GLOBAL LANDMARK METHOD FOR CRITICAL DIMENSION UNIFORMITY RECONSTRUCTION

摘要

Data associated with a substrate can be processed by measuring a property of at least a first type of specific features and a second type of specific features on a substrate. The first type of specific features is measured at a first plurality of locations on the substrate to generate a first group of measured values, and the second type of specific features is measured at a second plurality of locations on the substrate to generate a second group of measured values, in which the first and second groups of measured values are influenced by critical dimension variations of the substrate. A combined measurement function is defined based on combining the at least first and second groups of measured values. At least one group of measured values is transformed prior to combining with another group or other groups of measured values, in which the transformation is defined by a group of coefficients. Variations in the critical dimension across the substrate are determined based on the combined measurement function and a predetermined relationship between the measured values and the critical dimension.