Recurrent feature-incorporated convolutional neural network for virtual metrology of the chemical mechanical planarization process

Ki Bum Lee, Chang Ouk Kim

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

In semiconductor manufacturing, the chemical mechanical planarization (CMP) process produces higher thickness variability in the edge area of the wafer than that in the center area due to the characteristics of the polishing operation. To address this problem, advanced CMP equipment includes a function that controls the removal rate of each area. However, to take full advantage of this capability, effective advanced process control systems must be implemented with a virtual metrology (VM) model. However, the prediction performance of the VM model often deteriorates due to process drift. Here, we present a deep learning-based VM model that demonstrates high performance in the presence of nonlinear process drift. The proposed model combines a recurrent neural network and a convolutional neural network to extract time-dependent and time-independent features. A two-stage model training method is proposed that alternately updates the weights of each network to improve prediction performance. In the experiments using on-site CMP process data, the performance of the deep learning models exceeded that of standard machine learning models. The proposed model showed an 8.48% decrease in process variability relative to the best-performing machine learning model, which was elastic nets.

Original languageEnglish
Pages (from-to)1-14
Number of pages14
JournalJournal of Intelligent Manufacturing
DOIs
Publication statusAccepted/In press - 2018 Jul 17

Fingerprint

Chemical mechanical polishing
Neural networks
Learning systems
Recurrent neural networks
Polishing
Process control
Semiconductor materials
Control systems

All Science Journal Classification (ASJC) codes

  • Software
  • Industrial and Manufacturing Engineering
  • Artificial Intelligence

Cite this

@article{9ffa64d3aef141cfa0d9f2e9ba4de4cf,
title = "Recurrent feature-incorporated convolutional neural network for virtual metrology of the chemical mechanical planarization process",
abstract = "In semiconductor manufacturing, the chemical mechanical planarization (CMP) process produces higher thickness variability in the edge area of the wafer than that in the center area due to the characteristics of the polishing operation. To address this problem, advanced CMP equipment includes a function that controls the removal rate of each area. However, to take full advantage of this capability, effective advanced process control systems must be implemented with a virtual metrology (VM) model. However, the prediction performance of the VM model often deteriorates due to process drift. Here, we present a deep learning-based VM model that demonstrates high performance in the presence of nonlinear process drift. The proposed model combines a recurrent neural network and a convolutional neural network to extract time-dependent and time-independent features. A two-stage model training method is proposed that alternately updates the weights of each network to improve prediction performance. In the experiments using on-site CMP process data, the performance of the deep learning models exceeded that of standard machine learning models. The proposed model showed an 8.48{\%} decrease in process variability relative to the best-performing machine learning model, which was elastic nets.",
author = "Lee, {Ki Bum} and Kim, {Chang Ouk}",
year = "2018",
month = "7",
day = "17",
doi = "10.1007/s10845-018-1437-4",
language = "English",
pages = "1--14",
journal = "Journal of Intelligent Manufacturing",
issn = "0956-5515",
publisher = "Springer Netherlands",

}

TY - JOUR

T1 - Recurrent feature-incorporated convolutional neural network for virtual metrology of the chemical mechanical planarization process

AU - Lee, Ki Bum

AU - Kim, Chang Ouk

PY - 2018/7/17

Y1 - 2018/7/17

N2 - In semiconductor manufacturing, the chemical mechanical planarization (CMP) process produces higher thickness variability in the edge area of the wafer than that in the center area due to the characteristics of the polishing operation. To address this problem, advanced CMP equipment includes a function that controls the removal rate of each area. However, to take full advantage of this capability, effective advanced process control systems must be implemented with a virtual metrology (VM) model. However, the prediction performance of the VM model often deteriorates due to process drift. Here, we present a deep learning-based VM model that demonstrates high performance in the presence of nonlinear process drift. The proposed model combines a recurrent neural network and a convolutional neural network to extract time-dependent and time-independent features. A two-stage model training method is proposed that alternately updates the weights of each network to improve prediction performance. In the experiments using on-site CMP process data, the performance of the deep learning models exceeded that of standard machine learning models. The proposed model showed an 8.48% decrease in process variability relative to the best-performing machine learning model, which was elastic nets.

AB - In semiconductor manufacturing, the chemical mechanical planarization (CMP) process produces higher thickness variability in the edge area of the wafer than that in the center area due to the characteristics of the polishing operation. To address this problem, advanced CMP equipment includes a function that controls the removal rate of each area. However, to take full advantage of this capability, effective advanced process control systems must be implemented with a virtual metrology (VM) model. However, the prediction performance of the VM model often deteriorates due to process drift. Here, we present a deep learning-based VM model that demonstrates high performance in the presence of nonlinear process drift. The proposed model combines a recurrent neural network and a convolutional neural network to extract time-dependent and time-independent features. A two-stage model training method is proposed that alternately updates the weights of each network to improve prediction performance. In the experiments using on-site CMP process data, the performance of the deep learning models exceeded that of standard machine learning models. The proposed model showed an 8.48% decrease in process variability relative to the best-performing machine learning model, which was elastic nets.

UR - http://www.scopus.com/inward/record.url?scp=85049965534&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85049965534&partnerID=8YFLogxK

U2 - 10.1007/s10845-018-1437-4

DO - 10.1007/s10845-018-1437-4

M3 - Article

AN - SCOPUS:85049965534

SP - 1

EP - 14

JO - Journal of Intelligent Manufacturing

JF - Journal of Intelligent Manufacturing

SN - 0956-5515

ER -