Analytical predictive capabilities of Laser Induced Breakdown Spectroscopy (LIBS) with Principal Component Analysis (PCA) for plastic classification

Research output: Contribution to journalArticle

62 Citations (Scopus)

Abstract

A Laser Induced Breakdown Spectroscopy (LIBS) technique has been applied for the identification of four widely used plastics, polyethylene terephthalate (PET), high-density polyethylene (PE), polypropylene (PP) and polystyrene (PS), whose recycling is required from commercial and biosafety points of view. The 3rd harmonic (355 nm) nanosecond pulse from an Nd:YAG laser is used to create plasma on the sample surface and identification of the type of the plastic is derived from the plasma emission. Principal Component Analysis (PCA) of the LIBS spectra is employed for the classification of plastics. Distinct methods have been used, apart from principal components of PCA, to further confirm our results. Statistical parameters, viz., Mahalanobis distance (M-distance) and spectral residuals were used for decisive match/no match test which provided successful classification of plastics. Receiver Operating Characteristic (ROC) and Youden's index analyses were carried out to obtain the diagnostic threshold for classification of all four classes of plastics. Sensitivity, specificity, predictive values and discriminative accuracy of the classification tests based on the optimum threshold were calculated. This proves the analytical predictive capabilities of the LIBS technique for plastic identification and classification. The technique of LIBS, in future, can be routinely used in field applications such as plastic waste sorting and recycling.

Original languageEnglish
Pages (from-to)25872-25880
Number of pages9
JournalRSC Advances
Volume3
Issue number48
DOIs
Publication statusPublished - 28-12-2013

Fingerprint

Laser induced breakdown spectroscopy
Principal component analysis
Plastics
Recycling
Plasmas
Polyethylene Terephthalates
Polypropylenes
Polystyrenes
Polyethylene
High density polyethylenes
Sorting
Polyethylene terephthalates
Laser pulses
Lasers

All Science Journal Classification (ASJC) codes

  • Chemistry(all)
  • Chemical Engineering(all)

Cite this

@article{2abd3a93114a4521b8ee72b4c1c4b012,
title = "Analytical predictive capabilities of Laser Induced Breakdown Spectroscopy (LIBS) with Principal Component Analysis (PCA) for plastic classification",
abstract = "A Laser Induced Breakdown Spectroscopy (LIBS) technique has been applied for the identification of four widely used plastics, polyethylene terephthalate (PET), high-density polyethylene (PE), polypropylene (PP) and polystyrene (PS), whose recycling is required from commercial and biosafety points of view. The 3rd harmonic (355 nm) nanosecond pulse from an Nd:YAG laser is used to create plasma on the sample surface and identification of the type of the plastic is derived from the plasma emission. Principal Component Analysis (PCA) of the LIBS spectra is employed for the classification of plastics. Distinct methods have been used, apart from principal components of PCA, to further confirm our results. Statistical parameters, viz., Mahalanobis distance (M-distance) and spectral residuals were used for decisive match/no match test which provided successful classification of plastics. Receiver Operating Characteristic (ROC) and Youden's index analyses were carried out to obtain the diagnostic threshold for classification of all four classes of plastics. Sensitivity, specificity, predictive values and discriminative accuracy of the classification tests based on the optimum threshold were calculated. This proves the analytical predictive capabilities of the LIBS technique for plastic identification and classification. The technique of LIBS, in future, can be routinely used in field applications such as plastic waste sorting and recycling.",
author = "Unnikrishnan, {V. K.} and Choudhari, {K. S.} and Kulkarni, {Suresh D.} and Rajesh Nayak and Kartha, {V. B.} and C. Santhosh",
year = "2013",
month = "12",
day = "28",
doi = "10.1039/c3ra44946g",
language = "English",
volume = "3",
pages = "25872--25880",
journal = "RSC Advances",
issn = "2046-2069",
publisher = "Royal Society of Chemistry",
number = "48",

}

TY - JOUR

T1 - Analytical predictive capabilities of Laser Induced Breakdown Spectroscopy (LIBS) with Principal Component Analysis (PCA) for plastic classification

AU - Unnikrishnan, V. K.

AU - Choudhari, K. S.

AU - Kulkarni, Suresh D.

AU - Nayak, Rajesh

AU - Kartha, V. B.

AU - Santhosh, C.

PY - 2013/12/28

Y1 - 2013/12/28

N2 - A Laser Induced Breakdown Spectroscopy (LIBS) technique has been applied for the identification of four widely used plastics, polyethylene terephthalate (PET), high-density polyethylene (PE), polypropylene (PP) and polystyrene (PS), whose recycling is required from commercial and biosafety points of view. The 3rd harmonic (355 nm) nanosecond pulse from an Nd:YAG laser is used to create plasma on the sample surface and identification of the type of the plastic is derived from the plasma emission. Principal Component Analysis (PCA) of the LIBS spectra is employed for the classification of plastics. Distinct methods have been used, apart from principal components of PCA, to further confirm our results. Statistical parameters, viz., Mahalanobis distance (M-distance) and spectral residuals were used for decisive match/no match test which provided successful classification of plastics. Receiver Operating Characteristic (ROC) and Youden's index analyses were carried out to obtain the diagnostic threshold for classification of all four classes of plastics. Sensitivity, specificity, predictive values and discriminative accuracy of the classification tests based on the optimum threshold were calculated. This proves the analytical predictive capabilities of the LIBS technique for plastic identification and classification. The technique of LIBS, in future, can be routinely used in field applications such as plastic waste sorting and recycling.

AB - A Laser Induced Breakdown Spectroscopy (LIBS) technique has been applied for the identification of four widely used plastics, polyethylene terephthalate (PET), high-density polyethylene (PE), polypropylene (PP) and polystyrene (PS), whose recycling is required from commercial and biosafety points of view. The 3rd harmonic (355 nm) nanosecond pulse from an Nd:YAG laser is used to create plasma on the sample surface and identification of the type of the plastic is derived from the plasma emission. Principal Component Analysis (PCA) of the LIBS spectra is employed for the classification of plastics. Distinct methods have been used, apart from principal components of PCA, to further confirm our results. Statistical parameters, viz., Mahalanobis distance (M-distance) and spectral residuals were used for decisive match/no match test which provided successful classification of plastics. Receiver Operating Characteristic (ROC) and Youden's index analyses were carried out to obtain the diagnostic threshold for classification of all four classes of plastics. Sensitivity, specificity, predictive values and discriminative accuracy of the classification tests based on the optimum threshold were calculated. This proves the analytical predictive capabilities of the LIBS technique for plastic identification and classification. The technique of LIBS, in future, can be routinely used in field applications such as plastic waste sorting and recycling.

UR - http://www.scopus.com/inward/record.url?scp=84887969533&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84887969533&partnerID=8YFLogxK

U2 - 10.1039/c3ra44946g

DO - 10.1039/c3ra44946g

M3 - Article

VL - 3

SP - 25872

EP - 25880

JO - RSC Advances

JF - RSC Advances

SN - 2046-2069

IS - 48

ER -