skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Single Variable and Multivariate Analysis of Remote Laser-Induced Breakdown Spectra for Prediction of Rb, Sr, Cr, Ba, and V in Igneous Rocks

Abstract

Laser-induced breakdown spectroscopy (LIBS) will be employed by the ChemCam instrument on the Mars Science Laboratory rover Curiosity to obtain UV, VIS, and VNIR atomic emission spectra of surface rocks and soils. LIBS quantitative analysis is complicated by chemical matrix effects related to abundances of neutral and ionized species in the resultant plasma, collisional interactions within plasma, laser-to-sample coupling efficiency, and self-absorption. Atmospheric composition and pressure also influence the intensity of LIBS plasma. These chemical matrix effects influence the ratio of intensity or area of a given emission line to the abundance of the element producing that line. To compensate for these complications, multivariate techniques, specifically partial least-squares regression (PLS), have been utilized to predict major element compositions (>1 wt.% oxide) of rocks, PLS methods regress one or multiple response variables (elemental concentrations) against multiple explanatory variables (intensity at each pixel of the spectrometers). Because PLS utilizes all available explanatory variable and eliminates multicollinearity, it generally performs better than univariate methods for prediction of major elements. However, peaks arising from emissions from trace elements may be masked by peaks of higher intensities from major elements. Thus in PLS regression, wherein a correlation coefficient is determined for each elemental concentration atmore » each spectrometer pixel, trace elements may show high correlation with more intense lines resulting from optical emissions of other elements. This could result in error in predictions of trace element concentrations. Here, results of simple linear regression (SLR) and multivariate PLS-2 regression for determination of trace Rb, Sr, Cr, Ba, and V in igneous rock samples are compared. This study focuses on comparisons using only line intensities rather than peak areas to highlight differences between SLR and PLS.« less

Authors:
 [1];  [1];  [2];  [2];  [2]
  1. Los Alamos National Laboratory
  2. MT HOLYOKE COLLEGE
Publication Date:
Research Org.:
Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1044914
Report Number(s):
LA-UR-10-08457; LA-UR-10-8457
TRN: US201214%%579
DOE Contract Number:  
AC52-06NA25396
Resource Type:
Conference
Resource Relation:
Conference: 42nd Lunar and Planetary Science Conference ; March 7, 2011 ; The Woodlands, TX
Country of Publication:
United States
Language:
English
Subject:
37 INORGANIC, ORGANIC, PHYSICAL AND ANALYTICAL CHEMISTRY; 54 ENVIRONMENTAL SCIENCES; 97 MATHEMATICAL METHODS AND COMPUTING; ABUNDANCE; BREAKDOWN; EFFICIENCY; ELEMENTS; EMISSION SPECTRA; FORECASTING; IGNEOUS ROCKS; MULTIVARIATE ANALYSIS; PLASMA; SOILS; SPECTRA; SPECTROMETERS; SPECTROSCOPY; TRACE AMOUNTS

Citation Formats

Clegg, Samuel M, Wiens, Roger C, Speicher, Elly A, Dyar, Melinda D, and Carmosino, Marco L. Single Variable and Multivariate Analysis of Remote Laser-Induced Breakdown Spectra for Prediction of Rb, Sr, Cr, Ba, and V in Igneous Rocks. United States: N. p., 2010. Web.
Clegg, Samuel M, Wiens, Roger C, Speicher, Elly A, Dyar, Melinda D, & Carmosino, Marco L. Single Variable and Multivariate Analysis of Remote Laser-Induced Breakdown Spectra for Prediction of Rb, Sr, Cr, Ba, and V in Igneous Rocks. United States.
Clegg, Samuel M, Wiens, Roger C, Speicher, Elly A, Dyar, Melinda D, and Carmosino, Marco L. 2010. "Single Variable and Multivariate Analysis of Remote Laser-Induced Breakdown Spectra for Prediction of Rb, Sr, Cr, Ba, and V in Igneous Rocks". United States. https://www.osti.gov/servlets/purl/1044914.
@article{osti_1044914,
title = {Single Variable and Multivariate Analysis of Remote Laser-Induced Breakdown Spectra for Prediction of Rb, Sr, Cr, Ba, and V in Igneous Rocks},
author = {Clegg, Samuel M and Wiens, Roger C and Speicher, Elly A and Dyar, Melinda D and Carmosino, Marco L},
abstractNote = {Laser-induced breakdown spectroscopy (LIBS) will be employed by the ChemCam instrument on the Mars Science Laboratory rover Curiosity to obtain UV, VIS, and VNIR atomic emission spectra of surface rocks and soils. LIBS quantitative analysis is complicated by chemical matrix effects related to abundances of neutral and ionized species in the resultant plasma, collisional interactions within plasma, laser-to-sample coupling efficiency, and self-absorption. Atmospheric composition and pressure also influence the intensity of LIBS plasma. These chemical matrix effects influence the ratio of intensity or area of a given emission line to the abundance of the element producing that line. To compensate for these complications, multivariate techniques, specifically partial least-squares regression (PLS), have been utilized to predict major element compositions (>1 wt.% oxide) of rocks, PLS methods regress one or multiple response variables (elemental concentrations) against multiple explanatory variables (intensity at each pixel of the spectrometers). Because PLS utilizes all available explanatory variable and eliminates multicollinearity, it generally performs better than univariate methods for prediction of major elements. However, peaks arising from emissions from trace elements may be masked by peaks of higher intensities from major elements. Thus in PLS regression, wherein a correlation coefficient is determined for each elemental concentration at each spectrometer pixel, trace elements may show high correlation with more intense lines resulting from optical emissions of other elements. This could result in error in predictions of trace element concentrations. Here, results of simple linear regression (SLR) and multivariate PLS-2 regression for determination of trace Rb, Sr, Cr, Ba, and V in igneous rock samples are compared. This study focuses on comparisons using only line intensities rather than peak areas to highlight differences between SLR and PLS.},
doi = {},
url = {https://www.osti.gov/biblio/1044914}, journal = {},
number = ,
volume = ,
place = {United States},
year = {Thu Dec 23 00:00:00 EST 2010},
month = {Thu Dec 23 00:00:00 EST 2010}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share: