US 12,321,102 B2
Machine and deep learning methods for spectra-based metrology and process control
Barak Bringoltz, Rishon le Tzion (IL); Ran Yacoby, Jerusalem (IL); Noam Tal, Mishmar Hashivaa (IL); Shay Yogev, Kibbutz Kfar Menachem (IL); Boaz Sturlesi, Rehovot (IL); and Oded Cohen, Rehovot (IL)
Assigned to NOVA LTD., Rehovot (IL)
Filed by NOVA LTD., Rehovot (IL)
Filed on Nov. 13, 2023, as Appl. No. 18/508,177.
Application 18/508,177 is a continuation of application No. 17/995,706, granted, now 11,815,819, issued on Nov. 14, 2023, previously published as PCT/IL2021/050389, filed on Apr. 6, 2021.
Claims priority of provisional application 63/005,589, filed on Apr. 6, 2020.
Prior Publication US 2024/0310737 A1, Sep. 19, 2024
Int. Cl. G05B 13/02 (2006.01); G03F 7/00 (2006.01)
CPC G03F 7/70508 (2013.01) [G03F 7/705 (2013.01); G03F 7/70525 (2013.01); G05B 13/0265 (2013.01); G05B 2219/2602 (2013.01); G05B 2219/45031 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A system for Advance Process Control (APC) in semiconductor manufacturing comprising one or more processors having one or more associated non-transient memories comprising instructions that when executed by the one or more processors implement steps of:
for a plurality of wafer sites,
receiving a pre-process set of scatterometric training data, measured before implementation of a processing step,
receiving a corresponding post-process set of scatterometric training data measured after implementation of the process step, and
receiving a set of process control knob training data indicative of process control knob settings applied during implementation of the process step; and
generating a machine learning model correlating variations in the pre-process sets of scatterometric training data and the corresponding process control knob training data with the corresponding post-process sets of scatterometric training data, to train the machine learning model to recommend changes to process control knob settings to compensate for variations in the pre-process scatterometric data: wherein the generating of the machine learning model comprises training a neural network (NN), including multiple encoder layers wherein the pre-process sets of scatterometric training data are applied as model input, wherein the corresponding post-process sets of scatterometric training data are applied as model output, wherein the multiple process control knobs training data are applied as auxiliary inputs that intersect the NN at any one of the multiple encoder layers;
wherein the generating of the machine learning model comprises applying a pair of loss functions for backpropagation of the NN, wherein a first loss function of the pair maximizes a similarity between outputs of the NN and the post-process scatterometric training data, and a second loss function of the pair expresses a quality of similarity between auxiliary outputs and the set of process knob training data.