Paired martian breccia meteorites, Northwest Africa (NWA) 7034 and 7533, are the first martian rocks found to contain rare-earth-element (REE) phosphates and silicates. The most common occurrence is as clusters of anhedral monazite-(Ce) inclusions in apatite. Occasionally, zoned, irregular merrillite inclusions are also present in apatite. Monazite-bearing apatite is sometimes associated with alkali-feldspar and Fe-oxide. Apatite near merrillite and monazite generally contains more F and OH (F-rich region) than the main chlorapatite host and forms irregular boundaries with the main host. Locally, the composition of F-rich regions can reach pure fluorapatite. The chlorapatite hosts are similar in composition to isolated apatite without monazite inclusions, and to euhedral apatite in lithic clasts. The U-Th-total Pb ages of monazite in three apatite are 1.0 +/- 0.4 Ga (2 sigma), 1.1 +/- 0.5 Ga (2 sigma), and 2.8 +/- 0.7 Ga (2 sigma), confirming a martian origin. The texture and composition of monazite inclusions are mostly consistent with their formation by the dissolution of apatite and/or merrillite by fluid at elevated temperatures (>100 degrees C). In NWA 7034, we observed a monazite-chevkinite-perrierite-bearing benmoreite or trachyandesite clast. Anhedral monazite and chevkinite-perrierite grains occur in a matrix of sub-micrometer REE-phases and silicates inside the clast. Monazite-(Ce) and -(Nd) and chevkinite-perrierite-(Ce) and-(Nd) display unusual La and Ce depletion relative to Sm and Nd. In addition, one xenotime-(Y)-bearing pyrite-ilmenite-zircon clast with small amounts of feldspar and augite occurs in NWA 7034. One xenotime crystal was observed at the edge of an altered zircon grain, and a cluster of xenotime crystals resides in a mixture of alteration materials. Pyrite, ilmenite, and zircon in this clast are all highly altered, zircon being the most likely source of Y and HREE now present in xenotime. The association of xenotime with zircon, low U and Th contents, and the low Yb content relative to Gd and Dy in xenotime suggest the possible formation of xenotime as a byproduct of fluid-zircon reactions. On the basis of relatively fresh apatite grains and lithic clasts in the same samples, we propose that the fluid-rock/mineral reactions occurred in the source rocks before their inclusion in NWA 7034 and 7533. Additionally, monazite-bearing apatite and REE-mineral-bearing clasts are possibly derived from different crustal origins. Thus, our results imply the wide-occurrence of hydrothermal fluids in the martian crust at 1 Ga or older, which were probably induced by impacts or large igneous intrusions. (C) 2016 Elsevier B.V. All rights reserved.
