HR: 16:00h
AN: P14A-01    [Abstracts]
TI: QUANTITATIVE REMOTE LASER-INDUCED BREAKDOWN SPECTROSCOPY BY MULTIVARIATE ANALYSIS
AU: * Clegg, S M
EM: sclegg@lanl.gov
AF: Los Alamos National Laboratory, P.O. Box 1663, Los Alamos, NM 87545, United States
AU: Sklute, E C
EM: mdyar@mtholyoke.edu
AF: Dept. of Astronomy Mount Holyoke College, 50 College Street, South Hadley, MA 01075, United States
AU: Dyar, M D
EM: mdyar@mtholyoke.edu
AF: Dept. of Astronomy Mount Holyoke College, 50 College Street, South Hadley, MA 01075, United States
AU: Barefield, J E
EM: jbarefield@lanl.gov
AF: Los Alamos National Laboratory, P.O. Box 1663, Los Alamos, NM 87545, United States
AU: Wiens, R C
EM: rwiens@lanl.gov
AF: Los Alamos National Laboratory, P.O. Box 1663, Los Alamos, NM 87545, United States
AB: The ChemCam instrument selected for the Mars Science Laboratory (MSL) rover includes a remote Laser- Induced Breakdown Spectrometer (LIBS) that will quantitatively probe samples up to 9m from the rover mast. LIBS is fundamentally an elemental analysis technique. LIBS involves focusing a Nd:YAG laser operating at 1064 nm onto the surface of the sample. The laser ablates material from the surface, generating an expanding plasma containing electronically excited ions, atoms, and small molecules. As these electronically excited species relax back to the ground state, they emit light at wavelengths characteristic of the species present in the sample. Some of this emission is directed into one of three dispersive spectrometers. In this paper, we studied a suite of 18 igneous and highly-metamorphosed samples from a wide variety of parageneses for which chemical analyses by XRF were already available. Rocks were chosen to represent a range of chemical composition from basalt to rhyolite, thus providing significant variations in all of the major element contents (Si, Fe, Al, Ca, Na, K, O, Ti, Mg, and Mn). These samples were probed at a 9m standoff distance under experimental conditions that are similar to ChemCam. Extracting quantitative elemental concentrations from LIBS spectra is complicated by the chemical matrix effects. Conventional methods for obtaining quantitative chemical data from LIBS analyses are compared with new multivariate analysis (MVA) techniques that appear to compensate for these chemical matrix effects. The traditional analyses use specific elemental peak heights or areas, which compared with calibration curves for each element at one or more emission lines for a series of standard samples. Because of matrix effects, the calibration standards generally must have similar chemistries to the unknown samples, and thus this conventional approach imposes severe limitations on application of the technique to remote analyses. In this suite of samples, the use of traditional methods results in chemical analyses with significant uncertainties. Alternatively, greatly-improved quantitative elemental analysis was accomplished by using a Partial Least Squares (PLS) calibration model for all of the major elements of interest. Principal Components Analysis (PCA) and Soft Independent Modeling of Class Analogy (SIMCA) are then employed to predict the rock-type of the sample. These MVA techniques appear to compensate for these matrix effects because the analysis finds correlations between the spectra (independent variables), the individual elements of interest (dependent variables such as Si) as well as the other elements in the matrix.
DE: 1000 GEOCHEMISTRY
DE: 5464 Remote sensing
DE: 5470 Surface materials and properties
DE: 5494 Instruments and techniques
SC: Planetary Sciences [P]
MN: 2007 Fall Meeting