公开(公告)号：US20250067691A1

公开(公告)日：2025-02-27

申请号：US18725804

申请日：2022-12-30

Applicant: NOVA MEASURING INSTRUMENTS INC.

Inventor： Heath POIS ,  Dmitry KISLITSYN ,  Mark KLARE ,  Paul ISBESTER ,  Daniel Kandel ,  Michal Haim YACHINI

IPC: G01N23/223 ,  G01N23/2273

Abstract: Determining process excursions in a semiconductor processing using unsupervised machine learning on photoelectron emission dataset obtained by XPS or XRF tool. Principal component analysis is applied to the emission dataset and the variances of each principal component is analyzed to thereby select a number of N principal components whose variance is the highest. All data points of the dataset which do not correspond to any of the N principal components are removed from the dataset to obtain a filtered dataset. An emission intensity is then calculated from the filtered dataset and is plotted on a SPC chart to inspect for excursions.