Research Article

View ORCID ProfileChristopher H. House, View ORCID ProfileGregory M. Wong, Christopher R. Webster, Gregory J. Flesch, View ORCID ProfileHeather B. Franz, Jennifer C. Stern, View ORCID ProfileAlex Pavlov, Sushil K. Atreya, View ORCID ProfileJennifer L. Eigenbrode, Alexis Gilbert, View ORCID ProfileAmy E. Hofmann, Maëva Millan, Andrew Steele, View ORCID ProfileDaniel P. Glavin, Charles A. Malespin, and View ORCID ProfilePaul R. Mahaffy

  1. aDepartment of Geosciences, The Pennsylvania State University, University Park, PA 16802;

  2. bEarth and Environmental Systems Institute, The Pennsylvania State University, University Park, PA 16802;

  3. cNASA Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109;

  4. dSolar System Exploration Division, NASA Goddard Space Flight Center, Greenbelt, MD 20771;

  5. eClimate and Space Sciences and Engineering, University of Michigan, Ann Arbor, MI 48109;

  6. fDepartment of Earth and Planetary Sciences and Earth-Life Science Institute, Tokyo Institute of Technology, Tokyo 152-8550, Japan;

  7. gDepartment of Biology, Georgetown University, Washington, DC 20057;

  8. hEarth and Planets Laboratory, Carnegie Institution for Science, Washington, DC 20015

See allHide authors and affiliations

  1. Edited by Mark Thiemens, Department of Chemistry and Biochemistry, University of California San Diego, La Jolla, CA; received August 26, 2021; accepted December 1, 2021


Carbon isotopic analysis is among the most pervasive geochemical approaches because the fractionation of carbon isotopes produces a natural tracer of biological and chemical processes. Rover-based carbon isotopic analyses of sedimentary rocks on Mars have the potential to reveal modes of Martian carbon cycling. We report carbon isotopic values of the methane released during pyrolysis of samples obtained at Gale crater. The values show remarkable variation indicating different origins for the carbon evolved from different samples. Samples from multiple locations within Gale crater evolved methane with highly fractionated carbon isotopes. We suggest three routes by which highly fractionated carbon could be deposited on Mars, with each suggesting that Martian carbon cycling is quite distinct from that of the present Earth.


Obtaining carbon isotopic information for organic carbon from Martian sediments has long been a goal of planetary science, as it has the potential to elucidate the origin of such carbon and aspects of Martian carbon cycling. Carbon isotopic values (δ13CVPDB) of the methane released during pyrolysis of 24 powder samples at Gale crater, Mars, show a high degree of variation (−137 ± 8‰ to +22 ± 10‰) when measured by the tunable laser spectrometer portion of the Sample Analysis at Mars instrument suite during evolved gas analysis. Included in these data are 10 measured δ13C values less than −70‰ found for six different sampling locations, all potentially associated with a possible paleosurface. There are multiple plausible explanations for the anomalously depleted 13C observed in evolved methane, but no single explanation can be accepted without further research. Three possible explanations are the photolysis of biological methane released from the subsurface, photoreduction of atmospheric CO2, and deposition of cosmic dust during passage through a galactic molecular cloud. All three of these scenarios are unconventional, unlike processes common on Earth.

  • Gale crater
  • Mars
  • carbon isotopes
  • pyrolysis
  • methane

Carbon isotopic data from Martian sedimentary organic carbon can potentially elucidate the origin of indigenous organics and reveal aspects of the Martian carbon cycle. Extended exploration by the Mars Science Laboratory (MSL) Curiosity rover of the fluvio-lacustrine sedimentary system at Gale crater provides unique opportunities, as samples are collected from a variety of locations within a known stratigraphic context (1, 2). MSL has collected and analyzed more than 30 drilled samples on Mars between August 2012 and July 2021. These samples have been collected from varied lithologies from hundreds of meters of stratigraphy in Gale crater that represent the complex history and evolution of the region. A description of the mission samples studied can be found in the supplement, and a stratigraphic column is shown in Fig. 1A. Methods used in this work are also described in the supplement to this manuscript. This study considered the carbon isotopic values of methane evolved during pyrolysis as observed by the MSL tunable laser spectrometer (TLS) of the Sample Analysis at Mars (SAM) instrument suite (3) from 24 samples from Gale crater, Mars. Methane abundances and δ13C values were determined from analyzing TLS-SAM high-resolution spectra.

Fig. 1.

Geologic context of samples included in this study. (A) Stratigraphic column with labels for each of the MSL drill sites. (B) Moray_Firth Mastcam mosaic (mcam14053) from Sol 2685 showing Greenheugh pediment near the location of the EB drill hole, which was drilled on top. (C) HU drill hole in the Glasgow member of the Murray formation just below the Greenheugh pediment. (D) HF drill hole in gray-colored Jura member Murray mudstone at the top of the VRR. (E) Namib dune of the Bagnold dunes where the GB sample was taken. (F) Yellowknife Bay locality where the CB drill hole was drilled into mudstone of the Sheepbed member of the Bradbury group rocks.


For evolved gas analysis (EGA) at Gale crater, drilled powder or scooped fines are heated at 35 °C min−1 under a helium flow in quartz cups up to ∼850 °C. A fraction of the evolved gas mixture is directly analyzed with a quadrupole mass spectrometer, and a specific preselected temperature cut of the remaining gas is diverted to the Herriott cell of the TLS. The redox state of the oven, in practice, depends on the relative abundances of oxychlorine species and reduced minerals in each sample. During EGA, organic material can produce a range of products including CO2, CH4, CO, OCS, CS2, and molecular organic fragments. EGA δ13C CH4 values were measured by the TLS instrument of the SAM suite for 24 samples from Gale crater, Mars (Table 1) using methods described in this paper’s supplement. The amount of CH4 observed by the TLS-SAM instrument for the TLS temperature cut is also indicated in Table 1. This includes five analyses of the Cumberland (CB) sample drilled in the Sheepbed member of the Bradbury group rocks at Yellowknife Bay, as well as 15 samples from the Mount Sharp group, three from the overlying Stimson formation, and one scooped sand sample. Shown in Table 2 are the posterior probabilities of reduced sulfur presence (based on several evolved sulfur gases observed compared to laboratory data) and the δ34S values calculated from SO2 evolved between ∼500 and 600 °C during EGA and measured by the SAM quadrupole mass spectrometer (QMS).

Table 1.

MSL methane isotopic values from EGA

Table 2.

Relevant results from the MSL QMS during EGA and informal stratigraphic units for samples

As seen in Fig. 2, the TLS CH4 δ13C values that are highly 13C depleted correspond predominantly with the 34S-depleted δ34S QMS values observed for evolved SO2, reported by Franz et al. (4) and Wong (5). Four of the most depleted TLS CH4 δ13C values (Table 1) are from samples [CB2, CB3, CB5, and Edinburgh (EB)] that also clearly evolve 34S-depleted SO2 (Table 2 and Fig. 2). Such δ34S-depleted SO2 evolved at mid (500 to 600 °C) temperature has been interpreted to indicate Martian sulfides (4). Reduced sulfur has also been inferred based on the evolution of OCS and CS2 compared to analyses of laboratory sulfur samples (6). Interestingly, 8 out of the 10 samples showing strong TLS δ13C depletions (i.e., <−70‰) in evolved CH4 also evolved sulfur gases (OCS and CS2) indicative of a reduced sulfur detection based on statistical comparisons to laboratory runs (Table 2). One of the two samples in which the EGA data do not support a reduced sulfur detection is CB6, which had an unusual EGA procedure that prohibited an effective analysis of the sulfur gases. The other is Hutton (HU), which occurs just below the Greenheugh pediment (a gently sloping erosional surface downslope from Gediz Vallis) and the Basal Siccar Point group unconformity. HU shows evidence for geochemical alteration by later fluids flowing near the Basal Siccar Point group unconformity (e.g., ref. 7).

Fig. 2.

EGA TLS CH4 δ13CVPDB values versus EGA SO2 δ34SVCDT QMS values (4, 5). Error bars indicate 1 SE. For reference, the dashed lines separate the graph into quadrants around the origin, and the gray line shows a weighted linear fit (y = [8 ± 3]x − [59 ± 12], mean squared weighted deviation = 7). Most analyses from Gale crater that have a large negative δ13C value in evolved CH4 also have 34S-depleted evolved SO2.


Some of the 13C depletions reported here are anomalously large, especially with respect to the carbon isotopic composition of the Martian atmosphere, whose δ13C value reported earlier by TLS-SAM is ∼+46‰ (8). This atmospheric value reflects the integrated largescale loss of volatiles from the Martian atmosphere and thus, the δ13C composition of the atmosphere may have been less 13C enriched when Gale crater sediments were deposited. Because of the magnitude of the 13C depletions observed in CH4 evolved during EGA runs, the authors have considered potential rover-induced origins for the observations without uncovering any explanation. In fact, the TLS spectra obtained on Mars from evolved CH4 (SI Appendix, Fig. S1) are exceptionally clean and provide multiple 12C and 13C lines with which to calculate δ13C values, making it unlikely that the 13C depletions observed are due to an interfering organic molecule. From the repeat CB analyses, it appears that the isotope depletion observed in CH4 is most pronounced at lower temperatures. This observation suggests that precursors to the evolved CH4 are relatively volatile organic molecules. However, strong depletions were still observed in several samples using high-temperature cuts (>450 °C) in which the temperature cut includes a tail of a CH4 release centered at a lower temperature (e.g., SI Appendix, Fig. S2). The CH4 isotopic variation in CB samples, though, are not completely explained by differences in the temperature cut. Additionally, at Yellowknife Bay, the TLS analyses of CB using the 2.78-μm laser produced highly 13C-depleted CO2 δ13C values (SI Appendix, Table S1). While these TLS analyses of CB produced CO2 δ13C values that were, in some cases, comparable to CH4 δ13C (SI Appendix, Table S1), such depleted CO2 δ13C values have not been observed in later samples of the mission. In contrast, TLS CH4 δ13C values using the 3.27 μm laser show strong depletions in multiple different drill samples from vastly different parts of the mission. Because the observation of anomalous 13C values in evolved CH4 are repeatable with different samples spread out in space and time, we have focused on those values for this study.

We have considered the SAM instrument background of N-tert-butyldimethylsilyl-N-methyl-trifluoroacetamide (MTBSTFA (δ13C = −35‰; refs. 9 and 10) as a reasonable source of evolved CH4 to consider. Early in the mission, this background was estimated to contribute up to 900 nmol of CO2 during pyrolysis (11). In addition to oxidizing to CO2, MTBSTFA is known to react with water resulting in 1,3-bis(1,1-dimethylethyl)-1,1,3,3-tetramethyldisiloxane [or bisilylated water (BSW)], which can be monitored by the SAM QMS. Based on the levels of BSW detected, the level of MTBSTFA background has been variable between runs depending on how long it has been since a wet chemistry experiment and, for a few cases in which it was relevant, how long a sample was stored before analysis. For 16 of the samples, the mean of the total BSW observed during the course of the EGA runs was 1.0 nmol with an SD of 1.5 nmol (Table 2). HU had over 22 times this average, and Duluth (DU) had over 30 times more than this average. In the case of DU, the extreme amount of BSW observed was also concurrent with an anomalous amount of total EGA methane (774 nmol) and a TLS CH4 isotopic value (δ13C = −45‰) similar to MTBSTFA (δ13C = −35‰), which demonstrated that with elevated levels of MTBSTFA in the SAM background, a portion can end up as evolved CH4. Further, an intramolecular isotopic analysis of methyl-trifluoroacetamide from the hydrolysis of MTBSTFA showed that the types of carbon (methyl-carbon and carbonyl C) most likely to contribute CH4 from the MTBSTFA background were the most 13C enriched (SI Appendix, Fig. S3), making it unlikely that the observed anomalies stemmed from a site-specific carbon depletion obscured in the bulk δ13C value for MTBSTFA.

Major endmember carbon reservoirs (SI Appendix, Table S4) presently on Mars are the atmospheric CO213C = +46 ± 4‰; ref. 8) and the igneous carbon (δ13C = −20 ± 4‰; ref. 12). TLS δ13C values between −17 ± 2‰ and −57 ± 2‰ were found for 10 samples including DU. The isotopic composition of the CH4 evolved during pyrolysis of these samples may reflect the MTBSTFA SAM background, Martian igneous carbon, and/or meteoritic infall along with any isotopic fractionations the occur during oven reactions. Laboratory experiments using solid materials were conducted to explore the magnitude of carbon isotopic fractionation possible during pyrolysis under conditions similar to SAM (SI Appendix, Table S2). We found that cleavage and reduction of methyl groups to CH4, as would happen for most CH4 derived from MTBSTFA products, produced little 13C depletion (0.4 to 4.6‰; SI Appendix, Fig. S4). CH4 evolved from recalcitrant sources (graphite and diamond) showed 13C-depletion of 8 to 21‰, CH4 from oxalate/oxamide showed moderate 13C depletion (25.0 to 25.3‰), and bicarbonate reduction showed the largest 13C depletions (28.5 to 49‰). Both our laboratory pyrolysis experiments (SI Appendix, Table S2) and modeling of possible isotopic pseudoequilibration during pyrolysis (SI Appendix, Figs. S5–S8 and Table S3) considering several different scenarios suggested that oven processes would typically produce fractionations less than 50‰ and, therefore, cannot account for the large 13C depletions observed in multiple samples at Gale crater. Applying the most extreme oven fractionation imagined to these carbon reservoirs results in CH4 with δ13C values of about −5 to −70‰. CB (CB1, CB2, CB3, CB5, CB6), Gobabeb (GB2), Highfield (HF), Rock Hall (RH), Hutton (HU), and EB showed TLS CH4 values more 13C depleted than this range, indicating that oven reactions are not likely to be causing their anomalous 13C-depleted values observed in evolved CH4.

It may be notable that the highly depleted 13C values for evolved CH4 have so far been found in five distinct locations at Gale crater, Mars (Table 2 and Fig. 1 A–F). The highly 13C-depleted signal was first seen in the mudstones of Yellowknife Bay on the crater floor (Fig. 1F) in a location where high thermal inertia values were measured from orbit and have been potentially attributed to secondary alteration at the end of the Peace Vallis fan (13). N