Elsevier

Icarus

Volume 365, 1 September 2021, 114481
Icarus

Research Paper
Laser-Induced Breakdown Spectroscopy (LIBS) characterization of granular soils: Implications for ChemCam analyses at Gale crater, Mars

https://doi.org/10.1016/j.icarus.2021.114481Get rights and content

Highlights

  • Mechanical mixtures and grain coatings were analyzed using LIBS in a martian atmosphere.

  • Grain size and concentration were varied to study their influences on LIBS signal.

  • We show that mixed grains or coatings have different effects on LIBS analyses.

  • We provide constraints on the distribution of the amorphous component in martian soil.

Abstract

The Curiosity rover has been characterizing mineralogical and chemical compositions of Gale crater soils on Mars since 2012. Given its sub-millimeter scale of analysis, the ChemCam instrument is well suited to study the composition of soil constituents. However, the interpretation of LIBS data on soils in the martian environment is complicated by the large diversity of particle sizes (from dust to sand), combined with the unknown physical arrangement of their mineral constituents (i.e., the type of grain mixtures). For example, martian soils contain a significant amount of X-ray amorphous materials whose physical form remains unclear. In this study, we reproduced martian soil analyses in the laboratory to understand how the LIBS technique can provide specific insights into the physical and chemical properties of granular soils. For this purpose, different types of samples were studied with various ranges of grain sizes, mimicking two possible mixtures that may occur in martian soils: mechanical mixtures of two populations of grains made of distinct chemical compositions; and material forming a compositionally distinct coating at the surface of grains. Our results, also supported by in situ ChemCam data, demonstrate that both the sizes and the type of mixture of soil particles have a strong influence on the LIBS measurement. For mechanical mixtures of two populations of grains larger than 125–250 μm, the scatter of the data provides information about the chemical composition of the end-members. On the other hand, the chemistry recorded by LIBS for grains with surface coatings is fully dominated by the outer material for grains smaller than 500 μm in diameter. This is due to the small penetration depth of the laser (~0.3–1.5 μm per shot), combined with the ejection of small grains at each shot, which leads to a constant replenishment of fresh material. This experimental work will thus improve our understanding of martian soils analyzed by ChemCam, and more broadly, will benefit LIBS studies of granular materials.

Introduction

The surface of Mars is globally dominated by basaltic rocks, as successively shown by telescopic, orbital and in situ observations (e.g., Pinet and Chevrel, 1990; Christensen et al., 2000; Bandfield, 2002; McSween et al., 2004; Gasnault et al., 2010). Over time, these primary materials have undergone physical and chemical weathering, and have been partially turned into loose sediments, referred to as “soils” on Mars. Soils are considered as representative of the average martian crust after correcting for meteoric contributions (Taylor and McLennan, 2009, see Table 1). Indeed, they are thought to have sampled broad areas and large volumes of the exposed martian crust that have undergone variable degrees of chemical weathering and aqueous alteration as they are aggregates of local, regional, and global sources. In the context of in situ missions, the chemical composition of soils is accessible to rover analyses and they are particularly useful to characterize the alteration of the crust. Indeed, soil compositions may differ from pristine basalts due to the addition of volcaniclastic materials (that may be from different igneous provenances; Le Deit et al., 2016; Bedford et al., 2019) and secondary phases (McGlynn et al., 2012, and references therein). Due to their small size and large reactive surfaces compared to rocks, soils are also particularly effective in recording aqueous alteration processes during pedogenesis (McSween et al., 2010). The term “soil” in the context of martian studies is free of any biologic connotation and simply refers to all unconsolidated loose material, which can be distinguished from hardened rocks (Meslin et al., 2013). They consist of surficial deposits of fine-grained components mixed with rock fragments (Certini et al., 2020). It is important to make a distinction between soils and dust, although the boundary between both is sometimes difficult to define. It is mainly a particle size criterion. Dust often refers to particles smaller than a few micrometers in size that can be easily mobilized by the wind (Lasue et al., 2018), whereas the more general concept of soil encompasses micrometer- to millimeter-sized grains.

As seen from the ground, dust and soils seem to share a relatively similar chemical composition (Berger et al., 2016). Relative to soils, the dust has higher sulfur and chlorine abundances, but the S/Cl ratios are comparable (~3.7), suggesting that dust and soil formation may be genetically linked (Berger et al., 2016). Dust formation probably implies wind abrasion of soft rock units and the eolian breakdown of saltating sand-sized particles (e.g., Ohja et al., 2018). However, dust is not the most chemically altered component of the martian soils, and although dust may be a contributor to the amorphous component found in soils, some differences in composition indicate that the two materials are not equivalent (Lasue et al., 2018).

The chemical composition of martian soils seems to be relatively homogeneous among the different landing sites. Indeed, the two Viking landers first highlighted the similar bulk soil compositions of the Chryse Planitia and Utopia Planitia soils (Clark et al., 1977). Apart from slightly lower alkaline elements and higher silica, Sojourner (Mars Pathfinder) in Ares Vallis also measured a similar composition (Brückner et al., 2003; Foley et al., 2003). The Fe and Mg enrichments in soils compared to the neighboring bedrocks suggested that soil compositions are not entirely controlled by alteration of local rocks in this region. The Gusev crater and Meridiani Planum landing sites, respectively studied by the Spirit and Opportunity rovers, also revealed very concordant basaltic compositions (Gellert et al., 2004; Yen et al., 2005), consistent with orbital data from the Gamma-ray spectrometer onboard Mars Odyssey (Newsom et al., 2007). Consequently, soils seem to be quite uniform over these different landing sites with an influence of local rock chemistry (e.g., Yen et al., 2005). The relative chemical homogeneity of soils could be due to large-scale eolian mixing, or/and it could also result from a global homogeneity of the source regions. Recent works based on orbital data have shown a difference in soil S, Cl and H chemistries, especially between the northern lowlands and southern highlands, suggesting different levels of interaction with water (Hood et al., 2019), which argues against complete mixing at a global scale.

Increasing our knowledge of soil compositions is among the scientific objectives of the Mars Science Laboratory mission (Grotzinger et al., 2012) and its rover Curiosity. Since 2012, it has been investigating Gale crater, located south of Elysium Planitia, in the border of the Aeolis Mensae region. Its scientific payload includes tools capable of analyzing both the mineralogy and chemistry of soils. The bulk chemical composition of soils in Gale crater, measured with the Alpha Particle X-ray Spectrometer (APXS) instrument is within 2σ of the Mars Exploration Rovers (MER) analyses, except for TiO2 (Blake et al., 2013; Morris et al., 2013; O'Connell-Cooper et al., 2017), suggesting relatively similar basaltic soils. The CheMin instrument (Blake et al., 2012), the first X-ray Diffractometer (XRD) operating on the martian surface, showed that all major mineralogical phases of the fine fraction (<150 μm) of the Rocknest eolian soil (the Gale crater soil reference) are consistent with a basaltic source (Bish et al., 2013; Achilles et al., 2017; Morrison et al., 2018). CheMin detected igneous minerals like plagioclase (~An50), olivine (~Fo58) and pyroxenes (augite and pigeonite), but very few secondary products were identified and quantified. Only anhydrite (0.9 ± 0.2 wt%), magnetite (1.8 ± 0.3 wt%) and hematite (1.0 ± 0.1 wt%) were observed in the X-ray diffractograms, but no clay minerals (Bish et al., 2013; Achilles et al., 2017, Morrison et al., 2018). However, CheMin analyses also revealed that 35 ± 15 wt% of the Rocknest soil is X-ray amorphous, i.e., material that is insufficiently crystallized to be identifiable with the XRD method. The presence of amorphous phases has long been suspected on the surface of Mars (e.g., Evans and Adams, 1979; Singer, 1985; Morris et al., 2000; Milliken et al., 2008; Squyres et al., 2008), but CheMin provided the first in situ evidence with XRD data (Bish et al., 2013; Rampe et al., 2014). The composition of the amorphous component can be estimated with a mass balance calculation, i.e., by subtracting the chemistry of minerals identified by CheMin from the bulk chemistry measured by APXS. At Rocknest, this mass balance calculation suggests a silica-poor amorphous component (~35 wt% SiO2), and enrichments in iron (~23 wt% FeOT), sulfur (~14 wt% SO3) and possibly phosphorus (~3 wt% P2O5) compared to the bulk crystalline composition (Blake et al., 2013; Dehouck et al., 2014; Achilles et al., 2017). The SAM (Sample Analysis at Mars) instrument (Mahaffy et al., 2012), which contains a gas chromatograph paired with a quadrupole mass spectrometer, recorded volatiles released during the heating of the Rocknest soil samples up to ~835 °C, including H2O, SO2, CO2 and O2 (Leshin et al., 2013). The range of release temperatures suggests that water is bound to the amorphous component. The CheMin instrument did not detect any hydrated mineral in the soil, which also suggests that H2O is associated with the amorphous component.

ChemCam, the Laser-Induced Breakdown Spectroscopy (LIBS) instrument (Wiens et al., 2012; Maurice et al., 2012), enables elementary chemistry to be determined at a sub-millimeter scale, and so is a relevant tool to constrain the chemical constituents of soils. The laser beam has a sampling area that ranges from 350 to 550 μm diameter, depending on the distance to the target. This resolution allows us to directly isolate soil components at or above some size scale, thus providing the first discrimination between different populations of grains (Meslin et al., 2013; Cousin et al., 2015). Near the landing area, some grains coarser than the laser beam were found to have predominantly a felsic composition (enriched in Si, Al, Na, and K). These grains have the same composition as rocks encountered in the Bradbury landing site and so might come from locally eroded rocks (Sautter et al., 2014; Meslin et al., 2013; Cousin et al., 2015). The second population contains grains that are both coarser and finer than the laser sampling area, with a mafic composition (enriched in Fe and Mg). The Mn, Cr and Mg signals are lower for these grains than in the neighboring rocks, suggesting a lack of direct genetic connection between these soil grains and local mafic rocks (Meslin et al., 2013; Cousin et al., 2015). ChemCam data recorded a stronger hydrogen emission peak linked to the mafic part, which may be associated with weathering products among the non-crystalline components (Meslin et al., 2013). Consequently, constraining the nature and chemistry of these hydrated amorphous phases is important to better understand the aqueous history of these sediments.

The Bagnold Dunes, located near the base of Mount Sharp, are the first active dunes studied in situ on another planet (Bridges and Ehlmann, 2018). Two samples were scooped at different locations: Gobabeb on Sol 1225, and Ogunquit Beach on Sol 1650 (Fig. 1), corresponding to the upwind and downwind margins of the dune field. The CheMin instrument was used to determine their mineralogies. They are relatively similar to that of the Rocknest sand with a crystalline part mostly composed of igneous basaltic minerals (plagioclase, olivine, augite, and pigeonite), with minor abundances of hematite, magnetite, quartz, and anhydrite (Achilles et al., 2017; Rampe et al., 2018; Morrison et al., 2018). Similarly to Rocknest, the Gobabeb and Ogunquit Beach samples have substantial fractions of amorphous materials, with respectively 35 ± 15 (Achilles et al., 2017) and 40 ± 15 wt% (Rampe et al., 2018) of the bulk mineralogy. The X-ray amorphous component compositions show some differences between Gobabeb and Ogunquit Beach, especially in SiO2 and FeOT abundances that could be a result of different local sediment sources (Rampe et al., 2018). Second-order mixing of local bedrock contributions and mineral sorting were also observed through subtle mineralogical variation between the two locations, especially in olivine and plagioclase abundances (Achilles et al., 2017; Lapotre et al., 2017; Ehlmann et al., 2017; Rampe et al., 2018). Several scoops of the Gobabeb samples were sieved to supply portions with controlled grain sizes to the SAM and CheMin instruments. Leftovers with relatively well-constrained ranges of grain sizes were dumped onto the ground. The Sample Acquisition/Sample Processing and Handling (Sa/SPaH) subsystem of the rover (Anderson et al., 2012) delivered four piles of the Bagnold sand: dump A (<150 μm), dump B (>150 μm), dump C (150 μm–1 mm) and dump D (>1 mm). ChemCam measurements were performed on these four different dumps (Cousin et al., 2017). Grain size distributions were also measured with ChemCam RMI images and MAHLI (Mars Hand Lens Imager) images, the latter of which has a spatial resolution better than 19 μm per pixel at a ~ 1 cm standoff distance (Edgett et al., 2013). These images reveal that no grains larger than 500 μm are detectable in piles B and C, and, as a matter of fact, no pile was formed at the expected dump D location, due to the absence of grains larger than 1 mm (Ehlmann et al., 2017). ChemCam analyses in these dump piles reveal that at least two different chemical components are present in the Bagnold Dunes. Indeed, Cousin et al. (2017) show a mixing line between two end-members: the first one is enriched in Al, Na, Si, and Ca suggesting the presence of felsic compounds, while the other one is depleted in these elements, suggesting the presence of mafic minerals, which is in agreement with CheMin mineralogy. Finally, although the hydrogen abundances recorded in the dunes were lower compared to the Rocknest soil samples (respectively 1.5–3 and ~ 1 wt%, Leshin et al., 2013; Sutter et al., 2017; Gabriel et al., 2018), the amorphous component of the dunes is also thought to hold the H2O content (Cousin et al., 2017; Sutter et al., 2017; Achilles et al., 2017; Ehlmann et al., 2017; Rampe et al., 2018). Similarly to Rocknest soils, constraining the origin and nature of the hydrated amorphous phases is an important step to understand the aqueous alteration processes that made the history of these soil deposits.

Direct analyses of the amorphous component by ChemCam require first to understand the complexity of the interaction between the laser beam and the granular materials. The goal of this study is to conduct analyses of martian soil simulants with a replica of the ChemCam instrument in the laboratory, where experimental parameters and end-member compositions are well constrained, to better understand ChemCam flight data. Only a few studies were done on LIBS performance on non-indurated targets (see for example the review by Harmon et al., 2013), and even less under martian atmosphere conditions. Indeed, ChemCam represents the first use of the LIBS technique on another planet, and the interpretation of ChemCam data of soils is complicated in the martian environment because of the large diversity of grain sizes and the unknown physical form (i.e., the kind of grain mixtures) of soil particles, including the amorphous component. Here, we studied the interaction of the laser beam with two kinds of mixtures that occur or may occur in martian soils. The first kind corresponds to mechanical mixtures of two populations of grains of distinct chemical compositions (Fig. 2, left panel). This may be applied to two populations of crystalline grains (e.g., mafic and felsic minerals, Meslin et al., 2013; Cousin et al., 2015), but also to a population of amorphous grains mingled with crystalline grains. The second kind of mixtures that we studied corresponds to a material forming a compositionally-distinct coating at the surface of grains (Fig. 2, right panel). Such a coating may be formed by the precipitation of salts from a saturated solution, for which grains can serve as nuclei. Alternatively, grains may also develop alteration rinds due to aqueous weathering processes, or get coated by dust during their transport due to electrostatic forces. Here again, the coating may be crystalline or amorphous in nature. In martian soils, these two kinds of mixtures can occur simultaneously, for example, with the mixing of primary grains, grains with secondary coatings, and fragments derived from the coatings. In addition to chemistry, constraining the physical form of the non-crystalline part is important because it can yield information about its origin and consequently result in a better understanding of the geology and aqueous history of these sediments. Dust and precipitated coatings on indurated rocks were studied previously for ChemCam data interpretation purposes (Graff et al., 2011; Lanza et al., 2015) but not with the aim of understanding soil properties. The objective of our work is to show how the LIBS technique can provide specific insight into the physical and chemical properties of soils.

In addition to hydrogen, sulfur is also an important element to study aqueous alteration processes on the martian surface. Sulfur is present in abundance in martian soils (i.e., ~6.2 wt% SO3, Taylor and McLennan, 2009) mainly in the form of sulfates (e.g., King and McLennan, 2010). Sulfate minerals have been widely detected by remote-sensing (e.g., Gendrin et al., 2005; Langevin et al., 2005; Bibring et al., 2006), especially in the equatorial Valles Marineris canyon, in the adjacent Meridiani area, and in the circumpolar deposits. In situ analyses also reveal that sulfates correspond to abundant secondary products in martian soils. Magnesium, calcium and/or iron sulfates were identified at the Viking, Pathfinder, Phoenix and MER landing sites (e.g., Clark, 1993; Foley et al., 2003; Kounaves et al., 2010; Wang et al., 2006; Soderblom et al., 2004). In Gale crater soils, only anhydrite has been observed by CheMin, but the mass balance calculations indicate that sulfur is abundant in the amorphous component of the Rocknest soil with ~14 wt% SO3 (Achilles et al., 2017). Consequently, in this study, we also tested the sulfur LIBS emission peaks as a potential geochemical marker of S-bearing phases in martian granular media.

Section snippets

Preparation of martian analog samples

For the purpose of this study, we produced simple laboratory mixtures corresponding to the two model cases mentioned above: mechanical mixtures, and grains with a surface coating (Fig. 2). Preparations were made for several grain sizes, ranging from smaller to larger than the laser beam (~425 ± 25 μm, Chide et al., 2019). Although this is a simplification with respect to natural martian soils, we prepared only binary mixtures to facilitate data interpretation. To enable direct comparisons

Relationship between ICA scores and concentrations

The spectra of the three end-members used in our experiments are shown in Fig. 5a, illustrating their chemical differences. The Independent Component Analysis applied to these spectra successfully discriminated the end-member compositions. Fig. 5b represents the Fe ICA scores as a function of the Mg ICA scores obtained for the pure JSC-L, JSC-M, and magnesium sulfate samples. Although we observe some variability among the 150 shots of a given sample due to some chemical heterogeneity and

Sulfur detection

Magnesium sulfate used in our samples is assumed to contain from 0 to ~20 wt% H2O, considering partial and total hydration of the pristine kieserite powder into epsomite (~51 wt% H2O) during the experimental precipitation process and subsequent degassing of ~31 wt% H2O in the vacuum of the martian chamber (see Section 2.1.2). The lower limit expressed here is probably not realistic because a significant H signal is visible in the spectrum (Fig. 5a). Considering such hydration levels, our

Conclusion

The interpretation of the ChemCam soil measurements is made challenging by the complexity of the physical interaction between the laser and the grains, but also by the diversity of grain sizes and mineral phases found in martian soils. This study aimed to produce and analyze martian soil analogs in the laboratory to better understand the effects of all these parameters. We studied two kinds of mixtures (Fig. 2), each with several grain size ranges: (1) mechanical mixtures of two populations of

Declaration of Competing Interest

None.

Acknowledgments

We are grateful to the MSL science and engineering teams who operate the rover. ChemCam instrument development, operations and science support in the US were funded by the NASA Mars Exploration Program. ChemCam and MSL are supported in France by the Centre National d'Etudes Spatiales (CNES). All the ChemCam data used in this paper are released and can be found on the Planetary Data System (https://pds-geosciences.wustl.edu/missions/msl/index.htm). Laboratory LIBS experiments were conducted at

References (95)

  • R.S. Harmon et al.

    Applications of laser-induced breakdown spectroscopy for geochemical and environmental analysis: a comprehensive review

    Spectrochim. Acta B At. Spectrosc.

    (2013)
  • N.L. Lanza et al.

    Understanding the signature of rock coatings in laser-induced breakdown spectroscopy data

    Icarus

    (2015)
  • M. McCanta et al.

    Testing the veracity of LIBS analyses on Mars using the libssim program

    Planet. Space Sci.

    (2013)
  • W. Rapin et al.

    Roughness effects on the hydrogen signal in laser-induced breakdown spectroscopy

    Spectrochim. Acta B At. Spectrosc.

    (2017)
  • W. Rapin et al.

    Quantification of water content by laser-induced breakdown spectroscopy on Mars

    Spectrochim. Acta B At. Spectrosc.

    (2017)
  • F. Rivera-Hern’andez et al.

    Using ChemCam libs data to constrain grain size in rocks on Mars: proof of concept and application to rocks at Yellowknife Bay and Pahrump Hills, Gale crater

    Icarus

    (2019)
  • S. Schröder et al.

    Hydrogen detection with ChemCam at Gale crater

    Icarus

    (2015)
  • R.B. Singer

    Spectroscopic observation of Mars

    Adv. Space Res.

    (1985)
  • R. Wiens et al.

    Pre-flight calibration and initial data processing for the ChemCam laser-induced breakdown spectroscopy instrument on the Mars Science Laboratory rover

    Spectrochim. Acta B At. Spectrosc.

    (2013)
  • C.N. Achilles et al.

    Mineralogy of an active eolian sediment from the Namib dune, Gale crater, Mars

    J. Geophys. Res. Planets

    (2017)
  • C.C. Allen et al.

    JSC Mars-1: a martian soil simulant

    Space

    (1998)
  • R.C. Anderson et al.

    Collecting samples in Gale crater, Mars; an overview of the Mars science laboratory sample acquisition, sample processing and handling system

    Space Sci. Rev.

    (2012)
  • J.L. Bandfield

    Global mineral distributions on Mars

    J. Geophys. Res. Planets

    (2002)
  • J.A. Berger et al.

    A global Mars dust composition refined by the Alpha-Particle X-ray Spectrometer in Gale crater

    Geophys. Res. Lett.

    (2016)
  • J.-P. Bibring et al.

    Global mineralogical and aqueous Mars history derived from OMEGA/Mars Express data

    Science

    (2006)
  • D.L. Bish et al.

    X-ray diffraction results from Mars Science Laboratory: mineralogy of Rocknest at Gale crater

    Science

    (2013)
  • D. Blake et al.

    Characterization and calibration of the CheMin mineralogical instrument on Mars Science Laboratory

    Space Sci. Rev.

    (2012)
  • D.F. Blake et al.

    Curiosity at Gale crater, Mars: characterization and analysis of the Rocknest sand shadow

    Science

    (2013)
  • N.T. Bridges et al.

    The Mars Science Laboratory (MSL) Bagnold dunes campaign, Phase I: overview and introduction to the special issue: Bagnold dunes campaign overview

    J. Geophys. Res. Planets

    (2018)
  • J. Brückner et al.

    Refined data of alpha proton x-ray spectrometer analyses of soils and rocks at the Mars Pathfinder site: implications for surface chemistry

    J. Geophys. Res. Planets

    (2003)
  • J. Campbell et al.

    A fundamental parameters approach to calibration of the Mars Exploration Rover alpha particle x-ray spectrometer

    J. Geophys. Res. Planets

    (2009)
  • J.-F. Cardoso

    Infomax and maximum likelihood for blind source separation

    IEEE Signal Proc. Lett.

    (1997)
  • J.-F. Cardoso et al.

    Blind beamforming for non-gaussian signals

  • P.R. Christensen et al.

    Identification of a basaltic component on the martian surface from thermal emission spectrometer data

    J. Geophys. Res. Planets

    (2000)
  • B.C. Clark et al.

    The Viking x-ray fluorescence experiment: analytical methods and early results

    J. Geophys. Res.

    (1977)
  • S. Clegg et al.

    ChemCam sulfur quantitative analysis and interpretation

  • S. Clegg et al.

    ChemCam sulfur quantitative analysis of the Askival altered feldspathic cumulate target, gale crater, mars

  • F. Colao et al.

    LIBS application for analyses of martian crust analogues: search for the optimal experimental parameters in air and CO2 atmosphere

    Appl. Phys. A

    (2004)
  • A. Cousin et al.

    Geochemistry of the Bagnold dune field as observed by ChemCam and comparison with other aeolian deposits at Gale crater

    J. Geophys. Res. Planets

    (2017)
  • E. Dehouck et al.

    Constraints on abundance, composition, and nature of x-ray amorphous components of soils and rocks at Gale crater, Mars

    J. Geophys. Res. Planets

    (2014)
  • E. Dehouck et al.

    Independent characterization of the Amorphous Component of martian soils using ChemCam LIBS data

  • K. Edgett et al.

    Curiosity’s Mars hand lens imager (Mahli): Sol 0–179 activities, observations, range and scale characterization

  • J.A. Effenberger et al.

    Effect of atmospheric conditions on LIBS spectra

    Sensors

    (2010)
  • B. Ehlmann et al.

    Chemistry, mineralogy, and grain properties at Namib and high dunes, Bagnold dune field, Gale crater, Mars: a synthesis of Curiosity rover observations

    J. Geophys. Res. Planets

    (2017)
  • D.L. Evans et al.

    Comparison of Viking lander multispectral images and laboratory reflectance spectra of terrestrial samples

  • C.N. Foley et al.

    Final chemical results from the Mars pathfinder alpha proton x-ray spectrometer: Mars exploration rover mission and landing sites

    J. Geophys. Res.

    (2003)
  • A.A. Fraeman et al.

    The stratigraphy and evolution of lower Mount Sharp from spectral, morphological, and thermophysical orbital data sets

    J. Geophys. Res. Planets

    (2016)
  • Cited by (0)

    View full text