US 11,901,203 B2
Substrate process endpoint detection using machine learning
Pengyu Han, San Jose, CA (US); Lei Lian, Fremont, CA (US); Shu Yu Chen, Zhubei (TW); Todd Egan, Fremont, CA (US); Wan Hsueh Lai, New Taipei (TW); Chao-Hsien Lee, Taoyuan (TW); Pin Ham Lu, Taipei (TW); Zhengping Yao, Cupertino, CA (US); and Barry Craver, San Jose, CA (US)
Assigned to Applied Materials, Inc., Santa Clara, CA (US)
Filed by APPLIED MATERIALS, INC., Santa Clara, CA (US)
Filed on Jun. 10, 2021, as Appl. No. 17/344,787.
Prior Publication US 2022/0399215 A1, Dec. 15, 2022
Int. Cl. H01L 21/67 (2006.01); G01N 21/95 (2006.01); G06N 20/00 (2019.01); G05B 13/02 (2006.01)
CPC H01L 21/67253 (2013.01) [G01N 21/9501 (2013.01); G06N 20/00 (2019.01); G05B 13/0265 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
providing training data to train each of a plurality of machine learning models to provide a metrology measurement value associated with a particular type of metrology measurement for a substrate based on spectral data collected during a substrate process performed for the substrate, wherein each of the plurality of machine learning models are associated with a different type of a set of machine learning model types;
assigning a performance rating to each of the plurality of machine learning models based on an accuracy of a value for the metrology measurement provided by a respective machine learning model in view of a measured value for the metrology measurement, the measured value generated based on historical metrology data collected by metrology equipment for a prior substrate of a set of prior substrates; and
selecting, in view of the performance rating for each of the plurality of machine learning models, the respective machine learning model to be applied to future spectral data collected during a future substrate process performed for a future substrate.