Elbow twins are a distinct feature of cassiterite and rutile. Their occurrence is justified by the reticular theory of twinning because of the existence of a common sublattice — the twin lattice — formed by the overlap of a large fraction of the lattice nodes of the individuals. Yet, the atomic basis for the existence of twins falls outside the possibilities of the reticular theory. We present here the analysis of the pseudo–eigensymmetry of the crystallographic orbits building the structure of the elbow twins and show that their occurrence is justified by the complete restoration of the structure near the composition surface, with a deviation of less than 0.7 Å.
The appropriateness of Zr as an ‘immobile element’ during garnet–hornblende (Grt–Hbl) vein formation potentially caused by the Cl–rich fluid or melt infiltration under upper amphibolite facies condition is examined. The sample used is a Grt–Hbl vein from Brattnipene, Sør Rondane Mountains, East Antarctica that discordantly cuts the gneissose structure of the mafic gneiss. Modal analysis of the wall rock minerals combined with the quantitative determination of their Zr contents reveals that most of the whole–rock Zr resides in zircon whereas ~ 5% is hosted in garnet and hornblende. The Zr concentration of garnet and hornblende is constant irrespective of the distance from the vein. Zircon shows no resorption or overgrowth microstructures. Moreover, the grain size, chemical zoning (CL, Th/U ratio and REE pattern) and rim ages of zircon are also similar irrespective of the distance from the vein. LA–ICPMS U–Pb dating of zircon rims does not give younger ages than the granulite facies metamorphism reported by previous studies. All of these detailed observations on zircon support that zircon is little dissolved or overgrown, and that Zr is not added nor lost during the Grt–Hbl vein formation. Therefore, Zr can be described as an appropriate ‘immobile element’ during the Grt–Hbl vein formation. Detailed microstructural observation of zircon is thus useful in evaluating the appropriateness of Zr as an immobile element.
The decarbonation reaction boundary and the melting temperature in the MgCO3–SiO2 system were investigated up to 26 GPa using a multi–anvil apparatus. It was found that the decarbonation reaction (MgCO3 + SiO2 → MgSiO3 + CO2) occurs below 8 GPa, but it becomes a melting reaction above 8 GPa. The melting system shows a simple eutectic relation. The eutectic point shifts to a higher temperature and more MgCO3–rich composition with increasing pressure. The eutectic temperatures at ~ 9 GPa and ~ 26 GPa were found to be ~ 1700 °C and ~ 2000 °C, respectively. Our results show that magnesite remains stable through the mantle geotherm, at least up to uppermost lower mantle, and that carbon can be transported to the lower mantle as magnesite. If a hot plume brings magnesite back to the upper mantle, it generates carbonatite magma or CO2 fluid.
This study demonstrates the validity of a thin osmium coating for quantitative energy–dispersive spectroscopic (EDS) analysis, particularly for light elements such as O (and potentially C and N) in natural/synthetic minerals. An osmium coating prepared by chemical vapor deposition provides an extremely thin and uniform layer whose thickness can be controlled simply by coating time. Because of the high reproducibility and reliability of the osmium coating process, users have no difficulty in evaluating the actual coating thickness, which enables strict and precise absorption corrections (for the coating layer), even for low–energy characteristic X–rays, which are susceptible to attenuation by the coating layer itself. Our results show that oxygen concentrations in silicate and oxide minerals can be quantified correctly when using the osmium coating, whereas quantification using a carbon coating afforded values that were a few wt% lower than stoichiometry, probably due to the uncertainty of the actual coating thickness (i.e., the absorption correction was incorrect). The ability to accurately quantify oxygen may stimulate new analytical applications, such as the estimation of Fe2+/Fe3+ concentrations and water content in minerals. Furthermore, the Os–coated samples prepared for EDS analysis are also suitable for electron back–scattered diffraction (EBSD) analysis without re–polishing and re–coating, which are usually routine but time–consuming tasks in the case of carbon–coated samples.