13C-depleted methane pyrolyzed from contaminated Gale Crater sediment cores, Mars

Abstract

This work re-interprets published SAM-EGA (Sample Analysis at Mars-Evolved Gas Analysis) geochemical data for twenty-four sediment cores sampled by the Curiosity rover. The samples were pyrolyzed at 35 °C/min in the range ∼ 100-850 °C. The amount of methane generated from carbonaceous matter in the cores and its stable carbon isotope ratio (δ13C-CH4) in selected cuts within the full temperature range were determined by tunable laser spectrometry (TLS cut). Chemometric analysis of five independent variables for eighteen of the cores identifies four genetic families in which three endmembers explain most data variance. One endmember is contamination by silylating agent (MTBSTFA) introduced unintentionally from wet-chemistry cups during flight and/or after landing of the rover. Five subsamples from the Cumberland (CB) core show a linear relationship between δ13C-CH4 and mean pyrolysis temperature (R2 = 0.77) within TLS cuts in the range 99-786 °C, which conforms with temperature-dependent kinetic theory. Based on dates of pyrolysis for the five CB samples, BSW (bisylylated water, a marker of MTBSTFA contamination) peaked on sol 281 and progressively decreased to sol 382. Linear regression of δ13C-CH4 versus BSW concentration (R2 = 0.98) yields δ13C-CH4 ∼ -70‰ for carbonaceous matter in uncontaminated CB core. Higher BSW yields systematically more negative δ13C-CH4 to the most negative value of −133‰ for CB1, which equates to an apparent fractionation of ∼98‰ from bulk MTBSTFA (δ13C = −35‰) to CB1 methane. Previous workers suggested that δ13C-CH4 from MTBSTFA could not be more than ∼5‰ depleted compared to bulk MTBSTFA. However, their SAM-like laboratory pyrolysis experiments for analogs of MTBSTFA yield δ13C-CH4 values that correspond only to gas trapped within the analyzed pyrolysis temperature cut (455-755 °C). Lower temperature TLS cuts for CB1 and CB2 (220-349 °C and 99-349 °C) trapped more 13C-depleted methane as reflected in their anomalous δ13C-CH4 of −133 and − 115‰, respectively. In addition, pyrolysis of MTBFTSA byproducts, residual solvent, and internal standards with Martian minerals, perchlorates, and sorption or desorption on clays may contribute to more negative δ13C-CH4 than expected from laboratory experiments that lack these components. It is not possible to determine bulk δ13C or δ13C-CH4 for the remaining two carbonaceous endmembers because of the effects of three factors: (1) Chemometric results by alternating least squares regression (ALS) define the relative contributions of endmembers based on five independent variables, not just δ13C-CH4. (2) Samples from distinct locations and stratigraphic ages are subject to variations in δ13C of deposited carbonaceous matter. (3) δ13C-CH4 values for samples from different TLS cuts differ due to temperature-dependent kinetic fractionation.