Abstract: A metrology system is disclosed. In one embodiment, the metrology system includes a controller communicatively coupled to a reference metrology tool and an optical metrology tool, the controller including one or more processors configured to: generate a geometric model for determining a profile of a test HAR structure from metrology data from a reference metrology tool; generate a material model for determining one or more material parameters of a test HAR structure from metrology data from the optical metrology tool; form a composite model from the geometric model and the material model; measure at least one additional test HAR structure with the optical metrology tool; and determine a profile of the at least one additional test HAR structure based on the composite model and metrology data from the optical metrology tool associated with the at least one HAR test structure.

Abstract: Methods and systems for calibrating metrology tool offset values to match measurement results across a fleet of metrology tools are presented herein. The calibration of offset values is based on measurements of inline, production wafers and does not require the use of specially fabricated and characterized quality control (QC) wafers. In this manner, the entire process flow to calibrate metrology tool offset values is automated and fully integrated within a high volume semiconductor fabrication process flow. In a further aspect, the implementation of a new offset value is regulated by one or more predetermined control limit values. In another further aspect, the measured values of a parameter of interest are adjusted to compensate for the effects of measurement time on the wafer under measurement.

Abstract: Automatic wavelength or angle pruning for optical metrology is described. An embodiment of a method for automatic wavelength or angle pruning for optical metrology includes determining a model of a structure including a plurality of parameters; designing and computing a dataset of wavelength-dependent or angle-dependent data for the model; storing the dataset in a computer memory; performing with a processor an analysis of the dataset for the model including applying an outlier detection technology on the dataset, and identifying any data outliers, each data outlier being a wavelength or angle; and, if any data outliers are identified in the analysis of the dataset of the model, removing the wavelengths or angles corresponding to the data outliers from the dataset to generate a modified dataset, and storing the modified dataset in the computer memory.

Abstract: Embodiments include automatic selection of sample values for optical metrology. An embodiment of a method includes providing a library parameter space for modeling of a diffracting structure using an optical metrology system; automatically determining by a processing unit a reduced sampling set from the library parameter space, wherein the reduced space is based on one or both of the following recommending a sampling shape based on an expected sample space usage, or recommending a sampling filter based on correlation between two or more parameters of the library parameter space; and generating a library for the optical metrology system using the reduced sampling set.

Abstract: Methods and systems for optimizing a set of measurement library control parameters for a particular metrology application are presented herein. Measurement signals are collected from one or more metrology targets by a target measurement system. Values of user selected parameters of interest are resolved by fitting a pre-computed measurement library function to the measurement signals for a given set of library control parameters. Values of one or more library control parameters are optimized such that differences between the values of the parameters of interest estimated by the library based measurement and reference values associated with trusted measurements of the parameters of interest are minimized. The optimization of the library control parameter values is performed without recalculating the pre-computed measurement library.

Abstract: A metrology system is disclosed. In one embodiment, the metrology system includes a controller communicatively coupled to a reference metrology tool and an optical metrology tool, the controller including one or more processors configured to: generate a geometric model for determining a profile of a test HAR structure from metrology data from a reference metrology tool; generate a material model for determining one or more material parameters of a test HAR structure from metrology data from the optical metrology tool; form a composite model from the geometric model and the material model; measure at least one additional test HAR structure with the optical metrology tool; and determine a profile of the at least one additional test HAR structure based on the composite model and metrology data from the optical metrology tool associated with the at least one HAR test structure.

Abstract: Methods and systems for evaluating and ranking the measurement efficacy of multiple sets of measurement system combinations and recipes for a particular metrology application are presented herein. Measurement efficacy is based on estimates of measurement precision, measurement accuracy, correlation to a reference measurement, measurement time, or any combination thereof. The automated the selection of measurement system combinations and recipes reduces time to measurement and improves measurement results. Measurement efficacy is quantified by a set of measurement performance metrics associated with each measurement system and recipe. In one example, the sets of measurement system combinations and recipes most capable of measuring the desired parameter of interest are presented to the user in rank order based on corresponding values of one or more measurement performance metrics. A user is able to select the appropriate measurement system combination in an objective, quantitative manner.

Abstract: Dynamic removal of correlation of highly-correlated parameters for optical metrology is described. An embodiment of a method includes determining a model of a structure, the model including a set of parameters; performing optical metrology measurement of the structure, including collecting spectra data on a hardware element; during the measurement of the structure, dynamically removing correlation of two or more parameters of the set of parameters, an iteration of the dynamic removal of correlation including: generating a Jacobian matrix of the set of parameters, applying a singular value decomposition of the Jacobian matrix, selecting a subset of the set of parameters, and computing a direction of the parameter search based on the subset of parameters. If the model does not converge, performing one or more additional iterations of the dynamic removal of correlation until the model converges; and if the model does converge, reporting the results of the measurement.

Abstract: Methods and systems for optimizing a set of measurement library control parameters for a particular metrology application are presented herein. Measurement signals are collected from one or more metrology targets by a target measurement system. Values of user selected parameters of interest are resolved by fitting a pre-computed measurement library function to the measurement signals for a given set of library control parameters. Values of one or more library control parameters are optimized such that differences between the values of the parameters of interest estimated by the library based measurement and reference values associated with trusted measurements of the parameters of interest are minimized. The optimization of the library control parameter values is performed without recalculating the pre-computed measurement library.

Abstract: Model optimization approaches based on spectral sensitivity is described. For example, a method includes determining a first model of a structure. The first model is based on a first set of parameters. A set of spectral sensitivity variations data is determined for the structure. Spectral sensitivity is determined by derivatives of the spectra with respect to the first set of parameters. The first model of the structure is modified to provide a second model of the structure based on the set of spectral sensitivity variations data. The second model of the structure is based on a second set of parameters different from the first set of parameters. A simulated spectrum derived from the second model of the structure is then provided.

Abstract: Embodiments are generally directed to neural network training for library-based critical dimension metrology. An embodiment of a method includes optimizing a threshold for a principal component analysis of a spectrum data set to provide a principal component value, estimating a training target for one or more neural networks, training the one or more neural networks based both on the training target and on the principal component value provided from optimizing the threshold for the principal component analysis, and providing a spectral library based on the one or more trained neural networks.

Abstract: Methods and systems for evaluating and ranking the measurement efficacy of multiple sets of measurement system combinations and recipes for a particular metrology application are presented herein. Measurement efficacy is based on estimates of measurement precision, measurement accuracy, correlation to a reference measurement, measurement time, or any combination thereof. The automated the selection of measurement system combinations and recipes reduces time to measurement and improves measurement results. Measurement efficacy is quantified by a set of measurement performance metrics associated with each measurement system and recipe. In one example, the sets of measurement system combinations and recipes most capable of measuring the desired parameter of interest are presented to the user in rank order based on corresponding values of one or more measurement performance metrics. A user is able to select the appropriate measurement system combination in an objective, quantitative manner.

Abstract: Methods and tools for generating measurement models of complex device structures based on re-useable, parametric models are presented. Metrology systems employing these models are configured to measure structural and material characteristics associated with different semiconductor fabrication processes. The re-useable, parametric sub-structure model is fully defined by a set of independent parameters entered by a user of the model building tool. All other variables associated with the model shape and internal constraints among constituent geometric elements are pre-defined within the model. In some embodiments, one or more re-useable, parametric models are integrated into a measurement model of a complex semiconductor device. In another aspect, a model building tool generates a re-useable, parametric sub-structure model based on input from a user.

Abstract: Methods and systems for determining a meta-model to correct model based measurements are presented. Such systems are employed to measure structural and material characteristics (e.g., material composition, dimensional characteristics of structures and films, etc.) associated with different semiconductor fabrication processes. In one aspect, model-based measurement parameter values are corrected based on a meta-model that maps specimen parameter values determined based on the measurement model to reference parameter values determined based on a more accurate reference measurement. In another aspect, parameters of a meta-model are determined such that errors between reference parameter values and specimen parameter values determined based on the measurement model are minimized. In some embodiments, the accuracy of a corrected parameter value is an order of magnitude greater than the uncorrected parameter value.

Abstract: Methods and tools for generating measurement models of complex device structures based on re-useable, parametric models are presented. Metrology systems employing these models are configured to measure structural and material characteristics associated with different semiconductor fabrication processes. The re-useable, parametric sub-structure model is fully defined by a set of independent parameters entered by a user of the model building tool. All other variables associated with the model shape and internal constraints among constituent geometric elements are pre-defined within the model. In some embodiments, one or more re-useable, parametric models are integrated into a measurement model of a complex semiconductor device. In another aspect, a model building tool generates a re-useable, parametric sub-structure model based on input from a user.

Abstract: Embodiments include automatic selection of sample values for optical metrology. An embodiment of a method includes providing a library parameter space for modeling of a diffracting structure using an optical metrology system; automatically determining by a processing unit a reduced sampling set from the library parameter space, wherein the reduced space is based on one or both of the following recommending a sampling shape based on an expected sample space usage, or recommending a sampling filter based on correlation between two or more parameters of the library parameter space; and generating a library for the optical metrology system using the reduced sampling set.

Abstract: Dynamic removal of correlation of highly-correlated parameters for optical metrology is described. An embodiment of a method includes determining a model of a structure, the model including a set of parameters; performing optical metrology measurement of the structure, including collecting spectra data on a hardware element; during the measurement of the structure, dynamically removing correlation of two or more parameters of the set of parameters, an iteration of the dynamic removal of correlation including: generating a Jacobian matrix of the set of parameters, applying a singular value decomposition of the Jacobian matrix, selecting a subset of the set of parameters, and computing a direction of the parameter search based on the subset of parameters. If the model does not converge, performing one or more additional iterations of the dynamic removal of correlation until the model converges; and if the model does converge, reporting the results of the measurement.

Abstract: Automatic wavelength or angle pruning for optical metrology is described. An embodiment of a method for automatic wavelength or angle pruning for optical metrology includes determining a model of a structure including a plurality of parameters; designing and computing a dataset of wavelength-dependent or angle-dependent data for the model; storing the dataset in a computer memory; performing with a processor an analysis of the dataset for the model including applying an outlier detection technology on the dataset, and identifying any data outliers, each data outlier being a wavelength or angle; and, if any data outliers are identified in the analysis of the dataset of the model, removing the wavelengths or angles corresponding to the data outliers from the dataset to generate a modified dataset, and storing the modified dataset in the computer memory.

Abstract: Approaches for accurate neural network training for library-based critical dimension (CD) metrology are described. Approaches for fast neural network training for library-based CD metrology are also described.

Abstract: Approaches for accurate neural network training for library-based critical dimension (CD) metrology are described. Approaches for fast neural network training for library-based CD metrology are also described. In an example, a method includes optimizing a threshold for a principal component analysis (PCA) of a spectrum data set to provide a principal component (PC) value, estimating a training target for one or more neural networks, training the one or more neural networks based both on the training target and on the PC value provided from optimizing the threshold for the PCA, and providing a spectral library based on the one or more trained neural networks.