A method of automatic optimization of an examination recipe includes obtaining inspection data of a given layer of a semiconductor specimen acquired by an inspection tool during runtime examination, the inspection data including inspection images representative of defect candidates from a defect map of the given layer, extracting inspection features characterizing the inspection images, and using a classifier to classify the defect candidates based on the inspection features, giving rise to a list of defect candidates having a higher probability of being defects of interest (DOIs). The semiconductor specimen includes multiple layers, and the classifier is a general-purpose classifier (GPC) usable for runtime classification of inspection data from any layer of the multiple layers of the semiconductor specimen, the GPC being previously trained using training data including inspection features characterizing training inspection images of various types of DOIs and nuisances collected from the multiple layers and label data associated therewith.

A machine learning model may employ in situ chemical composition information, as an input, to characterize processes in real time, and optionally assist in process control. Chemical composition information may be obtained from an in situ emission spectrometer such an optical emission spectrometer.

A method of obtaining an array of a plurality of regions on a substrate, including obtaining position measurement data by measuring a mark assigned to each sample region among the plurality of regions on the substrate, and estimating a position of each non-measurement region excluding the sample region among the plurality of regions by using a regression model used to estimate the array from the position measurement data, wherein the regression model is a nonparametric regression model.

A method and associated computer program and apparatuses for determining a correction for at least one control parameter, the at least one control parameter for controlling a semiconductor manufacturing process so as to manufacture semiconductor devices on a substrate. The method includes: obtaining metrology data relating to the semiconductor manufacturing process or at least part thereof; obtaining associated data relating to the semiconductor manufacturing process or at least part thereof, the associated data providing information for interpreting the metrology data; and determining the correction based on the metrology data and the associated data, wherein the determining is such that the determined correction depends on a degree to which a trend and/or event in the metrology data should be corrected based on the interpretation of the metrology data.

A system and method are disclosed for generating metrology measurements with second sub-system such as an optical sub-system. The method may include performing a training and a run-time operation. The training may include receiving first metrology data for device features from the first metrology sub-system (e.g., optical); generating first metrology measurements (e.g., critical dimensions, etc.); binning the device features into two or more device bins based on the first metrology measurements; and identifying representative metrology targets for the two or more device bins based on distributions of the first metrology measurements. The run-time operation may include receiving run-time metrology data (e.g., optical) of the representative metrology targets; and generating run-time metrology measurements based on the run-time metrology data.

Disclosed is a method of determining a reliability metric describing a reliability of metrology signal and/or a parameter of interest value derived therefrom and associated apparatuses. The method comprises obtaining a trained inference model for inferring a value for a parameter of interest from a measurement signal and one or more measurement signals and/or respective one or more values of a parameter of interest derived therefrom using said trained inference model. At least one reliability metric value is determined for the one or more measurement signals and/or respective one or more values of a parameter of interest, the reliability metric describing a reliability of one or more measurement signals and/or respective one or more values of a parameter of interest with respect to an accurate prediction space associated with the trained inference model.

Proactive management of semiconductor substrate tools. A machine learning model is used to predict future performance characteristics for such tools. In some examples, the model can diagnose issues with tools or with ambient conditions of the tools' environment. In some examples, the model can recommend one or more remedial actions to maintain adequate performance of the substrate tool.

Methods and systems for measurements of semiconductor structures based on a trained whole wafer measurement model that is valid for all possible measurement locations on a wafer are described herein. A whole wafer measurement model is trained based on Design Of Experiments (DOE) measurement data collected across an entire wafer or set of wafers subjected to the same set of process steps. By employing DOE measurement data across an entire wafer or set of wafers, information about process behavior across the entire wafer is implicitly incorporated into the trained model at all locations across the wafer under measurement. The model training process encourages physical process behavior, which reduces the degrees of freedom of the underlying model, breaks correlations between parameters, and reduces the dimension of the solution space. As a result, measurement performance and robustness is improved.

A method, computer program and associated apparatuses for metrology. The method includes determining a reconstruction recipe describing at least nominal values for use in a reconstruction of a parameterization describing a target. The method includes obtaining first measurement data relating to measurements of a plurality of targets on at least one substrate, the measurement data relating to one or more acquisition settings and performing an optimization by minimizing a cost function which minimizes differences between the first measurement data and simulated measurement data based on a reconstructed parameterization for each of the plurality of targets. A constraint on the cost function is imposed based on a hierarchical prior. Also disclosed is a hybrid model method comprising obtaining a coarse model operable to provide simulated coarse data; and training a data driven model to correct the simulated coarse data so as to determine simulated data for use in reconstruction.

An inspection system includes a controller including a memory maintaining program instructions and one or more processors configured to execute the program instructions. The program instructions cause the one or more processors to generate a geometric model of a structure of a sample, generate an optical response function model of the structure of the sample to illumination based at least in part on the geometric model, receive measured data from a detector, generate a parametric sub-structure model based on at least the optical response function model and the measured data, and extract one or more parameters of the structure based on the measured data.