Abstract: A diffraction pattern guided source mask optimization (SMO) method that includes determining a source variable region from a diffraction pattern. The source variable region corresponds to one or more areas of a diffraction pattern in a pupil for which one or more pupil variables are to be adjusted. The source variable region in the diffraction pattern includes a plurality of pixels in an image of a selected region of interest in the diffraction pattern. Determining the source variable region can include binarization of the plurality of pixels in the image such that individual pixels are either included in the source variable region or excluded from the source variable region. The method can include adjusting the one or more pupil variables for the one or more areas of the pupil that correspond to the source variable region; and rendering a final pupil based on the adjusted one or more pupil variables.

Abstract: A diffraction pattern guided source mask optimization (SMO) method that includes determining a source variable region from a diffraction pattern. The source variable region corresponds to one or more areas of a diffraction pattern in a pupil for which one or more pupil variables are to be adjusted. The source variable region in the diffraction pattern includes a plurality of pixels in an image of a selected region of interest in the diffraction pattern. Determining the source variable region can include binarization of the plurality of pixels in the image such that individual pixels are either included in the source variable region or excluded from the source variable region. The method can include adjusting the one or more pupil variables for the one or more areas of the pupil that correspond to the source variable region; and rendering a final pupil based on the adjusted one or more pupil variables.

Abstract: Methods of training machine learning models related to a patterning process, including a method for training a machine learning model configured to predict a mask pattern. The method including obtaining (i) a process model of a patterning process configured to predict a pattern on a substrate, wherein the process model comprises one or more trained machine learning models, and (ii) a target pattern, and training the machine learning model configured to predict a mask pattern based on the process model and a cost function that determines a difference between the predicted pattern and the target pattern.