Predictive modeling and analysis of HfO2 thin film process based on Bayesian information criterion using PCA-based neural networks

Young Don Ko, Pyung Moon, Chang Eun Kim, Moon Ho Ham, Myong Kee Jeong, Alberto Garcia-Diaz, Jae Min Myoung, Ilgu Yun

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Principal component analysis (PCA)-based neural network (NNet) models of HfO2 thin films are used to study the process of efficient model selection and develop an improved model by using multivariate functional data such as X-ray diffraction data (XRD). The accumulation capacitance and the hysteresis index input parameters, both characteristic of HfO2 dielectric films, were selected for the inclusion in the model by analyzing the process conditions. Standardized XRD were used to analyze the characteristic variations for different process conditions; the responses and the electrical properties were predicted by NNet modeling using crystallinity-based measurement data. A Bayesian information criterion (BIC) was used to compare the model efficiency and to select an improved model for response prediction. Two conclusions summarize the results of the research documented in this paper: (i) physical or material properties can be predicted by the PCA-based NNet model using large-dimension data, and (ii) BIC can be used for the selection and evaluation of predictive models in semiconductor manufacturing processes.

Original languageEnglish
Pages (from-to)1334-1339
Number of pages6
JournalSurface and Interface Analysis
Volume45
Issue number9
DOIs
Publication statusPublished - 2013 Sep 1

Fingerprint

principal components analysis
Principal component analysis
Neural networks
Thin films
thin films
X ray diffraction
Dielectric films
diffraction
Hysteresis
crystallinity
Materials properties
Electric properties
x rays
Capacitance
manufacturing
Physical properties
physical properties
capacitance
hysteresis
electrical properties

All Science Journal Classification (ASJC) codes

  • Chemistry(all)
  • Condensed Matter Physics
  • Surfaces and Interfaces
  • Surfaces, Coatings and Films
  • Materials Chemistry

Cite this

Ko, Young Don ; Moon, Pyung ; Kim, Chang Eun ; Ham, Moon Ho ; Jeong, Myong Kee ; Garcia-Diaz, Alberto ; Myoung, Jae Min ; Yun, Ilgu. / Predictive modeling and analysis of HfO2 thin film process based on Bayesian information criterion using PCA-based neural networks. In: Surface and Interface Analysis. 2013 ; Vol. 45, No. 9. pp. 1334-1339.
@article{ef6cda469b84478892f25efb1ac110db,
title = "Predictive modeling and analysis of HfO2 thin film process based on Bayesian information criterion using PCA-based neural networks",
abstract = "Principal component analysis (PCA)-based neural network (NNet) models of HfO2 thin films are used to study the process of efficient model selection and develop an improved model by using multivariate functional data such as X-ray diffraction data (XRD). The accumulation capacitance and the hysteresis index input parameters, both characteristic of HfO2 dielectric films, were selected for the inclusion in the model by analyzing the process conditions. Standardized XRD were used to analyze the characteristic variations for different process conditions; the responses and the electrical properties were predicted by NNet modeling using crystallinity-based measurement data. A Bayesian information criterion (BIC) was used to compare the model efficiency and to select an improved model for response prediction. Two conclusions summarize the results of the research documented in this paper: (i) physical or material properties can be predicted by the PCA-based NNet model using large-dimension data, and (ii) BIC can be used for the selection and evaluation of predictive models in semiconductor manufacturing processes.",
author = "Ko, {Young Don} and Pyung Moon and Kim, {Chang Eun} and Ham, {Moon Ho} and Jeong, {Myong Kee} and Alberto Garcia-Diaz and Myoung, {Jae Min} and Ilgu Yun",
year = "2013",
month = "9",
day = "1",
doi = "10.1002/sia.5286",
language = "English",
volume = "45",
pages = "1334--1339",
journal = "Surface and Interface Analysis",
issn = "0142-2421",
publisher = "John Wiley and Sons Ltd",
number = "9",

}

Predictive modeling and analysis of HfO2 thin film process based on Bayesian information criterion using PCA-based neural networks. / Ko, Young Don; Moon, Pyung; Kim, Chang Eun; Ham, Moon Ho; Jeong, Myong Kee; Garcia-Diaz, Alberto; Myoung, Jae Min; Yun, Ilgu.

In: Surface and Interface Analysis, Vol. 45, No. 9, 01.09.2013, p. 1334-1339.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Predictive modeling and analysis of HfO2 thin film process based on Bayesian information criterion using PCA-based neural networks

AU - Ko, Young Don

AU - Moon, Pyung

AU - Kim, Chang Eun

AU - Ham, Moon Ho

AU - Jeong, Myong Kee

AU - Garcia-Diaz, Alberto

AU - Myoung, Jae Min

AU - Yun, Ilgu

PY - 2013/9/1

Y1 - 2013/9/1

N2 - Principal component analysis (PCA)-based neural network (NNet) models of HfO2 thin films are used to study the process of efficient model selection and develop an improved model by using multivariate functional data such as X-ray diffraction data (XRD). The accumulation capacitance and the hysteresis index input parameters, both characteristic of HfO2 dielectric films, were selected for the inclusion in the model by analyzing the process conditions. Standardized XRD were used to analyze the characteristic variations for different process conditions; the responses and the electrical properties were predicted by NNet modeling using crystallinity-based measurement data. A Bayesian information criterion (BIC) was used to compare the model efficiency and to select an improved model for response prediction. Two conclusions summarize the results of the research documented in this paper: (i) physical or material properties can be predicted by the PCA-based NNet model using large-dimension data, and (ii) BIC can be used for the selection and evaluation of predictive models in semiconductor manufacturing processes.

AB - Principal component analysis (PCA)-based neural network (NNet) models of HfO2 thin films are used to study the process of efficient model selection and develop an improved model by using multivariate functional data such as X-ray diffraction data (XRD). The accumulation capacitance and the hysteresis index input parameters, both characteristic of HfO2 dielectric films, were selected for the inclusion in the model by analyzing the process conditions. Standardized XRD were used to analyze the characteristic variations for different process conditions; the responses and the electrical properties were predicted by NNet modeling using crystallinity-based measurement data. A Bayesian information criterion (BIC) was used to compare the model efficiency and to select an improved model for response prediction. Two conclusions summarize the results of the research documented in this paper: (i) physical or material properties can be predicted by the PCA-based NNet model using large-dimension data, and (ii) BIC can be used for the selection and evaluation of predictive models in semiconductor manufacturing processes.

UR - http://www.scopus.com/inward/record.url?scp=84881368965&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84881368965&partnerID=8YFLogxK

U2 - 10.1002/sia.5286

DO - 10.1002/sia.5286

M3 - Article

AN - SCOPUS:84881368965

VL - 45

SP - 1334

EP - 1339

JO - Surface and Interface Analysis

JF - Surface and Interface Analysis

SN - 0142-2421

IS - 9

ER -