Deciphering the alteration processes by micro-scale characterization of vermiculite-rich sample from Granby Tuff, an analogue to Noachian clay deposits at Oxia Planum

Introduction:  ESA and NASA will launch ExoMars Rosalind Franklin rover mission to Oxia Planum on Mars. Oxia is a Noachian, phyllosilicate-bearing plainM located between Mawrth Vallis and Ares Vallis [1]. ExoMars will be first mission to perform ground truth measurements of truly Noachian phyllosilicate deposits and is expected to address past water-rich environment and habitability of Mars [2]. The Fe,Mg-rich phyllosilicate surfaces detected at Oxia Planum are some of the largest continuous exposures of this type on Mars. One of the key aspects with relevance to martian evolution is to understand processes that led to formation of so extensive, homogeneous deposits.

Orbital NIR spectral features of the phyllosilicates at Oxia suggest Fe-rich vermiculite and/or saponite [3]. To prepare for the exploration of Oxia Planum and future scientific investigations, characterization of analogue material is needed. Survey of Fe-rich terrestrial vermiculite-bearing rocks [4,5] showed that the best spectral analogy is shown by the basaltic rocks from Granby, Massachusetts, USA. The Granby formation is represented by basaltic flows, dikes and tuffs, all of which hydrothermally altered. Amygdales in vesicular basalts are filled with Fe-rich clay and similar Fe-rich clay material, although much more fine-grained is present as dispersed alteration of glass in tuff [6].

As identified by previous research [4] based on spectral bands between 2.3 and 2.5 μm in the near-infrared region and characteristic X-ray diffraction peaks, Fe-rich clay in Granby rocks is vermiculite±saponite, Phyllosilicates are intergrowing in various proportions in amygdales and tuff. This study provides a more detailed characterization of the phyllosilicate constituents. The aims is to demonstrate the potential of flight-ready technology on ExoMars mission for deciphering the alteration pathway for vermiculite formation and also to understand how relevant for Oxia can Granby analogue be in terms of phyllosilicate formation process.

Method: Mineralogy of two amygdular and tuffaceous samples of Granby was characterized in-situ (as opposed to bulk powder analyses of 4) at a micrometer scale using the laboratory emulator of the Raman Lasers Spectrometer (RLS-Sim) [7] and the spare flight model (FS) of the visible/near-infrared (VNIR) reflectance MicrOmega spectrometer onboard the ExoMars Rosalind Franklin rover.

Additionally, detailed analysis of chemical composition of minerals was performed using SEM-EDX mapping of the exact same surfaces as ones imaged by VNIR and Raman. SEM-EDX can serve as a context for interpretation of data from flight spare instruments and it provides ground-truth information on capabilities and limitations of onboard instruments. Analyzed areas were up to 5 mm x 1 cm in size.

Results: VNIR hyperspectral imaging shows abundant occurrences of phyllosilicate spectrally similar to vermiculite. Strength of characteristic vibrations changes locally, reflecting perhaps crystallinity degree of phyllosilicate. In a few spots, Al-phyllosilicate is also seen, especially in large infilled amygdales.

Raman spectroscopy detects feldspar, that can be defined as albitic in composition based on Raman shift peaks positions. Additionally, peaks suggestive of a mica, perhaps muscovite are present. In multiple places, fine-scale intergrowths of feldspar and muscovite are present, what is manifested by overlap of peaks from two phases in individual analytical spots.

Spatial correlation of the two datasets, VNIR and Raman, suggests that vermiculite, feldspar, muscovite and Al-rich clay are overgrown with each other, although no sufficient textural context is given allowing to infer on alteration sequence and replacements among the minerals.

Context SEM-EDX mapping identifies muscovite and biotite crystals embedded in feldspathic glass, extensively altered to Fe,Mg-clay. In places, zonal Al composition of mica is seen, indicative of leaching during the alteration. Textural and compositional information from SEM shows that biotite and muscovite were destabilized by feldspar or glass, most likely during hydrothermal alteration stage. Saponite or chlorite were most likely formed in this alteration stage as well, particularly in amygdales. In a following stage, most likely during surface weathering, muscovite, biotite and chlorite were altered to vermiculite.

Relevance for Oxia Planum: Tuffs and amygdular basalts from Granby are very good NIR spectral analogue to deposits at Oxia Planum as originally suggested [4,5]. The phase identified by VNIR micrOmega instrument and confirmed by laboratory characterization as Fe-vermiculite (±saponite), perhaps well reflects mineralogy of vast phyllosilicate deposits at Oxia. Is, however, Granby, an analogue for Oxia in terms of process that led to phyllosilicate deposits formation? History of Granby may be more complex that the one of Oxia, based on the higher heterogeneity of mineralogy of Granby as seen in scale of hand-size sample. On the other hand, one can question the apparent homogenous composition of Oxia Planum clays or its possible cryptic heterogeneity, both due to the scale at which it is currently investigated [e.g., 1,3].

Our study indicates some aspects that can be analogous for process of phyllosilicate deposits formation. Deposits at Oxia can be – as Granby Tuff – related to hydrothermal alteration of basalts and ashes. In such case, i.e, if in-situ observations by ExoMars rover collects information suggestive of volcanic ash at the plain, in-situ investigations should focus on amygdales and glasses to address alteration processes in details. Furthermore, our research demonstrates the value of using multiple analytical techniques in a coordinated approach to characterize complex geological samples. By combining mineralogical and chemical analyses, we can obtain a more comprehensive understanding of the samples' composition and formation history.

Acknowledgments: This project was supported by the EU Horizon 2020 Space program call H2020-COMPET-2015-Grant Agreement no 687302. The study got support from the National Planetology Program (PNP) of the INSU-CNRS and from the CNES Research Proposal Call (APR).

References: [1] P. Fawdon et al., (2024) Journal of Maps 20(1).  [2] J. Vago et al., (2017) Astrobiology 17:6-7. [3] L. Mandon et al, (2021) Astrobiology 21: 464-480. [4] A.M. Krzesińska et al, (2021) Astrobiology 21: 997-1016. [5] H. Dypvik et al, (2021) Planetary and Space Science, 208. [6] R. H. April and D. M. Keller (1992). Clays and Clay Minerals 40: 22-31. [7] G. Lopez-Reyes et al., (2021) Journal of Raman Spectroscopy 53: 382-395.