1J. Leitner, 2C. Vollmer, 3T. Henkel, 1,4U. Ott, 1P. Hoppe
Geochimica et Cosmochimica Acta (in Press) Link to Article [https://doi.org/10.1016/j.gca.2018.05.025]
1Max Planck Institute for Chemistry, Particle Chemistry Department, Hahn-Meitner-Weg 1, 55128 Mainz, Germany
2Institute for Mineralogy, Westfälische Wilhelms-Universität, Correnstrasse 24, 48149 Münster, Germany
3The University of Manchester, School of Earth and Environmental Sciences, Williamson Building, Oxford Road, Manchester, M13 9PL, UK
4MTA Atomki, Bem tér 18/c, 4826 Debrecen, Hungary
We report an in-situ investigation of silicon nitride (Si3N4) grains from several enstatite chondrites of low petrologic types (EL3, EH3, and EH4). The grains occurred in various host phases, including Fe,Ni metal, schreibersite, sulfides, and also in the silicate fraction, and are of Solar System origin. Energy-dispersive X-ray spectroscopy (EDS) showed that carbon and oxygen are present in all investigated grains. Carbon- and N-isotopic compositions of 288 grains were measured by NanoSIMS. Nitrogen is isotopically light compared to terrestrial air, with an average δ15N = –60±1 ‰. The average carbon isotopic composition does not deviate significantly from the terrestrial PDB standard. TOF-SIMS investigation of one particularly large Si3N4 grain (10 µm ×2 µm) showed that the O is located within the grain and not in adjacent particles, and also revealed the presence of chromium. Transmission electron microscopy (TEM) analysis showed that the Cr is present as carlsbergite (CrN) inclusions. TEM investigation of three Si3N4 grains showed them to be polycrystalline, with no consistent crystallographic relationship with the host material. Estimated Si3N4 abundances for four metal-sulfide assemblages demonstrate that the amount of nitrogen bound in the nitrides exceeds the maximum concentration of N that can be stored in Fe,Ni metal in solid solution. Thus, even if all of the N in the metal would have been exsolved, it would not have been enough to form the observed amounts of Si3N4 grains, giving further evidence against formation by exsolution. This clearly shows that they did not form by exsolution from the host materials, as has been suggested in earlier studies. Instead, the Si3N4 must have formed prior to incorporation into the enstatite chondrite parent bodies, either by shock wave-induced condensation processes, or by precipitation from the host phases in the presence of NH3.