Search

The Alpha Particle X-ray Spectrometer (APXS) onboard the Mars Science Laboratory (MSL) rover Curiosity has acquired approximately 1,300 geochemical analyses since landing in 2012. The APXS utilizes a combination of X-ray fluorescence and particle-induced X-ray emission to determine the chemical composition of materials within its 15+ mm diameter field of view (FOV) [1, 2]. Diagenetic features provide a means to further understand and constrain the habitability of Curiosity's landing site, Gale crater. These features often present as veins or nodules with an areal extent on the sub-cm scale. APXS analyses of these features therefore contain a mixture of signals from the feature and host substrate.To probe the composition of these sub-FOV features, Curiosity has developed a technique whereby multiple APXS measurements are conducted in close proximity to the primary target (referred to as a raster). The data are then analyzed to not only localize APXS FOVs, mitigating arm placement uncertainty which is on the order of 1-2 cm [2], but also infer the composition of the various endmembers within the workspace. The original raster analysis method (e.g., [2, 3]) has proven useful at deconvolving the chemistry of diagenetic features from the surrounding substrate. However, this method utilizes APXS oxide data as the primary input. These data are derived assuming a homogeneous sample for the purposes of calculating and correcting for matrix effects (the attenuation of induced X-rays by other elements in the sample). In instances of clear chemical heterogeneities, these matrix corrections can result in skewed compositions of the derived endmembers, such as a vein or nodule.Here we present an improvement to this method whereby we utilize low-level data products and isolate matrix effect calculations for each individual endmember (e.g., [4]). The derived results show significant improvements (10-30%) compared to the oxide method in stoichiometric ratios when applied to calcium sulfate veins, an ideal proof-of-concept sample. Subsequent analyses of magnesium-sulfate dominated nodules hint at other potential mobile elements within the fluids present during diagenesis, such as P, Mn, Ni, and/or Zn. Similar elements were enriched in nodules at the Ayton/Groken field site, where P2O5 and MnO concentrations in the nodular material totaled over 25 wt% at a ~2:1 P:Mn molar ratio [4]. The improved analytical method will be particularly useful as Curiosity continues to explore the Marker Band and sulfate unit.[1] Gellert & Clark (2015), Elements, 11.[2] VanBommel et al. (2016), XRS, 45.[3] VanBommel et al. (2017), XRS, 46.[4] VanBommel et al. (2023), Icarus, 392.

The Alpha Particle X-ray Spectrometers (APXS) onboard the Mars Exploration Rovers (MER) Spirit and Opportunity interrogated the bedrock, soil, and regolith at Gusev crater and Meridiani Planum, respectively. The APXS derives the composition of geologic materials through a combination of particle-induced X-ray emission (PIXE) and X-ray fluorescence (XRF) spectroscopy. Each measurement results in a histogram of energies with characteristic peak areas proportional to elemental concentrations. This spectrum reflects not only the composition of a target but also varies with experimental conditions (e.g., measurement duration, mission age, standoff, temperature), which must be accounted for to accurately quantify the elements present in a spectrum. Individual APXS measurements often provide sufficient counting statistics to resolve and quantify major, minor, and select trace elements (e.g., Ni, Zn, Br) while others (e.g., Ga, Ge) are more difficult to precisely quantify due to, in part, their typical low concentrations (e.g., sub-30 µg/g).To combat the effect of statistical noise on trace element quantification, individual spectra are summed together to create a composite spectrum. We have assembled a database of target characteristics, such as target type (e.g., rock, soil), location, feature, target, formation, and degree of sample preparation (e.g., as-is, brushed, abraded), for each individual APXS spectrum. Spectra of targets with shared geological context and geochemical characteristics (e.g., ratio of Fe3+ to FeT) were summed to create meaningful combinations of individual spectra (i.e., composite spectrum). The composite spectra were fit with a simplified (i.e., Gaussian peaks with a linear background) nonlinear least squares fitting routine to identify promising composites for quantification with the fitting routine developed and utilized for the quantification of other elements within MER APXS spectra. Composite spectra were also assessed visually and quantitatively to confirm they were representative of trace element peaks within each of the individual spectra rather than outliers.At both Meridiani Planum and Gusev crater, results indicate that the concentrations of Ga and Ge in outcrops are more than an order of magnitude higher than expected from meteoritic contribution alone. The ratios of Ga to Al and Ge to Si can also be used to infer the geologic history of a region due to their similar ionic radii and charges and therefore geochemical behavior. The Ga/Al molar ratio tends to be much more consistent at Meridiani Planum compared to that of Ge/Si, which shows more variation between formations. The divergence of the behaviors of Ge and Si could be explained by high temperature diagenetic fluids, as could the consistent behaviors of Ga and Al. We conclude that the elevated concentrations of trace elements such as Ga and Ge may be sourced in part from volcanic outgassing, and regional trends in the molar ratios of Ga/Al and Ge/Si are potentially due to high temperature diagenetic fluids.

Introduction: Curiosity has encountered eolian Stimson formation, Siccar Point group sandstones [1-3] at several locations during ascent of Mount Sharp, Gale crater, at: 1) the Emerson and Naukluft plateaus; 2) on the northern Greenheugh pediment (GP); and most recently 3) on the eastern GP (Figure 1). The Siccar Point group unconformably overlies [4] the predominantly lacustrine, Mount Sharp group, which itself is considered to overlie the Bradbury group (Figure 2). Alpha Particle X-ray spectrometer(APXS) analyses acquired during recent investigations of the eastern GP have expanded the range of eolian Stimson formation compositions and provide further evidence for a relationship with Bradbury group sandstones and cap rocks encountered early in the mission. The Bradbury group includes alkaline composition sandstones, clasts within breccio-conglomerates, and float rocks interpreted to represent primary igneous and volcaniclastic rocks derived from mixing of multiple igneous source rocks [5-12]. We report the results of APXS analyses of the Stimson formation and discuss implications for provenance and timing of events within Gale crater. APXS results: Basaltic composition sandstones, related to average Mars sand and soil, were identified at all three Stimson locations (Figures 1,2). All compositions are discussed here, relative to typical basaltic Stimson chemistry. Sandstones with 3xK2O and >MnO concentrations are identified on the northern GP overlying basaltic composition sandstones (Figures 1,2) [13-14]. Most recently, on the eastern GP, APXS analyses have revealed sandstones with even >K2O concentrations that fall into two groups (Figures 1,2): 1) with ~4xK2O, trending to >Na2O concentrations, and 2) with ~5xK2O, trending to 2O, Al2O3, and Ni, ≫Zn and Ge, and ≈MgO concentrations. The eastern GP, high-K sandstones also overlie more typical basaltic chemistry sandstones. Compositional relationship to Bradbury group and other targets previously encountered on the mission: The higher-K, Stimson sandstones share many compositional characteristics with Bradbury group rocks as well as several float rocks encountered along the Mount Sharp group traverse [13-14] (Figures 3,4). The CCAM instrument has also detected similar relationships [15]. The Bradbury group targets include sandstones near the landing site, high-K Bathurst and Kimberley formation sandstones, and capping sandstones situated just prior to driving on to the Mount Sharp group. Related composition float rocks analyzed along the Mount Sharp traverse are typically associated with Stimson outcrops. The higher-K Stimson sandstones do not plot on the sorting trend defined by the varying grainsize, modern Gale crater sands, and the basaltic Stimson sandstones on an A-CN-K diagram (Figure 4b). Instead, they are offset parallel to the plagioclase–potassium feldspar join. Thus, we do not attribute the compositional differences between the basaltic and high-K pediment Stimson to be the result of eolian sorting. Instead, based on sedimentology [3], mineralogy [16], and composition [13-14], we invoke a change in provenance, with input of a more potassium feldspar-rich source. Given the compositional relationship to Bradbury group sandstones, we suggest that the high-K Stimson sandstones are also derived from the mixing of multiple, igneous source rocks including basaltic and more alkaline compositions [13-14]. Many of the Bradbury group sandstones, Mount Sharp float and higher-K, pediment sandstones are also texturally alike. They are all relatively smooth, resistant, and blockier in appearance than typical basaltic Stimson sandstone (Figure 5). The high-K pediment and Bradbury sandstones also share similar orbital characteristics; both are manifest as crater-retaining units. Timing Implications: The more alkaline composition Bradbury group sandstones near the landing site are >9 km lateral distance and 400 m of elevation from the high-K pediment sandstones. Therefore, we have found evidence along the entire traverse for the presence of alkaline igneous source rocks. These source rocks have provided sediment input to both fluvial (Bradbury) and eolian (Siccar Point) sandstones [13-14]. Based on the compositional and textural relationships between the Bradbury and Siccar Point groups we revisit the timing of events at Gale. The Bradbury group is generally considered to be older than both the Mount Sharp and Siccar Pt groups (Figure 1). This work raises the possibility that they could instead be contemporaneous, with both being younger than the Mount Sharp group. In this scenario, the Mount Sharp group represents relatively old lacustrine/fluvial deposits, which were buried, lithified, and eroded, before deposition of the Bradbury and Siccar Point groups onto that erosional surface, thus explaining differences in elevation between the two groups (Figure 6a). Alternatively, the Bradbury group alkaline sedimentary rocks are the oldest unit and would have been deposited prior to the Mount Sharp group. The Siccar Point group eolian sandstones (basaltic and alkaline) would then have been deposited on to the Mount Sharp group (Figure 6b). In this case, a more alkaline source area in the vicinity of, or within, Gale crater has been eroded by both water and wind at different times during the history of the evolution of the crater and its infill [13-14]. Both scenarios are consistent with widespread evidence at Gale crater for the presence of multiple igneous source rocks. Acknowledgements: MSL APXS is managed and financed by the Canadian Space Agency (CSA). Thanks to the JPL engineers and MSL science team. References: [1] Banham, S.G., et al. (2018) Sedimentology, 65, 993–1042. doi.org/10.1111/sed.12469 [2] Banham, S.G., et al. (2021) JGR: Planets, 126 (4): e2020JE006554 [3] Banham, S.G., et al. (submitted) JGR: Planets [4] Watkins, J.A., et al. (2016) 47th LPSC, Abstract #2939 [5] Stolper, E.M., et al. (2013) Science 341 (6153), 1239463. doi:10.1126/science.1239463 [6] Schmidt, M.E., et al. (2014) JGR Planets, 119(1):64-81, doi:10.1002/2013JE004481 [7] Sautter, V., et al. (2015) Geoscience, 8(8), 605. [8] Thompson, L.M., et al. (2016) JGR Planets, 121(10):1981-2003, doi:10.1002/2016JE005055 [9] Treiman, A.H., et al. (2016) JGR Planets, 121(1), p.75 – 106. doi:10.1002/2015JE004932 [10] Cousin, A., et al. (2017) Icarus 288, 265-283. doi:10.1016/j.icarus.2017.01.014 [11] Edwards, P.H., et al. (2017) MAPS, 52 2931-2410. doi.org/10.1111/maps.12953 [12] Bedford, C.C., et al., (2019) Geochimica et Cosmochimica Acta, 246, 234-266. doi.org/10.1016/j.gca.2018.11.031 [13] Thompson, L.M., et al., (submitted) JGR Planets 2021JE007178 [14] Thompson, L.M., et al., (2022) 53rd LPSC, 1475.pdf [15] LeDeit, L., et al. (this conference) [16] Thorpe, M.T., et al. (Submitted) JGR Planets 2021JE00709