Evidences of fluid separation processes during subvolcanic emplacement of a pegmatite-like magma: The boron (F-Li) rich Capo Bianco aplite (Elba Island, Italy)

One of the main open questions about crystallization processes in aplite-pegmatite systems concerns the growing medium from which pegmatite minerals form: is it a single fluid that evolve continuously from an early silicate melt to a late aqueous fluid, or is it, at some stage, a composite fluid made up by silicate melt, aqueous- flux-rich melts and hydrosaline fluid linked each other by immiscibility processes? Fluid inclusion and experimental studies indicate that several silicate melts and aqueous fluid can coexist in pegmatitic environment showing immiscibility relationships. However, direct petrographic evidences of such a process are still lacking. In fact, pegmatites form at medium and deep crustal levels, and the observation of their pre-emplacement magma structure is inevitably hampered by the crystallization history that lead to the formation of a coarse-grained rock. Capo Bianco aplite (CBa; Elba Island, Italy) is a very peculiar case of a pegmatite-like magma emplaced at shallow crustal level (Dini et al. 2002, 2007). CBa crop out as a dismembered subvolcanic sill, about 3.5 km in length, >= 120 m thick and originally emplaced at a depth of 2.6 km (Rocchi et al., 2002). CBa has a porphyritic trachitoid texture made up of small euhedral phenocrysts (1-5 vol %; 1-4 mm in size) of brownish-grey Fe-rich muscovite, K-feldspar, oligoclase and quartz, set in a very fine-grained groundmass characterized by a fluidal distribution of euhedral albite laths (100-250 µm) and quartz microcrystals (~100 µm) into an equigranular K-feldspar, quartz, zinnwaldite aggregate (5-50 µm). CBa shows a broad rhythmic layering outlined by both the alternance of two different lithotypes having white and pink colours (the only difference being the absence of fine-grained zinwaldite into the white layers), and the inhomogeneous distribution of tourmaline orbicules (Figure1). The rhythmic layering is frequently folded and it is roughly parallel to the fluidal orientation of albite laths as well as to the sill boundaries. The lack of any subsolidus deformation is indicative of its magmatic origin. Tourmaline orbicules (from 1 mm up to 15 cm) have a fibrous-radiating texture and they incorporate subhedral quartz microcrystals, quartz phenocrysts and feldspar phenocrysts (the latter usually replaced by tourmaline). Chemical composition is quite homogeneous from core to rim (F-rich schorl-elbaite solid solution). Orbicules follow the structure of the magmatic flow field as evidenced by the fact that they are found aligned along streamlines defined by white and pink lamellae and folds. In some cases orbicules show a flattened ellipsoidal shape coherently oriented with the magmatic layering and locally, flow fields are found to fold around them. In all these cases the internal fine fibrous-radiating texture is preserved and do not show any subsolidus deformation. Finally, orbicule coalescence is frequently observed with all the steps preserved from touching, to compenetrating up to completely merged orbicules (Figure 2). Large orbicules commonly act as "attractors" of smaller ones resulting in spherical protuberances along the perimeter. Also in this case the complex geometry is reached before the onset of the crystallization because the fibrous-radiating tourmalines show optical continuity. Tourmaline orbicules of the Capo Bianco aplite have been interpreted as the result of a subsolidus autometasomatic process induced by the late circulation of hydrothermal fluids (Marinelli, 1954). However, the lack of a quartz-tourmaline vein network connecting the orbicules, as well as their textural relationships with the magmatic layering indicate that the orbicules represent an early character of this rock and that they behaved as an almost closed system in the course of their crystallization. A closed system like this cannot have a pure hydrothermal nature, because it would result in miarolitic cavities as commonly observed in most intrusions. Instead, this requires a relatively fast crystallization from a boron-rich silicate melt ("hydrosaline melt"?) as supported by (i) the presence of numerous halite-rich and crystal-rich inclusions (ii) the complete filling of orbicules, (iii) the outward crystal growth of radiating tourmaline fibres, (iv) the homogeneous, core to rim, composition of tourmalines (F-rich schorl-elbaite solid solution; in contrast with the extremely zoned tourmalines that usually crystallize from aqueous fluids). This means that during motion of magma into the sill the boron-rich silicate melt bubbles were already separated from the main silicate melt and they moved, coalesced and modified their shape according to flow fields. Only after the cessation of magma flow tourmaline crystallization rapidly occurred, concurrently with the crystallization of the very fine-grained quartz-feldspatic groundmass. These observations, coupled with geochemical composition of the rock and available experimental data on silicatic melts/glasses, indicate that tourmaline orbicules could represent the product of rapid crystallization of boron-rich silicate melt bubbles, earlier separated from the acidic magma, before the final emplacement (during decompression in magma conduit). Capo Bianco aplite can thus be regarded as a serendipitous occurrence of a boron-rich magma that escaped the source region and stopped/crystallized in a sub-volcanic setting, just at the right depth for maintaining a snapshot of silicate-melts immiscibility processes. The study of Capo Bianco aplite could help to unravel the complex crystallization history of pegmatites, taking into account immiscibility processes possibly triggered by rapid magma decompression.