Application of artificial neural networks to rapid data analysis in combinatorial nanoparticle syntheses

Yuuichi Orimoto, Kosuke Watanabe, Kenichi Yamashita, Masato Uehara, Hiroyuki Nakamura, Takeshi Furuya, Hideaki Maeda

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

The synthesis of nanomaterials is extremely sensitive to various factors under experimental conditions. Therefore, for controlling synthesis, it is important to ascertain comprehensively the relations between the conditions and nanomaterial properties. This study is intended to acquire the relations in data sets from combinatorial syntheses by means of an artificial neural network-based method toward property optimization. Recently, 3404 data sets were obtained systematically using microreactorbased combinatorial CdSe nanoparticle (NP) syntheses for examining conditionproperty relations. However, it is time-consuming to acquire the relations for the following reasons: (i) massiveness and complexity of the multivariate data sets, (ii) small numbers of points permitted for each experimental parameter to avoid 'combination explosion', and (iii) errors and missing data attributable to experimental reasons. In this work, an NN-based data analysis method was developed and applied for analyzing the data sets to acquire the relations. In the method, an exhaustive 1600 training processes and the following ensemble approach are performed for obtaining preferred NNs. Results show that NNs extract essential patterns on the condition-property relations on a realistic time scale. The trained NNs are capable of predicting the NP properties even for new experimental conditions with high accuracy. Moreover, data interpolation and sensitivity analysis based on the NNs provide us the relations as accessible descriptions such as multidimensional condition-property landscapes and key parameters for controlling the synthesis. Such information can guide us when optimizing the NP properties. Our approach is suitable to extract condition-property relations rapidly from the combinatorial synthesis data and is expected to be effective for various types of target materials, even with unknown properties, because of the flexibility of the NN analysis.

Original languageEnglish
Pages (from-to)17885-17896
Number of pages12
JournalJournal of Physical Chemistry C
Volume116
Issue number33
DOIs
Publication statusPublished - Aug 23 2012

Fingerprint

Nanoparticles
Neural networks
nanoparticles
synthesis
Nanostructured materials
Sensitivity analysis
Explosions
Interpolation
sensitivity analysis
interpolation
explosions
flexibility
education
optimization

All Science Journal Classification (ASJC) codes

  • Electronic, Optical and Magnetic Materials
  • Energy(all)
  • Physical and Theoretical Chemistry
  • Surfaces, Coatings and Films

Cite this

Orimoto, Y., Watanabe, K., Yamashita, K., Uehara, M., Nakamura, H., Furuya, T., & Maeda, H. (2012). Application of artificial neural networks to rapid data analysis in combinatorial nanoparticle syntheses. Journal of Physical Chemistry C, 116(33), 17885-17896. https://doi.org/10.1021/jp3031122

Application of artificial neural networks to rapid data analysis in combinatorial nanoparticle syntheses. / Orimoto, Yuuichi; Watanabe, Kosuke; Yamashita, Kenichi; Uehara, Masato; Nakamura, Hiroyuki; Furuya, Takeshi; Maeda, Hideaki.

In: Journal of Physical Chemistry C, Vol. 116, No. 33, 23.08.2012, p. 17885-17896.

Research output: Contribution to journalArticle

Orimoto, Y, Watanabe, K, Yamashita, K, Uehara, M, Nakamura, H, Furuya, T & Maeda, H 2012, 'Application of artificial neural networks to rapid data analysis in combinatorial nanoparticle syntheses', Journal of Physical Chemistry C, vol. 116, no. 33, pp. 17885-17896. https://doi.org/10.1021/jp3031122
Orimoto, Yuuichi ; Watanabe, Kosuke ; Yamashita, Kenichi ; Uehara, Masato ; Nakamura, Hiroyuki ; Furuya, Takeshi ; Maeda, Hideaki. / Application of artificial neural networks to rapid data analysis in combinatorial nanoparticle syntheses. In: Journal of Physical Chemistry C. 2012 ; Vol. 116, No. 33. pp. 17885-17896.
@article{28e5fa6502bf41ddbb2d0281756cc263,
title = "Application of artificial neural networks to rapid data analysis in combinatorial nanoparticle syntheses",
abstract = "The synthesis of nanomaterials is extremely sensitive to various factors under experimental conditions. Therefore, for controlling synthesis, it is important to ascertain comprehensively the relations between the conditions and nanomaterial properties. This study is intended to acquire the relations in data sets from combinatorial syntheses by means of an artificial neural network-based method toward property optimization. Recently, 3404 data sets were obtained systematically using microreactorbased combinatorial CdSe nanoparticle (NP) syntheses for examining conditionproperty relations. However, it is time-consuming to acquire the relations for the following reasons: (i) massiveness and complexity of the multivariate data sets, (ii) small numbers of points permitted for each experimental parameter to avoid 'combination explosion', and (iii) errors and missing data attributable to experimental reasons. In this work, an NN-based data analysis method was developed and applied for analyzing the data sets to acquire the relations. In the method, an exhaustive 1600 training processes and the following ensemble approach are performed for obtaining preferred NNs. Results show that NNs extract essential patterns on the condition-property relations on a realistic time scale. The trained NNs are capable of predicting the NP properties even for new experimental conditions with high accuracy. Moreover, data interpolation and sensitivity analysis based on the NNs provide us the relations as accessible descriptions such as multidimensional condition-property landscapes and key parameters for controlling the synthesis. Such information can guide us when optimizing the NP properties. Our approach is suitable to extract condition-property relations rapidly from the combinatorial synthesis data and is expected to be effective for various types of target materials, even with unknown properties, because of the flexibility of the NN analysis.",
author = "Yuuichi Orimoto and Kosuke Watanabe and Kenichi Yamashita and Masato Uehara and Hiroyuki Nakamura and Takeshi Furuya and Hideaki Maeda",
year = "2012",
month = "8",
day = "23",
doi = "10.1021/jp3031122",
language = "English",
volume = "116",
pages = "17885--17896",
journal = "Journal of Physical Chemistry C",
issn = "1932-7447",
publisher = "American Chemical Society",
number = "33",

}

TY - JOUR

T1 - Application of artificial neural networks to rapid data analysis in combinatorial nanoparticle syntheses

AU - Orimoto, Yuuichi

AU - Watanabe, Kosuke

AU - Yamashita, Kenichi

AU - Uehara, Masato

AU - Nakamura, Hiroyuki

AU - Furuya, Takeshi

AU - Maeda, Hideaki

PY - 2012/8/23

Y1 - 2012/8/23

N2 - The synthesis of nanomaterials is extremely sensitive to various factors under experimental conditions. Therefore, for controlling synthesis, it is important to ascertain comprehensively the relations between the conditions and nanomaterial properties. This study is intended to acquire the relations in data sets from combinatorial syntheses by means of an artificial neural network-based method toward property optimization. Recently, 3404 data sets were obtained systematically using microreactorbased combinatorial CdSe nanoparticle (NP) syntheses for examining conditionproperty relations. However, it is time-consuming to acquire the relations for the following reasons: (i) massiveness and complexity of the multivariate data sets, (ii) small numbers of points permitted for each experimental parameter to avoid 'combination explosion', and (iii) errors and missing data attributable to experimental reasons. In this work, an NN-based data analysis method was developed and applied for analyzing the data sets to acquire the relations. In the method, an exhaustive 1600 training processes and the following ensemble approach are performed for obtaining preferred NNs. Results show that NNs extract essential patterns on the condition-property relations on a realistic time scale. The trained NNs are capable of predicting the NP properties even for new experimental conditions with high accuracy. Moreover, data interpolation and sensitivity analysis based on the NNs provide us the relations as accessible descriptions such as multidimensional condition-property landscapes and key parameters for controlling the synthesis. Such information can guide us when optimizing the NP properties. Our approach is suitable to extract condition-property relations rapidly from the combinatorial synthesis data and is expected to be effective for various types of target materials, even with unknown properties, because of the flexibility of the NN analysis.

AB - The synthesis of nanomaterials is extremely sensitive to various factors under experimental conditions. Therefore, for controlling synthesis, it is important to ascertain comprehensively the relations between the conditions and nanomaterial properties. This study is intended to acquire the relations in data sets from combinatorial syntheses by means of an artificial neural network-based method toward property optimization. Recently, 3404 data sets were obtained systematically using microreactorbased combinatorial CdSe nanoparticle (NP) syntheses for examining conditionproperty relations. However, it is time-consuming to acquire the relations for the following reasons: (i) massiveness and complexity of the multivariate data sets, (ii) small numbers of points permitted for each experimental parameter to avoid 'combination explosion', and (iii) errors and missing data attributable to experimental reasons. In this work, an NN-based data analysis method was developed and applied for analyzing the data sets to acquire the relations. In the method, an exhaustive 1600 training processes and the following ensemble approach are performed for obtaining preferred NNs. Results show that NNs extract essential patterns on the condition-property relations on a realistic time scale. The trained NNs are capable of predicting the NP properties even for new experimental conditions with high accuracy. Moreover, data interpolation and sensitivity analysis based on the NNs provide us the relations as accessible descriptions such as multidimensional condition-property landscapes and key parameters for controlling the synthesis. Such information can guide us when optimizing the NP properties. Our approach is suitable to extract condition-property relations rapidly from the combinatorial synthesis data and is expected to be effective for various types of target materials, even with unknown properties, because of the flexibility of the NN analysis.

UR - http://www.scopus.com/inward/record.url?scp=84868588462&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84868588462&partnerID=8YFLogxK

U2 - 10.1021/jp3031122

DO - 10.1021/jp3031122

M3 - Article

AN - SCOPUS:84868588462

VL - 116

SP - 17885

EP - 17896

JO - Journal of Physical Chemistry C

JF - Journal of Physical Chemistry C

SN - 1932-7447

IS - 33

ER -