STRATIGRAPHIC DEPENDENCE OF CHLORINE ISOTOPE VALUES IN GALE CRATER

Chlorine isotope ratios have been measured on HCl released during evolved gas analysis (EGA) of drilled rock and scooped soil samples using the Sample Analysis at Mars (SAM) instrument on board the Curiosity rover. The source of HCl is thought to be oxychlorines, such as perchlorate, but chloride salts are also a possibility. Previous work revealed δ³⁷Cl values in Gale crater that are exceptionally light and more variable than any other measured material in the solar system (Farley et al., 2016), observations tentatively attributed to the unusual behavior of Cl isotopes in oxychlorine compounds.

New analyses sampling bedrock and a sand dune have been acquired as Curiosity ascends Mount Sharp. These new data reveal a statistically significant correlation between stratigraphic height and δ³⁷Cl in the fluvio-lacustrine sedimentary rocks above the Bradbury landing site. Starting in the Kimberly formation (-58±3‰), rising through the Murray formation (-49±8‰ to -31±9‰) and ending in the altered Stimson unit (-25±5‰), a monotonic relationship between elevation and δ³⁷Cl is observed.

We are presently seeking a cause for this correlation. Long-term temporal variations in instrument response are considered because the rover has largely driven up-section, so stratigraphic height correlates crudely with time. However, the instrumental drift scenario may be ruled out by multiple lines of evidence including: 1) duplicate δ³⁷Cl measurements separated by 533 sols on a sample stored on board the rover are repeatable within error; 2) samples outside of the fluvio-lacustrine stratigraphy (one rock and two sediment samples) measured contemporaneously with those in the sequence do not fit the trend. We suspect the δ³⁷Cl correlation with stratigraphic height is a primary feature of the rock, and not due to post-depositional alteration. A correlation between δ³⁷Cl and Cl or O2 would imply that the isotopic ratio is determined by the relative amounts of oxychlorine and chloride phases, making the trend a simple mixing relationship, but these values do not correlate. Comparing δ³⁷Cl results to other data acquired by Curiosity, we observe no compelling correlations with chemistry or mineralogy, although weak correlations with K₂O, Zn, and SiO₂/TiO₂ may exist; the extent and implications of these correlations of this are being explored.