Al-Si order and spinodal decomposition texture of a sanidine from igneous clasts of Stromboli (southern Italy): insights into the timing between the emplacement of a shallow basic sheet intrusion and the eruption of related ejecta

A sanidine from subvolcanic cognate monzonite clasts within the debris flow deposits belonging to the 5 ka Secche di Lazzaro pyroclastics of Stromboli volcano (Neostromboli period) was studied by X-ray diffraction (single crystal and powder) and transmission electron microscopy (TEM). The sanidine is poikilitic, Ab₃₁Or₆₅An₄, coexisting with plagioclase, iron-rich diopside and olivine, phlogopite, Ti-magnetite, ilmenite and apatite. The subvolcanic clasts are the slowly cooled equivalent of the potassic magmas erupted as lavas during the Neostromboli period. They are representative of basic magmas arrested at shallow levels as sheet intrusions (sills or dykes). Single-crystal X-ray diffraction of the sanidine (R_4σ = 4.4 % for 1317 unique reflections) enabled us to obtain Al-Si site ordering [X_A1^T1 = 0.301(2)] from average tetrahedral bond lengths. TEM microstructural analysis showed spinodal decomposition modulations, but no cryptoperthitic exsolution lamellae were found. Although quantitative application of the above microstructures and single-crystal results could have been hindered by the uncertainty due to the effect of anorthitic component in the sanidine, Al-Si ordering data suggest that, at 700 °C, the sheet intrusion was cooling on the order of 1–10 °C/day. The spinodal decomposition texture seems to be the result of the disruption of the shallow subvolcanic rocks at time of the 5 ka paroxystic explosive eruption which allowed a syn-eruptive very rapid cooling and prevented a long, in situ, subsolidus cooling history of the ejecta. In fact, the lack of cryptoperthitic exsolution lamellae in the sanidine suggests the monzonite fragments were rapidly ejected to the surface before the temperature of the albite-sanidine solvus was reached. At time of disruption, the basic sheet intrusion had to be characterized by a temperature just below its solidus and therefore it should have been emplaced shortly before the 5 ka eruption of the Secche di Lazzaro pyroclastics.