another example, an effective medium model can be generated that incorporates values ​​of variability parameters from multiple locations on the training sample. For example, because at least one of the number or design of the embedded CMOS structures 206 varies locally, the determined values ​​of the variability parameters may indeed vary as a function of location along any of the CuA devices 202. Thus, in the second measurement model, the CMOS structures 206 can be characterized as an effective medium having a weighted average of the determined values ​​of multiple location-derived variability parameters. For example, the second measurement model may use Bruggeman's Effective Medium Approximation (BEMA) which includes values ​​of variability parameters from multiple locations on the training sample (e.g., multiple locations within a particular CuA device 202 or multiple locations across multiple CuA devices 202). Furthermore, this second measurement model can be parameterized by the geometric and variability parameters of the memory array structure 204, as well as the floating parameters of the BEMA model describing the CMOS structure 206 (e.g., the effective layer thickness and / or the weight of any model).

[0067] In some embodiments, method 300 includes step 314 of generating optical measurement data for one or more test samples, including a CuA device 202. In this way, the optical measurement data for the test samples can correspond to full-loop data.

[0068] In some embodiments, Method 300 includes step 316 of generating one or more metric measurements of the CuA device 202 from optical measurement data for one or more test samples based on a second measurement model. For example, the floating parameter values ​​of the second measurement model can be determined by fitting them to the optical measurement data of one or more test samples using any preferred technique, such as but not limited to regression. Various measurements (e.g., metric measurements, inspection measurements, etc.) can be determined directly from the fitted values ​​derived from the second measurement model, or by analyzing the fitted values. Non-limiting examples of such measurements include, but are not limited to, superposition measurements, critical dimension (CD) measurements, shape measurements (e.g., height measurements, tilt measurements, sidewall angle measurements, etc.), stress measurements, composition measurements, bandgap measurements, electrical property measurements, or process state measurements (e.g., focus and / or dose state, resist state, partial pressure, temperature, focusing model, etc.).

[0069] Subsequently, the measurements of the CuA device 202 on the test sample can be used for a variety of purposes. In some embodiments, the measurements are used for process control, deployment, and / or performance evaluation of the CuA device 202 on the test sample. For example, the measurements can be used to generate modifiable values ​​for one or more process tools (e.g., scanner, stepper, etc.) in a feedback process and / or feedforward process.

[0070] This specification further assumes that, in order to generate measurements of the CuA device 202, it may be desirable to combine physical foundation modeling (e.g., the second measurement model in step 312) with machine learning techniques. For example, such a combined approach may increase the robustness and / or accuracy of the measurements depending on the sensitivity of critical parameters, robustness requirements, reference sampling rate, etc. However, machine learning models typically require, or at least benefit from, a large training dataset to yield robust and reliable behavior. In some application examples, it may be impractical or undesirable to experimentally generate an entire training dataset.

[0071] In some embodiments, Method 300 includes step 318 of generating synthetic measurement data using a second measurement model having a wide range of parameters describing the synthetic CuA device 202. In some embodiments, Method 300 includes step 320 of training a machine learning model to generate one or more measurements based on training data which includes at least one of optical measurement data for one or more training samples, reference data for one or more training samples, or synthetic data describing the synthetic CuA device 202. Thus, the second measurement model can be used to provide supplemental training data other than the experimental data (e.g., optical measurement data and / or reference data) associated with the training and / or test samples described above.

[0072] For example, the synthetic measurement data may include a synthetic equivalent of optical measurement data (which may be generated, for example, by the optical characterization subsystem 102) based on various combinations of geometric and variability parameters of the CuA device 202. For example, various combinations of geometric and variability parameters of the effective medium models of the memory array structure 204 and the CMOS structure 206 can be provided as input to a second measurement model to generate synthetic measurement data as output. Thus, this synthetic measurement data can be characterized as a synthetic DOE suitable for providing training data for machine learning models.

[0073] In some embodiments, additional experimental data may be collected to further supplement the training dataset for the machine learning model. For example, the training data may include reference data that can be generated for one or more test samples and / or any additional samples (e.g., using the X-ray characterization subsystem 102, the particle beam characterization subsystem 102, etc.).

[0074] In some embodiments, the training data or a portion thereof is manipulated during the training process. For example, a dimensionality reduction process can be applied to any of the training data, including but not limited to optical measurement data and / or synthetic measurement data (e.g., from training and / or test samples). Any dimensionality reduction process, including but not limited to principal component analysis (PCA), can be used.

[0075] The machine learning model may include any suitable type of machine learning technique known in the art, and may include any combination of learning techniques, including but not limited to supervised learning techniques, unsupervised learning techniques, or reinforcement learning techniques. In some embodiments, the machine learning model is a neural network model.

[0076] Once trained, the machine learning model can be used to generate measurements of the CuA device 202 (e.g., superposition, CD, height, tilt, defect identification, defect classification, etc.) based on optical measurement data from additional test samples. In some embodiments, method 300 includes step 322 of generating optical measurement data for one or more additional test samples containing the CuA device 202. In some embodiments, method 300 includes step 324 of generating one or more measurement values ​​of the CuA device 202 on one or more additional test samples based on optical measurement data for one or more additional test samples using a machine learning model.

[0077] Referring here to Figures 1B to 1D, various non-limiting configurations of the characterization subsystem 102 are described in more detail in accordance with one or more embodiments of this disclosure.

[0078] In some embodiments, the characterization subsystem 102 is an optical measurement subsystem that generates measurement data based on the interaction of the sample 104 with light. Figure 1B is a simplified schematic diagram of the characterization subsystem 102 configured as an optical characterization subsystem 102 according to one or more embodiments of the present disclosure. For example, the characterization subsystem 102 may include, but is not limited to, a spectroscopic ellipsometer (SE), an SE having multiple irradiation angles, an SE measuring a Müller matrix component (e.g., using a rotational compensator), a single-wavelength ellipsometer, a beam profile ellipsometer (e.g., an angle-resolved ellipsometer), a beam profile reflectometer (e.g., an angle-resolved reflectometer), a broadband reflectometer (e.g., a spectroscopic reflectometer), a single-wavelength reflectometer, an angle-resolved reflectometer, an imaging system, a scattermeter (e.g., a speckle analyzer), or any combination thereof.

[0079] In some embodiments, the characterization subsystem 102 includes an illumination source 112 configured to generate at least one illumination line 114. The illumination light from the illumination source 112 may include one or more selected wavelengths of light, including but not limited to ultraviolet (UV) illumination, visible light illumination, or infrared (IR) illumination. For example, the characterization subsystem 102 may include one or more apertures in the illumination pupil plane for splitting the illumination light from the illumination source 112 into one or more illumination lines 114 or illumination light areas. In this regard, the characterization subsystem 102 may provide bipolar illumination, quadrupole illumination, etc. Furthermore, the spatial profiles of one or more illumination lines 114 on the sample 104 can be controlled by a field plane aperture to have any selected spatial profile.

[0080] The irradiation source 112 may include any type of irradiation source suitable for providing at least one irradiation line 114. In some embodiments, the irradiation source 112 is a laser source. For example, the irradiation source 112 may include, but is not limited to, one or more narrowband laser sources, broadband laser sources, supercontinuum laser sources, white light laser sources, etc. In some embodiments, the irradiation source 112 includes a laser-sustained plasma (LSP) source. For example, the irradiation source 112 may include, but is not limited to, an LSP lamp, LSP bulb, or LSP chamber suitable for containing one or more elements that can emit broadband irradiation light when excited to a plasma state by the laser source. In some embodiments, the irradiation source 112 includes a lamp source. In some embodiments, the irradiation source 112 may include, but is not limited to, an arc lamp, a discharge lamp, an electrodeless lamp, etc.

[0081] The irradiation source 112 can provide one or more irradiation lines 114 using free-space techniques and / or optical fibers.

[0082] In some embodiments, the characterization subsystem 102 directs the irradiation beam 114 to the sample 104 via an irradiation path 118 through at least one irradiation lens 116 (e.g., an objective lens). The irradiation path 118 may include one or more optical components suitable for correcting and / or adjusting the irradiation beam 114 and directing the irradiation beam 114 to the sample 104. In some embodiments, the irradiation path 118 includes one or more irradiation path optics 120 for shaping or otherwise controlling the irradiation beam 114. For example, the irradiation path optics 120 may include, but are not limited to, one or more field diaphragms, one or more pupil diaphragms, one or more polarizers, one or more filters, one or more beam splitters, one or more diffusers, one or more homogenizers, one or more apodizers, one or more beam shapers, or one or more mirrors (e.g., stationary mirrors, translational mirrors, scanning mirrors, etc.).

[0083] The characterization subsystem 102 can position the sample 104 for measurement using any preferred technique. As shown in Figure 1B, the characterization subsystem 102 includes a sample stage 122 which includes one or more actuators (e.g., linear actuators, tilt / inclination actuators, rotary actuators, etc.) for positioning the sample 104 relative to the irradiation beam 114. In some embodiments, although not explicitly shown, the characterization subsystem 102 includes beam scanning optics (e.g., galvanometer mirrors, scanning prisms, etc.) for adjusting the position and / or scanning one or more irradiation beams 114.

[0084] In some embodiments, the characterization subsystem 102 includes at least one collecting lens 124 for capturing light or other radiation emitted from a sample 104, referred herein to as collected light 126, and guiding this collected light 126 through a collection path 130 to one or more detectors 128. The collection path 130 may include one or more optical elements suitable for correcting and / or adjusting the collected light 126 from the sample 104. In some embodiments, the collection path 130 includes one or more collection path optics 132 for shaping or otherwise controlling the collected light 126. For example, the collection path optics 132 may include, but are not limited to, one or more field diaphragms, one or more pupil diaphragms, one or more polarizers, one or more filters, one or more beam splitters, one or more diffusers, one or more homogenizers, one or more apodizers, one or more beam shapers, or one or more mirrors (e.g., stationary mirrors, translationable mirrors, scanning mirrors, etc.).

[0085] The characterization subsystem 102 may generally include any number or type of detectors 128. For example, the characterization subsystem 102 may include at least one single-pixel detector 128, such as a photodiode, avalanche photodiode, or single-photon detector, but not limited to these. Another example is that the characterization subsystem 102 may include at least one multi-pixel detector 128, such as a charge-coupled device (CCD) or complementary metal-oxide-semiconductor (CMOS) device, a line detector, or a time-delay integral (TDI) detector, but not limited to these.

[0086] The detector 128 may be placed at any selected position within the collection path 130. In some embodiments, the characterization subsystem 102 includes the detector 128 in the field of view plane (e.g., a plane joined to the sample 104) to generate an image of the sample 104. In some embodiments, the characterization subsystem 102 includes the detector 128 in the pupil plane (e.g., the diffraction plane) to generate a pupil image. In this regard, the pupil image may correspond to the angular distribution of light from the sample 104 to the detector 128. For example, the diffraction order related to the diffraction of the irradiation line 114 from the sample 104 may be imaged or otherwise observed within the pupil plane. In a general sense, the detector 128 can capture any combination of reflected (or transmitted), scattered, or diffracted light from the sample 104.

[0087] The irradiation path 118 and collection path 130 of the characterization subsystem 102 can be oriented in a wide range of settings. For example, as shown in Figure 1B, the irradiation path 118 and collection path 130 may include non-overlapping optical paths. In some embodiments, although not explicitly shown, the characterization subsystem 102 may include a beam splitter oriented so that a common objective lens simultaneously directs the irradiation beam 114 to the sample 104 and captures the light to be collected 126.

[0088] Figure 1C is a simplified schematic diagram of a characterization subsystem 102 configured as an X-ray characterization subsystem 102 according to one or more embodiments of the present disclosure. Such a characterization subsystem 102 may, but is not limited to, a small-angle X-ray scattermeter (SAXR) or a soft X-ray reflectometer (SXR). X-ray characterization systems and related measurement techniques are generally described in U.S. Patent No. 7,929,667, issued April 11, 2011; U.S. Patent No. 9,885,962, issued February 6, 2018; U.S. Patent No. 1,001,3518, issued July 3, 2018; U.S. Patent No. 1,032,4050, issued June 18, 2019; U.S. Patent No. 1,035,2695, issued July 16, 2019; U.S. Patent No. 1,077,5323, issued September 15, 2020; Germer et al., "Intercomparison between optical and x-ray scatterometry measurements of FinFET structures," Proc. SPIE, v. 8681, p. 86810Q (2013); Kline et al., "X-ray scattering critical dimensional metrology using a compact x-ray source for next generation semiconductor devices," Journal of This is described in Micro / Nanolithography, MEMS, and MOEMS 16.1 (2017), U.S. Patent No. 11333621 issued on 17 May 2022, and U.S. Patent Application Publication No. 2021 / 0207956 published on 8 July 2021, all of which are incorporated herein by reference in their entirety.

[0089] In some embodiments, the irradiation source 112 is an X-ray source configured to generate an X-ray irradiation beam 114 having an arbitrary particle energy (e.g., soft X-rays, hard X-rays, etc.). The characterization subsystem 102 may also include any combination of components suitable for capturing an associated collection signal 134, which may include, but is not limited to, X-ray emission, optical emission, or particle emission.

[0090] For example, the characterization subsystem 102 may include an X-ray irradiation lens 116 suitable for collimating or focusing the X-ray irradiation beam 114, and a collection path lens (not shown) suitable for collecting, collimating, and / or focusing the collected signal 134 from the sample 104. Furthermore, the characterization subsystem 102 may include, but not limited to, various irradiation path optical instruments (not shown) and / or collection path optical instruments (not shown), such as X-ray collimating mirrors, grazing-entry elliptical mirrors, multi-capillary optical instruments such as hollow capillary X-ray waveguides, multilayer optical instruments, or systems, or any combination thereof. In embodiments, the characterization subsystem 102 includes an X-ray detector 128, such as, but not limited to, an X-ray monochromator (e.g., a crystalline monochromator such as a Loxley-Tanner-Bowen monochromator), an X-ray aperture, an X-ray beam diaphragm, or a diffractive optical instrument (e.g., a zone plate).

[0091] Figure 1D is a simplified schematic diagram of a characterization subsystem 102 configured as a particle beam characterization subsystem 102 according to one or more embodiments of the present disclosure.

[0092] In one embodiment, the irradiation source 112 includes a particle source (e.g., an electron source, an ion source, etc.) such that the irradiation beam 114 includes a particle beam (e.g., an electron beam, a particle beam, etc.). The irradiation source 112 may include any particle source known in the art that is suitable for generating the particle irradiation beam 114. For example, the irradiation source 112 may include, but is not limited to, an electron gun or an ion gun. In another embodiment, the irradiation source 112 is configured to provide a particle beam having variable energy. For example, the irradiation source 112 may provide an acceleration voltage in the range of 0.1 kilovolts (kV) to 30 kV, but is not limited to that. As another example, the irradiation source 112 including an ion source may provide an ion beam having an energy in the range of 1 kiloelectron volt (keV) to 50 keV, but is not required.

[0093] In another embodiment, the irradiation path 118 includes one or more particle focusing elements (e.g., an irradiation lens 116, a collecting lens 124, etc.). For example, the one or more particle focusing elements may include, but are not limited to, individual particle focusing elements or one or more particle focusing elements forming a composite system. In another embodiment, the one or more particle focusing elements include an irradiation lens 116 configured to guide the particle irradiation beam 114 to the sample 104. Furthermore, the one or more particle focusing elements may include, but are not limited to, electrostatic lenses, magnetic lenses, uniform potential lenses, or dual potential lenses, as well as any type of electron lens known in the art.

[0094] In another embodiment, the characterization subsystem 102 includes one or more particle detectors 128 for imaging or otherwise detecting particles emitted from the sample 104. For example, the detectors 128 may include electron collectors (e.g., secondary electron collectors, backscatter electron detectors, etc.). In another example, the detectors 128 may include photon detectors (e.g., photodetectors, X-ray detectors, scintillate elements coupled to photomultiplier tube (PMT) detectors, etc.) for detecting electrons and / or photons from the sample surface.

[0095] The subject matter described herein may illustrate different components that are contained within or linked to other components. Such depicted structures are merely illustrative, and it should be understood that other structures that actually achieve the same function can be implemented. Conceptually, any arrangement of components that achieve the same function is effectively “related” in such a way that the desired function is achieved. Thus, any two components described herein combined to achieve a particular function can be seen as “related” to each other, regardless of structure or intermediate components, in such a way that the desired function is achieved. Similarly, any two such related components can be seen as “linked” or “joined” to each other in such a way that the desired function is achieved, and any two components that can be so related can be seen as “joinable” to each other in such a way that the desired function is achieved. Specific examples of joinability include, but are not limited to, components that are physically interactable and / or interacting physically, as well as components that are wirelessly interactable and / or interacting wirelessly, and / or components that are logically interactable and / or interacting logically.

[0096] Many of the present disclosure and its associated advantages are to be understood from the above description, and it is clear that various modifications to the shape, structure, and arrangement of the components can be made without departing from the disclosed subject matter or sacrificing any of its essential advantages. The described shapes are for illustrative purposes only, and it is the intent of the following claims to encompass and include such modifications. Furthermore, it should be understood that the present invention is defined by the appended claims.

Claims

1. A controller including one or more processors configured to execute program instructions, Receiving optical measurement data for one or more training samples, including an array-under-array CMOS (CuA) device containing a complementary metal-oxide-semiconductor (CMOS) structure disposed beneath a periodic memory array structure, To create a first measurement model for determining one or more measurements of the CuA device based on the optical measurement data relating to one or more training samples, the first measurement model incorporating a model of photomatter interaction with the CuA device, wherein the CMOS structure is modeled as a CMOS effective medium within the first measurement model, and the first measurement model is parameterized by geometric parameters of the memory array structure, variability parameters of the memory array structure, geometric parameters of the CMOS effective medium, and variability parameters of the CMOS effective medium. Receiving reference data relating to one or more training samples, Determining the values ​​of the geometric parameters of the memory array structure based on the aforementioned reference data, The value of the variation parameter of the CMOS effective medium is determined by fixing the value of the geometric parameter of the memory array structure with the optical measurement data relating to one or more training samples as input. Updating the first measurement model to a second measurement model that includes the values ​​of the variation parameters of the CMOS effective medium, To receive optical measurement data for one or more test samples including a CuA device, and, A system characterized in that, based on the second measurement model, it generates one or more measurement method measurement values ​​of the CuA device on the one or more test samples using the optical measurement data for the one or more test samples, thereby causing one or more processors to execute a measurement recipe.

2. A system according to claim 1, wherein the second measurement model includes the first measurement model having the variation parameter of the CMOS effective medium, which is fixed by the determined value of the geometric parameter.

3. The system according to claim 1, wherein the CMOS effective medium in the second measurement model includes a combination of determined values ​​of the geometric parameters derived from multiple locations of one or more training samples, based on Braggman's Effective Medium Approximation (BEMA) model.

4. The system according to claim 1, wherein the reference data relating to one or more training samples is Includes at least one of the following: transmission electron microscope (TEM) data, scanning electron microscope (SEM) data, SAXS data, X-ray photoelectron spectroscopy (XPS) data, or X-ray diffraction (XRD) data. A system characterized by the following features.

5. The system according to claim 1, wherein the reference data relating to one or more training samples is Includes critical dimension X-ray small-angle spectroscopy (CD-SAXS) data. A system characterized by the following features.

6. The system according to claim 1, wherein the one or more processors are To generate synthetic measurement data using the second measurement model having a wide range of parameters describing one or more synthetic CuA devices, and Training a machine learning model to generate one or more measurements based on training data that includes at least one of the following: optical measurement data relating to one or more training samples, reference data relating to one or more training samples, or synthetic data describing one or more synthetic CuA devices. The system is further configured to execute the program instructions that cause one or more processors to carry out the measurement recipe.

7. The system according to claim 6, wherein the one or more processors are To generate reference data for one or more synthetic CuA devices that are further included in the training data. The system is further configured to execute the program instructions that cause one or more processors to carry out the measurement recipe.

8. The system according to claim 7, wherein the reference data relating to one or more synthetic CuA devices is Includes at least one of the following: transmission electron microscope (TEM) data, scanning electron microscope (SEM) data, SAXS data, X-ray photoelectron spectroscopy (XPS) data, or X-ray diffraction (XRD) data. A system characterized by the following features.

9. The system according to claim 7, wherein the reference data relating to one or more synthetic CuA devices is Critical dimension X-ray small-angle spectroscopy (CD-SAXS) data A system characterized by including

10. The system according to claim 6, To generate optical measurement data for one or more additional test samples, including one or more CuA devices, and Based on the machine learning model, generate one or more measurement method values ​​for one or more CuA devices on one or more additional test samples using the optical measurement data for one or more additional test samples. A system characterized by further including the following.

11. The system according to claim 1, wherein the one or more measured values ​​are Measurement Method Measurement Value A system characterized by including

12. The system according to claim 11, wherein the measurement method value is At least one of the following: superposition measurement, critical dimension (CD) measurement, shape measurement, stress measurement, composition measurement, bandgap measurement, electrical property measurement, or process state measurement. A system characterized by including

13. The system according to claim 1, wherein the one or more measured values ​​are Test measurement values A system characterized by including

14. The system according to claim 13, wherein the inspection measurement value is Identifying or classifying defects in the second structure, at least one of the above. A system characterized by including

15. The system according to claim 1, wherein the optical measurement data relating to one or more training samples and one or more test samples is At least one of the following: polarization analysis data, reflected light analysis data, or scattered light analysis data. A system characterized by including

16. A system according to claim 1, characterized in that at least one of the first measurement model or the second measurement model uses at least one of the following methods: exact coupled wave analysis (RCWA), finite element method (FEM), method of moment, area integral analysis, volume integral analysis, or finite difference time-domain analysis (FDTD).

17. Optical properties evaluation system, Reference characteristic evaluation system, A controller that is communicably coupled to the optical characterization system and the reference characterization system. The controller includes one or more processors configured to execute program instructions, and the program instructions are Receiving optical measurement data from the optical characterization system for one or more training samples, including an under-array CMOS (CuA) device containing a complementary metal-oxide-semiconductor (CMOS) structure disposed below a periodic memory array structure, To create a first measurement model for determining one or more measurements of the CuA device based on the optical measurement data relating to one or more training samples, the first measurement model incorporating a model of photomatter interaction with the CuA device, the CMOS structure being modeled as a CMOS effective medium within the first measurement model, and the first measurement model being parameterized by geometric parameters of the memory array structure, variability parameters of the memory array structure, geometric parameters of the CMOS effective medium, and variability parameters of the CMOS effective medium. Receiving reference data relating to one or more training samples from the reference characteristic evaluation system, Determining the values ​​of the geometric parameters of the memory array structure based on the aforementioned reference data, The value of the variation parameter of the CMOS effective medium is determined by fixing the value of the geometric parameter of the memory array structure with the optical measurement data relating to one or more training samples as input. Updating the first measurement model to a second measurement model that includes the values ​​of the variation parameters of the CMOS effective medium, Receiving optical measurement data for one or more test samples including a CuA device from the optical characterization system, and Based on the second measurement model, one or more measurement method values ​​of the CuA device on the one or more test samples are generated using the optical measurement data for the one or more test samples. A system characterized by causing one or more processors to execute a measurement recipe.

18. The system according to claim 17, wherein the optical characteristic evaluation system is At least one of the following: ellipsometer, reflectometer, or scattermeter A system characterized by including

19. The system according to claim 17, wherein the reference specific evaluation system is At least one of the following: an X-ray characterization system or a particle beam characterization system. A system characterized by including

20. To generate optical measurement data for one or more training samples, including an array-under-array CMOS (CuA) device containing a complementary metal-oxide-semiconductor (CMOS) structure disposed beneath a periodic memory array structure. To create a first measurement model for determining one or more measurements of the CuA device based on the optical measurement data relating to one or more training samples, the first measurement model incorporating a model of photomatter interaction with the CuA device, wherein the CMOS structure is modeled as a CMOS effective medium within the first measurement model, and the first measurement model is parameterized by geometric parameters of the memory array structure, variability parameters of the memory array structure, geometric parameters of the CMOS effective medium, and variability parameters of the CMOS effective medium. To generate reference data relating to one or more training samples, Determining the values ​​of the geometric parameters of the memory array structure based on the aforementioned reference data, The value of the variation parameter of the CMOS effective medium is determined by fixing the value of the geometric parameter of the memory array structure with the optical measurement data relating to one or more training samples as input. Updating the first measurement model to a second measurement model that includes the values ​​of the variation parameters of the CMOS effective medium, To generate optical measurement data for one or more test samples including a CuA device, and, Based on the second measurement model, one or more measurement method values ​​of the CuA device on the one or more test samples are generated using the optical measurement data for the one or more test samples. A method characterized by including the following.

21. A method according to claim 20, characterized in that the second measurement model includes the first measurement model having the variation parameter of the CMOS effective medium, which is fixed by the determined value of the geometric parameter.

22. A method according to claim 20, characterized in that the CMOS effective medium in the second measurement model includes a combination of determined values ​​of the geometric parameters derived from multiple locations of one or more training samples, based on Braggman's Effective Medium Approximation (BEMA) model.

23. A method according to claim 20, wherein the reference data relating to one or more training samples is The data includes at least one of the following: transmission electron microscope (TEM) data, scanning electron microscope (SEM) data, small-angle X-ray spectroscopy (SAXS) data, X-ray photoelectron spectroscopy (XPS) data, or X-ray diffraction (XRD) data. A method characterized by the following:

24. A method according to claim 20, wherein the reference data relating to one or more training samples is Includes critical dimension X-ray small-angle spectroscopy (CD-SAXS) data. A method characterized by the following:

25. The method according to claim 20, To generate synthetic measurement data using the second measurement model having a wide range of parameters describing one or more synthetic CuA devices, and Training a machine learning model to generate one or more measurements based on training data that includes at least one of the following: optical measurement data relating to one or more training samples, reference data relating to one or more training samples, or synthetic data describing one or more synthetic CuA devices. A method characterized by further comprising the following.

26. The method according to claim 25, To generate reference data relating to one or more synthetic CuA devices, which are further included in the aforementioned training data. A method characterized by further comprising the following.

27. A method according to claim 26, wherein the reference data relating to one or more synthetic CuA devices is Includes at least one of the following: transmission electron microscope (TEM) data, scanning electron microscope (SEM) data, SAXS data, X-ray photoelectron spectroscopy (XPS) data, or X-ray diffraction (XRD) data. A method characterized by the following:

28. A method according to claim 26, wherein the reference data relating to one or more synthetic CuA devices is Critical dimension X-ray small-angle spectroscopy (CD-SAXS) data A method characterized by including the following.

29. The method according to claim 25, To generate optical measurement data for one or more additional test samples, including one or more CuA devices, and Based on the machine learning model, generate one or more measurement method values ​​for one or more CuA devices on one or more additional test samples using the optical measurement data for one or more additional test samples. A method characterized by further comprising the following.

30. The method according to claim 20, wherein the one or more measured values ​​are Measurement Method Measurement Value A method characterized by including the following.

31. The method according to claim 30, wherein the measurement method measurement value is At least one of the following: superposition measurement, critical dimension (CD) measurement, shape measurement, stress measurement, composition measurement, bandgap measurement, electrical property measurement, or process state measurement. A method characterized by including the following.

32. The method according to claim 20, wherein the one or more measured values ​​are Test measurement values A method characterized by including the following.

33. The method according to claim 32, wherein the inspection measurement value is Identifying or classifying defects in the second structure, at least one of the above. A method characterized by including the following.

34. The method according to claim 20, wherein the optical measurement data relating to one or more training samples and one or more test samples is At least one of the following: polarization analysis data, reflected light analysis data, or scattered light analysis data. A method characterized by including the following.

35. A method according to claim 20, characterized in that at least one of the first measurement model or the second measurement model uses at least one of the following: exact coupled wave analysis (RCWA), finite element method (FEM), moment method, area integral analysis, volume integral analysis, or finite difference time-domain analysis (FDTD).

Images