Abstract:
Method, metrology modules and RCA tool are provided, which use the behavior of resonance region(s) in measurement landscapes to evaluate and characterize process variation with respect to symmetric and asymmetric factors, and provide root cause analysis of the process variation with respect to process steps. Simulations of modeled stacks with different layer thicknesses and process variation factors may be used to enhance the analysis and provide improved target designs, improved algorithms and correctables for metrology measurements. Specific targets that exhibit sensitive resonance regions may be utilize to enhance the evaluation of process variation.
Abstract:
A metrology performance analysis system includes a metrology tool including one or more detectors and a controller communicatively coupled to the one or more detectors. The controller is configured to receive one or more metrology data sets associated with a metrology target from the metrology tool in which the one or more metrology data sets include one or more measured metrology metrics and the one or more measured metrology metrics indicate deviations from nominal values. The controller is further configured to determine relationships between the deviations from the nominal values and one or more selected semiconductor process variations, and determine one or more root causes of the deviations from the nominal values based on the relationships between values of the one or more metrology metrics and the one or more selected semiconductor process variations.
Abstract:
Metrology methods and targets are provided, for estimating inter-cell process variation by deriving, from overlay measurements of at least three target cells having different designed misalignments, a dependency of a measured inaccuracy on the designed misalignments (each designed misalignment is between at least two overlapping periodic structures in the respective target cell). Inaccuracies which are related to the designed misalignments are reduced, process variation sources are detected and targets and measurement algorithms are optimized according to the derived dependency.