Nonlinear mapping technique for data visualization and clustering assessment of LIBS data: application to ChemCam data

ChemCam is a remote laser-induced breakdown spectroscopy (LIBS) instrument that will arrive on Mars in 2012, on-board the Mars Science Laboratory Rover. The LIBS technique is crucial to accurately identify samples and quantify elemental abundances at various distances from the rover. In this study, we compare different linear and nonlinear multivariate techniques to visualize and discriminate clusters in two dimensions (2D) from the data obtained with ChemCam. We have used principal components analysis (PCA) and independent components analysis (ICA) for the linear tools and compared them with the nonlinear Sammon’s map projection technique. We demonstrate that the Sammon’s map gives the best 2D representation of the data set, with optimization values from 2.8% to 4.3% (0% is a perfect representation), together with an entropy value of 0.81 for the purity of the clustering analysis. The linear 2D projections result in three (ICA) and five times (PCA) more stress, and their clustering purity is more than twice higher with entropy values about 1.8. We show that the Sammon’s map algorithm is faster and gives a slightly better representation of the data set if the initial conditions are taken from the ICA projection rather than the PCA projection. We conclude that the nonlinear Sammon’s map projection is the best technique for combining data visualization and clustering assessment of the ChemCam LIBS data in 2D. PCA and ICA projections on more dimensions would improve on these numbers at the cost of the intuitive interpretation of the 2D projection by a human operator.     

Strategies for Mars remote Laser-Induced Breakdown Spectroscopy analysis of sulfur in geological samples

The key to understanding the sulfur history on Mars is to identify and determine sulfate and sulfide compositions and then to draw from them geologic clues about their environments of formation. To lay a foundation for use of remote LIBS to sulfur analysis in planetary exploration, we have undertaken a focused study of sulfur LIBS in geological samples in a simulated Mars atmosphere, with experimental parameters replicating the ChemCam LIBS instrument. A suite of twelve samples was selected, including rocks rich in minerals representative of sulfates and sulfides that might be encountered on Mars. Univariate analysis of sulfur emission lines did not provide quantitative information. Partial least squares (PLS) analysis was successful at modeling sulfur concentrations for a subset of samples with similar matrices. Sulfide minerals were identified on the basis of other siderophile or chalcophile peaks, such as those arising from Zn and Cu. Because the S lines are very weak compared to those of other elements, optimal PLS results were obtained by restricting the wavelength range to channels close to the most intense sulfur lines ~ 540-570 nm. Principal components analysis was attempted on the dataset, but did not differentiate the samples into meaningful groups because the sulfur lines are not strong enough. However, areas of the relatively weak S, H, and O peaks may be used to correctly classify all samples. Based on these outcomes, a flowchart that outlines a possible decision tree for identification and quantification of sulfur in remote LIBS analysis was constructed. Results suggest that LIBS data acquired under Mars conditions can meet the science requirements for the ChemCam instrument.     

Onboard calibration igneous targets for the Mars Science Laboratory Curiosity rover and the Chemistry Camera laser induced breakdown spectroscopy instrument

Accurate characterization of the Chemistry Camera (ChemCam) laser-induced breakdown spectroscopy (LIBS) on-board composition targets is of prime importance for the ChemCam instrument. The Mars Science Laboratory (MSL) science and operations teams expect ChemCam to provide the first compositional results at remote distances (1.5–7 m) during the in situ analyses of the Martian surface starting in 2012. Thus, establishing LIBS reference spectra from appropriate calibration standards must be undertaken diligently. Considering the global mineralogy of the Martian surface, and the possible landing sites, three specific compositions of igneous targets have been determined. Picritic, noritic, and shergottic glasses have been produced, along with a Macusanite natural glass. A sample of each target will fly on the MSL Curiosity rover deck, 1.56 m from the ChemCam instrument, and duplicates are available on the ground. Duplicates are considered to be identical, as the relative standard deviation (RSD) of the composition dispersion is around 8%. Electronic microprobe and laser ablation inductively coupled plasma mass spectrometry (LA ICP-MS) analyses give evidence that the chemical composition of the four silicate targets is very homogeneous at microscopic scales larger than the instrument spot size, with RSD < 5% for concentration variations > 0.1 wt.% using electronic microprobe, and < 10% for concentration variations > 0.01 wt.% using LA ICP-MS. The LIBS campaign on the igneous targets performed under flight-like Mars conditions establishes reference spectra for the entire mission. The LIBS spectra between 240 and 900 nm are extremely rich, hundreds of lines with high signal-to-noise, and a dynamical range sufficient to identify unambiguously major, minor and trace elements. For instance, a first LIBS calibration curve has been established for strontium from [Sr] = 284 ppm to [Sr] = 1480 ppm, showing the potential for the future calibrations for other major or minor elements.     

Efficient Implementation of High Dimensional Model Representations

Physical models of various phenomena are often represented by a mathematical model where the output(s) of interest have a multivariate dependence on the inputs. Frequently, the underlying laws governing this dependence are not known and one has to interpolate the mathematical model from a finite number of output samples. Multivariate approximation is normally viewed as suffering from the curse of dimensionality as the number of sample points needed to learn the function to a sufficient accuracy increases exponentially with the dimensionality of the function. However, the outputs of most physical systems are mathematically well behaved and the scarcity of the data is usually compensated for by additional assumptions on the function (i.e., imposition of smoothness conditions or confinement to a specific function space). High dimensional model representations (HDMR) are a particular family of representations where each term in the representation reflects the individual or cooperative contributions of the inputs upon the output. The main assumption of this paper is that for most well defined physical systems the output can be approximated by the sum of these hierarchical functions whose dimensionality is much smaller than the dimensionality of the output. This ansatz can dramatically reduce the sampling effort in representing the multivariate function. HDMR has a variety of applications where an efficient representation of multivariate functions arise with scarce data. The formulation of HDMR in this paper assumes that the data is randomly scattered throughout the domain of the output. Under these conditions and the assumptions underlying the HDMR it is argued that the number of samples needed for representation to a given tolerance is invariant to the dimensionality of the function, thereby providing for a very efficient means to perform high dimensional interpolation. Selected applications of HDMR's are presented from sensitivity analysis and time-series analysis.     

Laser induced breakdown spectroscopy library for the Martian environment

The NASA Mars Science Laboratory rover will carry the first Laser Induced Breakdown Spectroscopy experiment in space: ChemCam. We have developed a laboratory model which mimics ChemCam's main characteristics. We used a set of target samples relevant to Mars geochemistry, and we recorded individual spectra. We propose a data reduction scheme for Laser Induced Breakdown Spectroscopy data incorporating de-noising, continuum removal, and peak fitting. Known effects of the Martian atmosphere are confirmed with our experiment: better Signal-to-Noise Ratio on Mars compared to Earth, narrower peak width, and essentially no self-absorption. The wavelength shift of emission lines from air to Mars pressure is discussed. The National Institute of Standards and Technology vacuum database is used for wavelength calibration and to identify the elemental lines. Our Martian database contains 1336 lines for 32 elements: H, Li, Be, B, C, N, O, F, Na, Mg, Al, Si, P, S, Cl, K, Ar, Ca, Ti, V, Cr, Mn, Fe, Ni, Cu, Zn, As, Rb, Sr, Cs, Ba, and Pb. It is a subset of the National Institute of Standards and Technology database to be used for Martian geochemistry. Finally, synthetic spectra can be built from the Martian database. Correlation calculations help to distinguish between elements in case of uncertainty. This work is used to create tools and support data for the interpretation of ChemCam results.     

Evaluation of a compact spectrograph for in-situ and stand-off Laser-Induced Breakdown Spectroscopy analyses of geological samples on Mars missions

Laser-induced Breakdown Spectroscopy (LIBS) is actively under development for future use on surface probes to Mars. The analytical method can be deployed for in-situ and/or stand-off analysis with the latter embodiment providing the greatest advantages compared to previous and current elemental analysis methods used for planetary surface analysis. For this application, LIBS must be thoroughly investigated in terms of analytical capabilities and flight-rated instruments must be developed. Because of the low pressure of the predominantly CO₂ atmosphere on Mars, studies are needed to understand analytical requirements and to determine performance under these conditions. Stand-off analysis demands the most stringent requirements on instrumentation. Therefore, it must be determined if the high performance components that are normally used in a typical LIBS laboratory setup, which are generally not optimized for small size and weight, are essential to obtain the maximum scientific return from a mission. A key component of a LIBS apparatus is the detection system consisting of a spectrograph and a detector. Here we present an evaluation of one design of a compact spectrograph (Ocean Optics HR2000) for in-situ and stand-off LIBS analyses of geological samples under Mars atmospheric conditions.     

Comparative study of different methodologies for quantitative rock analysis by Laser-Induced Breakdown Spectroscopy in a simulated Martian atmosphere

Laser-Induced Breakdown Spectroscopy was selected by NASA as part of the ChemCam instrument package for the Mars Science Laboratory rover to be launched in 2009. ChemCam's Laser-Induced Breakdown Spectroscopy instrument will ablate surface coatings from materials and measure the elemental composition of underlying rocks and soils at distances from 1 up to 10 m. The purpose of our studies is to develop an analytical methodology enabling identification and quantitative analysis of these geological materials in the context of the ChemCam's Laser-Induced Breakdown Spectroscopy instrument performance. The study presented here focuses on several terrestrial rock samples which were analyzed by Laser-Induced Breakdown Spectroscopy at an intermediate stand-off distance (3 m) and in an atmosphere similar to the Martian one (9 mbar CO₂). The experimental results highlight the matrix effects and the measurement inaccuracies due to the noise accumulated when low signals are collected with a detector system such as an Echelle spectrometer equipped with an Intensified Charge-Coupled Device camera. Three different methods are evaluated to correct the matrix effects and to obtain quantitative results: by using an external reference sample and normalizing to the sum of all elemental concentrations, by using the internal standardization by oxygen, a major element common to all studied matrices, and by applying the Calibration Free Laser-Induced Breakdown Spectroscopy method. The three tested methods clearly demonstrate that the matrix effects can be corrected merely by taking into account the difference in the amount of vaporized atoms between the rocks, no significant variation in plasma excitation temperatures being observed. The encouraging results obtained by the three methods indicate the possibility of meeting ChemCam project objectives for stand-off quantitative analysis on Mars.     

Laser-Induced Breakdown Spectroscopy for Mars surface analysis: capabilities at stand-off distances and detection of chlorine and sulfur elements

An international consortium is studying the feasibility of performing in situ geochemical analysis of Mars soils and rocks at stand-off distances up to several meters using the Laser-Induced Breakdown Spectroscopy (LIBS) technique. Stand-off analysis for Martian exploration imposes particular requirements on instrumentation, and it is necessary to first test the performance of such a system in the laboratory. In this paper, we test the capabilities of two different experimental setups. The first one is dedicated to the qualitative analysis of metals and rocks at distances between 3 and 12 m. With the second one, we have obtained quantitative results for aluminum alloys and developed a spectral database under Martian conditions for sulfur and chlorine, two elements that are geologically interesting but generally difficult to detect by LIBS under standard conditions (atmospheric pressure, close distance). These studies were carried out to determine an optimal instrumental design for in situ Mars analysis. The quality of analytical results affected by the optical elements and spectrometer has been particularly highlighted.     

One-point calibration for calibration-free laser-induced breakdown spectroscopy quantitative analysis

We present a new method for improving the reliability of quantitative analysis by laser-induced breakdown spectroscopy (LIBS). The method can be considered as a variation of the calibration-free LIBS approach; although not completely standard-less, only one standard of known composition and similar matrix to the one to be analyzed is needed. On the other hand, the one-point calibration approach allows the empirical determination of essential experimental and spectroscopic parameters, whose knowledge is often imprecise or lacking; the result is a definite improvement of the trueness of LIBS analysis with respect to the traditional calibration-free approach.The characteristics and advantages of the proposed one-point calibration LIBS approach will be demonstrated on a set of copper-based samples of known composition.     

Supervised neighborhood preserving embedding for feature extraction and its application for soft sensor modeling

Neighborhood preserving embedding (NPE) is a useful tool for learning the manifold of high‐dimensional data. As a linear approximation of nonlinear locally linear embedding, NPE can be applied to dimensionality reduction by neighborhood preserving. However, the original NPE algorithm is an unsupervised method, which extracts features without any reference to the output information. In this paper, a supervised NPE framework is proposed for output‐related feature extraction in soft sensor applications. In the supervised NPE framework, the output information is utilized to guide the procedures for constructing the adjacent graph and calculating the weight matrix, with which the intrinsic structure of the data can be better described. For performance evaluation of the proposed method, experiments on a numerical example and an industrial debutanizer column process are carried out. The results show the effectiveness of the proposed framework.     

Locally linear embedding method for dimensionality reduction of tissue sections of endometrial carcinoma by near infrared spectroscopy

Locally linear embedding (LLE) is introduced here as a nonlinear compression method for near infrared reflectance spectra of endometrial tissue sections. The LLE has been evaluated by using support vector machine (SVM) classifiers and the projected difference resolution (PDR) method. Synthetic data sets devised to resemble near-infrared spectra of tissue samples were used to characterize the performance of the LLE. The LLE was compared using principal component compression (PCC) method to evaluate nonlinear and linear compression. For a set of real tissue samples, if the compressed data were not range-scaled prior to SVM classification, the principal component compressed data gave an average prediction rate of 39 ± 2% while the LLE 94 ± 2%; if range-scaled after compression, the LLE and PCC performed evenly, with maximum average prediction values of 94 ± 2% and 93 ± 2%, respectively. The SVM without compression yielded a classification rate of 92 ± 2%. The prediction accuracy was consistent with PDR results. Without the second derivative preprocessing, the classification rates were 90 ± 3%, 89 ± 2%, and 78 ± 2% for the LLE compressed, the PCC, and no compression classifications by the SVM, respectively.     

Laser-Induced Breakdown Spectroscopy in open-path configuration for the analysis of distant objects

A review of recent results on stand-off Laser-Induced Breakdown Spectroscopy (LIBS) analysis and applications is presented. Stand-off LIBS was suggested for elemental analysis of materials located in environments where any physical access was not possible but optical access could be envisaged. This review only refers to the use of the open-path LIBS configuration in which the laser beam and the returning plasma light are transmitted through the atmosphere. It does not present the results obtained with a transportation of the laser pulses to the target through an optical fiber. Open-path stand-off LIBS has mainly been used with nanosecond laser pulses for solid sample analysis at distances of tens of meters. Liquid samples have also been analyzed at distances of a few meters. The distances achievable depend on many parameters including the laser characteristics (pulse energy and power, beam divergence, spatial profile) and the optical system used to focus the pulses at a distance. A large variety of laser focusing systems have been employed for stand-off analysis comprising refracting or reflecting telescope. Efficient collection of the plasma light is also needed to obtain analytically useful signals. For stand-off LIBS analysis, a lens or a mirror is required to increase the solid angle over which the plasma light can be collected. The light collection device can be either at an angle from the laser beam path or collinear with the optical axis of the system used to focus the laser pulses on the target surface. These different configurations have been used depending on the application such as rapid sorting of metal samples, identification of material in nuclear industry, process control and monitoring in metallurgical industry, applications in future planetary missions, detection of environmental contamination or cleaning of objects of cultural heritage. Recent stand-off analyses of metal samples have been reported using femtosecond laser pulses to extend LIBS capabilities to very long distances. The high-power densities achievable with these laser pulses can also induce self-guided filaments in the atmosphere which produce LIBS excitation of a sample. The first results obtained with remote filament-induced breakdown spectroscopy predict sample analysis at kilometer ranges.     

Laser-induced breakdown spectroscopy for space exploration applications: Influence of the ambient pressure on the calibration curves prepared from soil and clay samples

Recently, there has been an increasing interest in the laser-induced breakdown spectroscopy (LIBS) technique for stand-off detection of geological samples for use on landers and rovers to Mars, and for other space applications. For space missions, LIBS analysis capabilities must be investigated and instrumental development is required to take into account constraints such as size, weight, power and the effect of environmental atmosphere (pressure and ambient gas) on flight instrument performance. In this paper, we study the in-situ LIBS method at reduced pressure (7 Torr CO₂ to simulate the Martian atmosphere) and near vacuum (50 mTorr in air to begin to simulate the Moon or asteroids' pressure) as well as at atmospheric pressure in air (for Earth conditions and comparison). Here in-situ corresponds to distances on the order of 150 mm in contrast to stand-off analysis at distance of many meters. We show the influence of the ambient pressure on the calibration curves prepared from certified soil and clay pellets. In order to detect simultaneously all the elements commonly observed in terrestrial soils, we used an Echelle spectrograph. The results are discussed in terms of calibration curves, measurement precision, plasma light collection system efficiency and matrix effects.     

Study of sub-mJ-excited laser-induced plasma combined with Raman spectroscopy under Mars atmosphere-simulated conditions

Laser-Induced Breakdown Spectroscopy (LIBS) and Raman spectroscopy are complimentary techniques. LIBS yields elemental information while Raman spectroscopy yields molecular information about a sample, and both share similar instrumentation configurations. The combination of LIBS and Raman spectroscopy in a single instrument for planetary surface exploration has been proposed, however challenges exist for developing a combined instrument. We present LIBS and Raman spectroscopy results obtained using a diode pumped, intracavity doubled, Q-switched, Nd:YLF laser operating at 523 nm, which overcomes some of the difficulties associated with a combined instrument. LIBS spectra were obtained with 170 μJ per pulse at 4 Hz repetition rate in a low pressure Mars-simulated atmosphere and Raman spectra produced with 200 mW at 100 kHz. The Nd:YLF laser is switchable between LIBS and Raman spectroscopy modes only by a change in Q-switch repetition rate. Emissions from Ca, Ca II, Fe, Fe II, Mg, Na, and O atom were identified in the μ-LIBS spectrum of oolithic hematite. Evidence was found for a change in plasma dynamics between 7 and 5 Torr that could be explained as a decrease in plasma temperature and electron density below 5 Torr. This is relevant to future Mars exploration using LIBS as the mean surface pressure on Mars varies from 3.75 to 6 Torr. LIBS plasma dynamics should be carefully evaluated at the pressures that will be encountered at the specific Mars landing site.     

Correlations Between Laser Induced Breakdown Spectroscopy (LIBS) and Dynamical Mechanical Analysis (DMA) for Assessment of Aging Effect on Plastic Bonded Explosives (PBX)

Plastic Bonded Explosives (PBXs) are explosive materials, in which sensitive explosive powders with high detonation performance are bound together in a matrix using small quantities of a synthetic polymer. They provide good thermal stability and improve cook‐off properties with negligible shrinkage. Laser‐induced breakdown spectroscopy (LIBS) has been introduced as a novel method for assessment of thermal aging in PBXs. Thermal aging of fifteen samples of PBXs with the same composition has been studied in two temperatures 60 and 70 °C for 5 period times during 29 d. Dynamical mechanical analysis (DMA) has also been done on different aged samples in order to validate the LIBS results. Among different molecular bands, LIBS data showed that the changes in line intensity in CN and AlO molecular radiations could be used to assess aging effects. Principle component analysis (PCA) chemometric method was also implemented in order to obtain more accurate discrimination between different samples. Since there is a good agreement for the outputs of LIBS data and DMA results, LIBS technique can be introduced a reliable method for assessment of thermal aging of PBXs.     

Parametric study of a fiber-optic laser-induced breakdown spectroscopy probe for analysis of aluminum alloys

In the present work we demonstrate a fiber-optic laser-induced breakdown spectroscopy (FO LIBS) system for delivering laser energy to a sample surface to produce a spark as well as to collect the resulting radiation from the laser-induced spark. In order to improve the signal/background (S/B) ratio, various experimental parameters, such as laser energy, gate delay and width, detector gain, lenses of different focal lengths and sample surface, were tested. In order to provide high reliability and repeatability in the analysis, we also measured plasma parameters, such as electron density and plasma temperature, and determined their influence on the measurement results. The performance of FO LIBS was also compared with that of a LIBS system that does not use a fiber to transmit the laser beam. LIBS spectra with a good S/B were recorded at 2-μs gate delay and width. LIBS spectra of six different Al alloy samples were recorded to obtain calibration data. We were able to obtain linear calibration data for numerous elements (Cr, Zn, Fe, Ni, Mn, Mg and Cu). A linear calibration curve for LIBS intensity ratio vs. concentration ratio reduces the effect of physical variables (i.e. shot-to-shot power fluctuation, sample-to-surface distance, and physical properties of the samples). Our results reveal that this system may be useful in designing a high-temperature LIBS probe for measuring the elemental composition of Al melt.     

Use of the vacuum ultraviolet spectral region for laser-induced breakdown spectroscopy-based Martian geology and exploration

Several elements important to planetary geology (e.g. Br, C, Cl, P, S) and the human exploration of Mars (e.g. toxic elements such as As) have strong emission lines in the purge and vacuum ultraviolet (VUV) spectral region (100–200 nm). This spectral region has not been extensively studied for space applications using geological samples. We studied emissions from the laser-induced breakdown spectroscopy (LIBS) plasma in this region using a sample chamber filled with 7 torr (930 Pa) of CO₂ to simulate the Martian atmosphere. Pressures down to 0.02 torr were also used to evaluate the effect of the residual CO₂ on the spectra and to begin investigating the use of VUV-LIBS for airless bodies such as asteroids and the Moon. Spectra were recorded using a 0.3-m vacuum spectrometer with an intensified CCD (ICCD) camera. The effects of time delay and laser energy on LIBS detection at reduced pressure were examined. The effect of ambient CO₂ on the detection of C in soil was also evaluated. Lines useful for the spectrochemical analysis of As, Br, C, Cl, P, and S were determined and calibration curves were prepared for these elements. Although LIBS is being developed for stand-off analysis at many meters distance, the experiments reported here were aimed at in-situ (close-up) analysis.     

Laser induced breakdown spectroscopy on soils and rocks: Influence of the sample temperature,moisture and roughness

ExoMars, ESA's next mission to Mars, will include a combined Raman/LIBS instrument for the comprehensive in-situ mineralogical and elemental analyses of Martian rocks and soils. It is inferred that water exists in the upper Martian surface as ice layers, “crystal” water or adsorbed pore water. Thus, we studied Laser Induced Breakdown Spectroscopy (LIBS) on wet and dry rocks under Martian environmental conditions in the temperature range − 60 °C to + 20 °C and in two pressure regimes, above and below the water triple point. Above this point, the LIBS signals from the rock forming elements have local minima that are accompanied by hydrogen (water) emission maxima at certain temperatures that we associate with phase transitions of free or confined water/ice. At these sample temperatures, the plasma electron density and its temperature are slightly lowered. In contrast to powder samples, a general increase of the electron density upon cooling was observed on rock samples. By comparing the LIBS signal behavior from the same rock with different grades of polishing, and different rocks with the same surface treatment, it was possible to distinguish between the influence of surface roughness and the bulk material structure (pores and grains). Below the triple point of water, the LIBS signal from the major sample elements is almost independent of the sample temperature. However, at both considered pressures we observed a hydrogen emission peak close to − 50 °C, which is attributed to a phase transition of supercooled water trapped inside bulk pores.