The proportions of the two end-members in the impact melts can be estimated by a binary mixing calculation along the regression line. We assume that Mdivi 1 the volatile-rich end-member has 2 wt.% H2O with 4600‰ in δDδD, 0.5 wt.% CO2, and 750 ppm Cl (Point A in Figs. 2a and 2b). We assume that the volatile-poor end-member has 200 ppm H2O with 0‰ in δDδD, 15 ppm CO2, and 10 ppm Cl (Point B in Figs. 2a and 2b). These assumptions are based on the data from Tissint and literature data from other martian samples. The details are discussed in Sections 4.2 and 4.3. The model results suggest that at maximum level (in the bubbly zone), the volatile-rich end-member contributed ≤10 wt.% of the impact melts. For the other major and minor elements in the volatile-rich end-member, their abundances would be diluted by a factor of 10 or more, and then mixed with the volatile-poor end-member. If the volatile-rich end-member contained more than 2 wt.% H2O, its proportion in the impact melts would be even less, and the dilution factor for other elements would be even larger. Such a dilution effect would hinder the identification of other chemical signatures in somatostatin end-member, such as major-element compositions and oxygen fugacity.
↧