A method for determining substrate deformation includes obtaining first measurement data associated with mark positions, from measurements of a plurality of substrates; obtaining second measurement data associated with mark positions, from measurements of the plurality of substrates; determining a mapping between the first measurement data and the second measurement data; and decomposing the mapping, by calculating an eigenvalue decomposition for the mapping, to separately determine a first deformation (e.g. mark deformation) that scales differently from a second deformation (e.g. substrate deformation) in the mapping between the data. The steps of determining a mapping and decomposing the mapping may be performed together using non-linear optimization.

Corrections are calculated for use in controlling a lithographic apparatus. Using a metrology apparatus a performance parameter is measured at sampling locations across one or more substrates to which a lithographic process has previously been applied. A process model is fitted to the measured performance parameter, and an up-sampled estimate is provided for process-induced effects across the substrate. Corrections are calculated for use in controlling the lithographic apparatus, using an actuation model and based at least in part on the fitted process model. For locations where measurement data is available, this is added to the estimate to replace the process model values. Thus, calculation of actuation corrections is based on a modified estimate which is a combination of values estimated by the process model and partly on real measurement data.

The invention provides a position sensor (300) which comprises an optical system (305, 306) configured to provide measurement radiation (304) to a substrate (307). The optical system is arranged to receive at least a portion of radiation (309) diffracted by a mark (308) provided on the substrate. A processor (313) is applied to derive at least one position-sensitive signal (312) from the received radiation. The measurement radiation comprises at least a first and a second selected radiation wavelength. The selection of the at least first and second radiation wavelengths is based on a position error swing-curve model.

A substrate processing apparatus is provided. The apparatus includes an imaging unit that images a mark on a substrate, and a processor that aligns the substrate based on an image of the mark obtained by the imaging unit. If the alignment has failed, the processor identifies a factor of the failure based on information including the image and executes at least one of a plurality of recovery processes based on the identified factor. The processor includes an output unit that outputs a condition for the at least one of recovery processes in accordance with an inference model, and a learning unit that learns the inference model based on an execution result of the at least one of the recovery processes under the condition output from the output unit.

The present disclosure generally relates to photolithography systems, and methods for correcting positional errors in photolithography systems. When a photolithography system is first started, the system enters a stabilization period. During the stabilization period, positional readings and data, such as temperature, pressure, and humidity data, are collected as the system prints or exposes a substrate. A model is created based on the collected data and the positional readings. The model is then used to estimate errors in subsequent stabilization periods, and the estimated errors are dynamically corrected during the subsequent stabilization periods.

A position detection apparatus configured to detect a pattern including a plurality of pattern elements formed on an object includes a control unit configured to detect the pattern by performing pattern matching between a template including a plurality of feature points and the plurality of pattern elements. While, performing pattern matching, the control unit changes positions of the plurality of feature points such that a correlation between an image and the template is within a predetermined allowable range.

Embodiments described herein relate to a system, methods, and non-transitory computer-readable mediums that accurately align subsequent patterned layers in a photoresist utilizing a deep learning model and utilizing device patterns to replace alignment marks in lithography processes. The deep learning model is trained to recognize unique device patterns called alignment patterns in the FOV of the camera. Cameras in the lithography system capture images of the alignment patterns. The deep learning model finds the alignment patterns in the field of view of the cameras. An ideal image generated from a design file is matched with the camera with respect to the center of the field of view of the camera. A shift model and a rotation model are output from the deep learning model to create an alignment model. The alignment model is applied to the currently printing layer.

Combination of a stage and a level sensor configured to sense a height level at a target location on an object is described, the stage comprising an object table configured to hold the object and a positioning device for displacing the object table relative to the level sensor in a first direction, the level sensor comprising a projection system configured to project a measurement beam onto a measurement area of the object, the measurement area having a measurement area length in the first direction, a detector system configured to receive different portions of the measurement beam after being reflected off different sub-areas within the measurement area, the different sub-areas being arranged in the first direction, and to supply output signals representative of the different portions received, a signal processing system configured to process the output signals from the detector system.

A method of determining positions of marks, the marks comprising periodic structures, at least some of the structures comprising periodic sub-structures, the sub-structures having a smaller period than the structures, the marks formed with positional offsets between the sub-structures and structures, the positional offsets caused by a combination of both known and unknown components, the method comprising illuminating a plurality of the marks with radiation having different characteristics, detecting radiation diffracted by the marks using one or more detectors which produce output signals, discriminating between constituent parts of the signals, the discriminating based on a variation of the signals as a function of spatial positions of the marks on a substrate, selecting at least one of the constituent parts of the signals, and using the at least one selected constituent part, and information relating to differences between the known components, to calculate a corrected position of at least one mark.

A method of applying a measurement correction includes calculating a first correction value based on a first coefficient and the measurement; calculating a second correction value based on a second coefficient, greater than the first coefficient, and the measurement; and calculating a third correction value based on a third coefficient, greater than the second coefficient, and the measurement. The method also includes applying the third correction value to the measurement if a difference between the first correction value and the third correction value is above a first threshold value; applying the second correction value to the measurement if a difference between the first correction value and the second correction value is above a second threshold value; and applying the first correction value to the measurement if the difference between the first correction value and the second correction value is below or equal to the second threshold value.