Tourmaline is commonly an accessory, locally abundant, mineral in granite-related hydrothermal Sn deposits and records information about the nature and evolution of mineralizing fluids and ore-forming processes. The world-class San Rafael lode-type Sn deposit, located in the northern part of the Andean Tin Belt in southern Peru, is characterized by volumetrically important tourmaline alteration, which resulted from massive fluid-rock interaction. Based on their paragenetic position, several generations of tourmaline have been identified in the San Rafael deposit both of magmatic and hydrothermal origin. They correspond to three major episodes of tourmaline formation that were characterized texturally and compositionally (major and trace elements, and oxygen isotopic signatures) by in situ analysis (SEM, EPMA, LA-ICP-MS, SIMS). The first one, magmatic tourmaline, is found in peraluminous granites. It is texturally homogeneous, has dravite composition with high Li, K, Na, and Zn contents, and shows a narrow range of δ18O values (10.4-11.6 ‰). The second episode corresponds to hydrothermal tourmaline formed during post-magmatic subsolidus alteration and veining-brecciation. It shows complex textural features at microscopical scale, and ranges in composition from dravite to schorl, with similar trace element and δ18O signatures (10.4-14.1 ‰) as the magmatic tourmalines. The last episode formed tourmaline with a schorl-foitite composition, high Sr, Be, Cr, Ni, and HREE contents, and lower δ18O values (7.1-9.7 ‰). It precedes the main quartz-cassiterite-chlorite ore stage and shows particularly high Sn contents (up to 1’000 ppm). Tourmaline formation at San Rafael results thus from multiple fluid circulation episodes in a protracted magmatic-hydrothermal system. Oxygen isotope data indicate precipitation from magmatic-dominated fluid, as recorded previously also for the main ore stage at San Rafael.
Melt inclusions are minute melts trapped during crystal growth [e.g., 1]. When trapped in euhedral olivine they are believed to represent the melt in which olivine grew[1], despite some post-entrapment modifications [e.g., 2, 3]. In this study, we analysed olivine hosted melt inclusions (OHMIs) in Mid-Ocean Ridge Basalts (MORBs) from two different locations on the Mid-Atlantic Ridge: FAMOUS and the 14° MAR triple junction (respectively Normal-MORB and Enriched-MORB).Silicon, as oxygen,is a major element inolivines and it diffusesslowly in dry basaltic melt [4]. Both isotope systems (O and Si) do not fractionate during melting [e.g., 5]. The isotopic ranges reported for bulk MORBs are narrow(<0.5‰)[e.g., 6, 7].Thus, we do not expectδ30Siand δ18Ovariability in OHMIs. However, previous analyses ofδ18Oin these OHMIs showup to 2.5‰ variation (8 times the analytical error)in each sample[8]. Any difference or similarity betweenthe δ30Si and δ18O signaturesinthe OHMIs would give valuable insight on the process responsible for the δ18Ovariation in the OHMIs.Silicon isotopes in OHMIs were measured by ion probe. OHMIs from the two samples displaysimilar δ30Sivalues, rangingfrom 0.08±0.23‰to 2.8±0.32‰for the N-MORB, and from 0.51±0.34‰to 2.8±0.31‰for the E-MORB(2se).The variation is larger than the accuracy (~0.4‰).δ18Oand δ30Sidisplay a slight positive trend. While theaverage δ18Ovalue in the OHMIs from the N-MORB sample is similar to the values previoulsy reported for bulk MORBs [6], the average δ30Siis muchhigher, and similar for both of our samples.Such high valuesare not representative of any high temperature terrestrial reservoirs. The large δ30Sioffset with bulk MORBsand the large variation within the OHMI population thus most likelyreflect either a localizedor OHMIs-specific process(e.g. boundary layer or local dissolution).
Carbonatitic dykes surrounded by K-Na-fenites were discovered in the Pelagonian Zone in Greece. Their carbonate portions have an isotopic mantle signature of δ13C and δ18O ranging from −5.18 to −5.56 (‰ vs. VPDB) and from 10.68 to 11.59 (‰ vs. VSMOW) respectively, whereas their mafic silicate portions have high Nb, Ta and ɛNd values, typical of alkaline basalts. Textural relationships hint at a cogenetic intrusion of silicate and carbonate liquids that according to antithetic REE profiles segregated at shallow depths (<0.6 GPa) from a parental melt sourced deeper in the mantle. Fenites bear similar REE abundances to mafic rocks but with high Rb-Ba and low Nb-Ta values. SHRIMP II U-Pb analyses of magmatic zircon cores (δ18O = 7.21–7.51) from a carbonate-bearing syenitic amphibolite yielded a Permian intrusion age at 278 ± 2 Ma, considerably older than a Cretaceous (118 ± 4 Ma) greenschist overprint obtained from metamorphic zircon rims (δ18O = 6.78–7.02). From 300 to 175 Ma the ɛNd of the Pelagonian magmatism rose irregularly to more primitive values attesting to a higher increment of asthenosphere-derived melts. In this context, the carbonatite formed within a transtensional regime of an intra-Pangaea dextral transform fault that signalled the forthcoming penetrating breakoff of the supercontinent, manifested in the Permo-Triassic.
Abstract Recent advances in microscale 40 Ar/ 39 Ar geochronology have revealed argon concentration gradients in naturally deformed muscovite that are incompatible with volume diffusion uniquely, and have been interpreted to result from intragranular defect-enhanced diffusion. Defects and heterogeneously spaced stacking faults observed by transmission electron microscopy in such muscovites are evaluated as potential fast pathways for argon diffusion. Two-dimensional defects, such as stacking faults, are of particular interest for noble gas diffusion because of the net dilatation effect that a stacking fault is able to generate in minerals. In micas, partial dislocations (and the area between them known as stacking faults) within the interlayer displace the potassium atoms from a stable hexagonally centred position between opposing tetrahedral layers to an unstable position relative to one of the tetrahedral layers such that repulsive forces lead to a localized net dilatation effect within the interlayer. Such a dilatation effect may have direct consequences for argon retention in micas. Numerical modelling of the effects stacking faults have on argon diffusion was performed on the basis of the calculated interlayer spacing, measured isotope data, and observed linear stacking fault density. These calculations result in effective diffusivity ratios defined by volume diffusion to defect-enhanced diffusion of 10 6 to 10 7 , which are comparable with diffusivity ratios in other materials (ceramics or metals). In the absence of defects causing physical grain size reduction (e.g. kink bands or subgrain boundaries), stacking faults are potentially the main defect in sheet silicates exerting a measurable influence on intragranular argon diffusion. Stacking-fault-enhanced argon diffusion differs from pipe diffusion, whose significance on bulk diffusion depends on high dislocation densities, by the small volume fraction of dislocations required to affect bulk diffusivities. In contrast to pipe diffusion, the linked occurrence of dislocations and stacking faults within mica interlayers represents a potentially significant volume fraction, even in samples that do not have high apparent dislocation densities.
