Merrillite is a ubiquitous accessory phase in a variety of Martian meteorite lithologies. The Martian merrillites exhibit a positive correlation between Mg# and Na and a negative correlation between Mg# and both Mn and vacancies in the octahedral Na‐site. Their REE patterns are varied and range from LREE‐depleted to LREE‐enriched. The dominant cation substitutions in the Martian merrillites are Fe2+VI Mg‐site⇔Mg2+VI Mg‐site and Ca2+VI Na‐site +  □VI Na‐site⇔2Na+VI Na‐site. The REE substitution into the 8‐fold coordinated Ca‐site is accommodated by the coupled substitution CaVIII Ca‐site + (Na)VI Na‐site ⇔(Y3+ + REE3+)VIII Ca‐site + □VI Na‐site. The REE substitution is significantly more prevalent in lunar merrillite and can be used as a “fingerprint” to distinguish lunar from Martian meteorites. The substitution of OH− (whitlockite) and/or F− (bobdownsite) for O2− on one of the phosphate tetrahedrons appears to be rather insignificant. The correlations among Na, Mg#, Mn, and Na‐site vacancies are linked to the premerrillite crystallization history of the melt and the crystal chemical behavior of the Mg‐ and Na‐sites. The former reflects the sequence and extent of plagioclase and pyroxene crystallization. The differences in REE pattern shapes among the merrillites reflect source regions for the Martian basalts and the shapes are not greatly perturbed by the crystallization history. The occurrence of merrillite does not imply low‐volatile component in the Martian magmas. However, the low whitlockite and bobdownsite contents suggest that these samples were not altered by hydrothermal fluids and therefore not reset owing to aqueous fluid interactions. Consequently, the young ages of the shergottites are probably true igneous crystallization ages.

Apatite-melt partitioning experiments were conducted in a piston-cylinder press at 1.0-1.2 GPa and 950-1000 °C using an Fe-rich basaltic starting composition and an oxygen fugacity within the range of DIW-1 to DIW+2. Each experiment had a unique F:Cl:OH ratio to assess the partitioning as a function of the volatile content of apatite and melt. The quenched melt and apatite were analyzed by electron probe microanalysis and secondary ion mass spectrometry techniques. The mineral-melt partition coefficients (D values) determined in this study are as follows: D F Ap-Melt = 4.4-19, D Cl Ap-Melt = 1.1-5, D OH Ap-Melt = 0.07-0.24. This large range in values indicates that a linear relationship does not exist between the concentrations of F, Cl, or OH in apatite and F, Cl, or OH in melt, respectively. This nonNernstian behavior is a direct consequence of F, Cl, and OH being essential structural constituents in apatite and minor to trace components in the melt. Therefore mineral-melt D values for F, Cl, and OH in apatite should not be used to directly determine the volatile abundances of coexisting silicate melts. However, the apatite-melt D values for F, Cl, and OH are necessarily interdependent given that F, Cl, and OH all mix on the same crystallographic site in apatite. Consequently, we examined the ratio of D values (exchange coefficients) for each volatile pair (OH-F, Cl-F, and OH-Cl) and observed that they display much less variability: K d Cl-F Ap-Melt = 0.21± 0.03, K d OH-F Ap-Melt = 0.014 ± 0.002, and K d OH-Cl Ap-Melt = 0.06 ± 0.02. However, variations with apatite composition, specifically when mole fractions of F in the apatite X-site were low (X F < 0.18), were observed and warrant additional study. To implement the exchange coefficient to determine the H 2 O content of a silicate melt at the time of apatite crystallization (apatitebased melt hygrometry), the H 2 O abundance of the apatite, an apatite-melt exchange K d that includes OH (either OH-F or OH-Cl), and the abundance of F or Cl in the apatite and F or Cl in the melt at the time of apatite crystallization are needed (F if using the OH-F K d and Cl if using the OH-Cl K d ). To determine the H 2 O content of the parental melt, the F or Cl abundance of the parental melt is needed in place of the F or Cl abundance of the melt at the time of apatite crystallization. Importantly, however, exchange coefficients may vary as a function of temperature, pressure, melt composition, apatite composition, and/or oxygen fugacity, so the combined effects of these parameters must be investigated further before exchange coefficients are applied broadly to determine volatile abundances of coexisting melt from apatite volatile abundances.

No abstract

No abstract

No abstract