J. Ferdousa, A.D. Brandona, A.H. Peslierb, Z. Pirottec
Geochimica et Cosmochimica Acta (in Press) Link to Article [https://doi.org/10.1016/j.gca.2017.05.032]
aEarth and Atmospheric Sciences Department, University of Houston, Houston, TX 77204-5007, USA
bJacobs, NASA-Johnson Space Center, Mail Code X13, Houston TX 77058, USA
cDept. de Géologie, Université Libre de Bruxelles, 1050 Brussels, Belgium
The origin of the incompatible trace element (ITE) characteristics of enriched shergottites has been critical for examining two contradicting scenarios to explain how these Martian meteorites form. The first scenario is that it reflects ITE enrichment in an early-formed mantle reservoir whereas the second scenario attributes it to assimilation of ancient Martian crust (∼4-4.5 Ga) by ITE-depleted magmas. Strongly differentiated shergottite magmas may yield added constraints for determining which scenario can best explain this signature in enriched shergottites. The meteorite Northwest Africa (NWA) 856 is a basaltic shergottite that, unlike many enriched shergottites, lacks olivine and has undergone extensive differentiation from more primitive parent magma. In similarity to other basaltic shergottites, NWA 856 is comprised primarily of compositionally zoned clinopyroxenes (45% pigeonite and 23% augite), maskelynite (23%) and accessory minerals such as ulvöspinel, merrillite, Cl-apatite, ilmenite, pyrrhotite, baddeleyite and silica polymorph. The CI-chondrite normalized rare earth element (REE) abundance patterns for its maskelynite, phosphates, and its whole rock are flat with corresponding light-REE depletions in clinopyroxenes. The 87Rb-87Sr and 147Sm-143Nd internal isochron ages are 162±14 (all errors are ±2σ) Ma and 162.7±5.5 Ma, respectively, with an initial εNdI= -6.6±0.2. The Rb-Sr isotope systematics are affected by terrestrial alteration resulting in larger scatter and a less precise internal isochron age. The whole rock composition is used in MELTS simulations to model equilibrium and fractional crystallization sequences to compare with the crystallization sequence from textural observations and to the mineral compositions. These models constrain the depth of initial crystallization to a pressure range of 0.4-0.5 GPa (equivalent to 34-42 km) in anhydrous conditions at the Fayalite-Magnetite-Quartz buffer, and consistently reproduce the observed mineralogy throughout the sequence with progressive crystallization. The Ti/Al ratios in the clinopyroxenes are consistent with initial crystallization occurring at these depths followed by polybaric crystallization as the parent magma ascended to the surface. The REE abundances in the clinopyroxenes and maskelynite are consistent with progressive crystallization in a closed system.
The new results for NWA 856 are combined with other shergottite data and are compared to mixing and assimilation and fractional crystallization (AFC) models using depleted shergottite magmas and ancient Martian crust as end-members. The models indicate that the range of REE abundances and ratios, when taken in isolation, can be successfully explained for all shergottites by crustal contamination. However, no successful crustal contamination model can explain the restricted εNdI of -6.8±0.2 over the wide range of Mg# (0.65 to 0.25), and corresponding trace element variations from enriched shergottites to depleted shergottites. The findings indicate that the origin of the long-term ITE-enriched signature in enriched shergottites and the geochemical variability seen in shergottites is not a result of crustal contamination but instead reflects ancient mantle heterogeneity.