Olivine-clinopyroxene-melt cation exchange reactions as a tool for quantitative understanding magma dynamics

The overall understanding of volcanic processes has evolved exponentially in the last decades, thanks to the rapid advances of geophysical and geochemical monitoring instruments making possible real-time and in situ observations of pre-eruptive dynamics. Thus, the study of active volcanoes provides a plethora of data that must be interpreted in light of what we know about the evolution of magmas (e.g., fractional crystallization, magma mixing, assimilation of host rock by magma, degassing and ascent processes). Albeit this goal is still far from being achieved, the modern style of petrology applied to volcanology offers innovative and visionary scenarios in the comprehension of magma dynamics. Following this line of reasoning, the chemistry of melts and crystals recording the intensive thermodynamic variables (i.e., pressure, P; temperature, T; volatile content and composition of the system, H2O, X) can be used to model the most important processes during the evolution of a magmatic system. In this study, I have adopted a petrological approach based on the partitioning behavior of major and trace elements between crystals (i.e., olivine and clinopyroxene) and melts on both experimental and natural products, in order to gain quantitative constraints on specific pre-eruptive volcanic processes. In the first part of this PhD thesis, I have experimentally explored the effect of carbonate assimilation on primitive basaltic magmas through the compositional information enclosed in olivine crystals. Divalent cation partitioning (Ni2+, Mg2+, Fe2+, Mn2+, Ca2+) between olivine and decarbonated melts have been discussed and compared with previous models to gain refined predictive equations that more accurately quantify the geochemical evolution of primitive skarn environments. Concomitantly, Rare Earth Elements (REE), Y, and Sc partition coefficients between olivine and basaltic melts assimilating variable CaCO3 contents have been debated in the framework of the crystal lattice strain theory (i.e., optimum ionic radius, r0; Young Modulus, E; strain-free partition coefficient, D0). The comparison between cumulates and magmatic skarns from the Colli Albani Volcanic District (Italy) and experiments from this PhD study provides quantitative 2 constraints on the geochemical changes of olivine phenocrysts and their melt inclusions as a function of carbonate assimilation in terms of both major and trace element partitioning. In the second part of this PhD thesis, the compositional information recorded in chemically zoned clinopyroxene phenocrysts from the recent 2003-2017 activity of Stromboli volcano (Aeolian Islands, Italy) have been used to unravel the most effective pre-eruptive volcanic processes driving the geochemical evolution of the magmatic plumbing system. The Present-day (<1.2 kyr) activity of Stromboli is fed by a vertically-extended mush column with an open-conduit configuration in which the hot and mafic magma (low porphyritic or lp-magma) from depth is continuously injected in the homogeneous shallow reservoir (highly porphyritic or hp-magma). Clinopyroxene phenocrysts exhibit marked diopside-augite heterogeneities caused by continuous lp-hp magma mixing and antecryst recycling testifying to the continuous disruption and cannibalism of relic antecrysts from the mush. The transition between diopside and augite has been used to establish the P-T-H2O crystallization conditions over times and to quantify minimum residence times of diopsidic antecrysts and diopsidic recharge bands. On this basis, I argued a distinct phase in the life of Stromboli volcano commenced at least after the 2003 paroxysm. This phase is characterized by more efficient mechanisms of mush disruption and cannibalism, in concert with gravitational instability of the solidification front, involving diopsidic antecrysts remobilization and transport by lp-magmas permeating the mush.