Volume 25, Nº 4 (2017)

The paper presents newly obtained original data on the morphology, internal structure (as seen in cathodoluminescence images, CL), and composition of more than 400 zircon grains separated from gabbroids and plagiogranites (OPG) sampled at the axial zone of the Mid-Atlantic Ridge (MAR). The zircons were analyzed for REE by LA-ICP-MS and for Hf, U, Th, Y, and P by EPMA. Magmatic zircon in the gabbroids crystallized from differentiating magmatic melt in a number of episodes, as follows from systematic rimward increase in the Hf concentration, and also often from the simultaneous increase in the (U + Th) and (Y + P) concentrations. These tendencies are also discernible (although much less clearly) in zircons from the OPG. Zircon in the OPG is depleted in REE compared to the least modified zircons in the gabbro, which suggests that the OPG were derived via partial melting of gabbro in the presence of seawater-derived concentrated aqueous salt fluid. Another reason for the REE depletion might be simultaneous crystallization of zircon and apatite. The CL-dark sectors, which are found in practically all of the magmatic zircon grains, have Y/P (a.p.f.u.) ≫ 1 which most likely resulted from OH accommodation in the zircon structure, a fact suggesting that the OPG parental melt contained water. High-temperature hydrothermal processes induced partial to complete recrystallization of zircon (via dissolution-reprecepitation), a process that was associated with ductile and brittle deformations of the zircon-hosting rocks. The morphology of the hydrothermal zircons varies depending on pH and silica activity in the fluid from weakly corroded subhedral crystals with typical vermicular microtopography of the crystal faces to completely modified grains of colloform structure. Geochemically, the earlier hydrothermal transformations of the zircons resulted in their enrichment in La and other LREE, except only Ce, whose concentration, conversely, decreases compared to that of the unmodified magmatic zircons. The hydrothermal zircon displays a reduced Ce anomaly and its most altered domains typically host minute inclusions of xenotime, U and Th oxides and silicates, and occasionally also baddeleyite, which suggests that the hydrothermal fluid was reduced and highly alkaline. These features were acquired by the seawater-derived fluid when it circulated within the axial MAR zone area due to phase separation in the H₂O–NaCl system and particularly as a result of fluid interaction with the abyssal peridotites of oceanic core complexes. Our data demonstrate that zircon is a sensitive indicator of tectonic and physicochemical processes in the oceanic crust.

Isotopic dates newly obtained for the northwestern portion of the Angara–Vitim batholith are consistent with preexisting data on the duration of the Late Paleozoic magmatic cycle: 55–60 Ma (from 325 to 280 Ma). These data also indicate that alkaline mafic magmatism in western Transbaikalia began simultaneously with the transition from crustal granite-forming processes to the derivation of granites of a mixed mantle–crustal nature, with gradual enrichment of the juvenile component in the source of the magmas. Analysis of the currently discussed geodynamic models of Late Paleozoic magmatism shows that a key role in all models of extensive granite-forming processes in the region is assigned to mafic mantle magmas, which can be generated in various geotectonic environments: subduction, delamination, decompression, and a mantle plume. The plume model is most consistent with the intraplate character of the Angara–Vitim batholith. The derivation of the vast volume of granitic material (approximately 1 million km³) should have required a comparable volume of mafic magma that should have been pooled in the middle crust of the Baikal fold area. However, the density structure of the region does not provide evidence of significant volumes of mafic rocks. This suggests that the mechanism of plume–lithospheric interaction that should have induced extensive crustal melting and the origin of vast granite areas was more complicated than simply conductive melting of crustal protoliths in contact with mafic intrusions.

The paper presents data on primary carbonate–silicate melt inclusions hosted in diopside phenocrysts from kalsilite melilitite of Cupaello volcano in Central Italy. The melt inclusions are partly crystalline and contain kalsilite, phlogopite, pectolite, combeite, calcite, Ba–Sr carbonate, baryte, halite, apatite, residual glass, and a gas phase. Daughter pectolite and combeite identified in the inclusions are the first finds of these minerals in kamafugite rocks from central Italy. Our detailed data on the melt inclusions in minerals indicate that the diopside phenocrysts crystallized at 1170–1190°C from a homogeneous melilitite magma enriched in volatile components (CO₂, 0.5–0.6 wt % H₂O, and 0.1–0.2 wt % F). In the process of crystallization at the small variation in P-T parameters two-phase silicate-carbonate liquid immiscibility occurred at lower temperatures (below 1080–1150°C), when spatially separated melilitite silicate and Sr-Ba-rich alkalicarbonate melts already existed. The silicate–carbonate immiscibility was definitely responsible for the formation of the carbonatite tuff at the volcano. The melilitite melt was rich in incompatible elements, first of all, LILE and LREE. This specific enrichment of the melt in these elements and the previously established high isotopic ratios are common to all Italian kamafugites and seem to be related to the specific ITEM mantle source, which underwent metasomatism and enrichment in incompatible elements.