US 12,447,577 B2
Polishing apparatus using neural network for monitoring
Kun Xu, Sunol, CA (US); Hassan G. Iravani, Sunnyvale, CA (US); Denis Ivanov, St. Petersburg (RU); Boguslaw A. Swedek, Morgan Hill, CA (US); Shih-Haur Shen, Sunnyvale, CA (US); Harry Q. Lee, Los Altos, CA (US); and Benjamin Cherian, San Jose, CA (US)
Assigned to Applied Materials, Inc., Santa Clara, CA (US)
Filed by Applied Materials, Inc., Santa Clara, CA (US)
Filed on Apr. 15, 2021, as Appl. No. 17/232,080.
Application 17/232,080 is a continuation of application No. 15/953,203, filed on Apr. 13, 2018, granted, now 10,994,389.
Claims priority of provisional application 62/488,688, filed on Apr. 21, 2017.
Prior Publication US 2021/0229234 A1, Jul. 29, 2021
Int. Cl. B24B 37/013 (2012.01); G06N 3/08 (2023.01); G06N 3/084 (2023.01); H01L 21/306 (2006.01); G06N 3/048 (2023.01); H01L 21/66 (2006.01)
CPC B24B 37/013 (2013.01) [G06N 3/08 (2013.01); G06N 3/084 (2013.01); H01L 21/30625 (2013.01); G06N 3/048 (2023.01); H01L 22/12 (2013.01); H01L 22/26 (2013.01)] 16 Claims
OG exemplary drawing
 
1. A method of polishing a substrate, comprising:
polishing a layer on the substrate at a polishing station;
monitoring the layer during polishing at the polishing station with an in-situ monitoring system to generate a plurality of initial values for a plurality of different locations on the layer;
inputting a first multiplicity of initial values from the plurality of initial values into a first plurality of corresponding input nodes of a neural network, the first multiplicity of initial values corresponding to a first multiplicity of locations from the plurality of different locations that are within a first region on the substrate, and wherein initial values from the plurality of initial values that are not input into input nodes of the neural network include a second multiplicity of initial values corresponding to a second multiplicity of locations from the plurality of different locations that are within a different second region on the substrate;
receiving a first multiplicity of modified values from a plurality of corresponding output nodes of the neural network, wherein each output node is connected to multiple hidden nodes and each hidden node is connected to multiple input nodes such that each modified value corresponding to a different location on the layer is calculated based on at least two input values from the first multiplicity of initial values corresponding to at least two successive locations of the first multiplicity of locations within a predetermined distance from the different location on the layer from which the modified value is calculated; and
at least one of detecting a polishing endpoint or modifying a polishing parameter based on the first multiplicity of modified values and the second multiplicity of initial values.