Kla Corporation

Shiming Wei

Michael E Adel

Mark Ghinovker

Weihua Li

Inna Tarshish-Shapir

Metrology target design methods and verification targets are provided. Methods comprise using OCD data related to designed metrology target(s) as an estimation of a discrepancy between a target model and a corresponding actual target on a wafer, and adjusting a metrology target design model to compensate for the estimated discrepancy. The dedicated verification targets may comprise overlay target features and be size optimized to be measureable by an OCD sensor, to enable compensation for inaccuracies resulting from production process variation. Methods also comprise modifications to workflows between manufacturers and metrology vendors which provide enable higher fidelity metrology target design models and ultimately higher accuracy of metrology measurements.

Verification metrology targets and their design

An overlay metrology system is disclosed. The overlay metrology system includes a controller configured to be communicatively coupled with an overlay metrology subsystem. The controller receives overlay measurements from the overlay metrology subsystem and generates one or more quality metrics. The controller extracts a set of principle components from the one or more quality metrics. The controller generates input data and inputs the input data into an input matrix of a supervised machine learning algorithm to train a predictive model. The controller then identifies a recipe or hardware configuration with a minimum residual value.

Growing community of inventors

Shanghai, China

Shiming Wei