The architecture of the magmatic feeding system of the Campi Flegrei Volcanic District: constraints from experimental petrology

This work is aimed to provide an experimental contribution to a better definition of the evolutionary processes of Phlaegrean Volcanic District (PVD). To reach this goal high pressure experiments were performed on a natural sample of mafic composition (K-basalt), at variable T, P, ƒO2 and XH2O conditions. Pressures of 800, 400 and 200 MPa have been chosen to simulate the magmatic differentiation in the PVD system, supposed to be a multilevel one. A temperature range of about 120°C starting from the liquidus temperature has been set up for the experimental runs. Water effect was evaluated by the addition of 1 to 6 wt% of H2O to the system. Two piston-cylinder apparata were used: an end loaded type and a non end loaded one. The experimental ƒO2 conditions, determined by the assemblies, were close to the QFM buffer for the first apparatus and to NNO+2 for the second one. Both equilibrium and thermal gradient experiments were performed. The sub-liquidus experiments show variable mineralogical assemblages: stable phases observed by use of a scanning electron microscope (SEM) are olivine, clinopyroxene, plagioclase, spinel and amphibole. The results from this work well matches the primitive compositions of the PVD and can furnish useful information on the plumbing system of this densely inhabited and extremely dangerous volcanic district. The high pressure (800 MPa) experiments give a contribution in clarifying the possible mantle source mineralogy under the Phlegraean area. Indeed the T-H2O phase relations from this work compared to other experimental work on primitive alkali-basalt composition show that no phlogopite crystallizes even in similar T, H2O and pressure conditions. This may suggest that this phase does not occur in the mantle source beneath the PVD. This work demonstrates that a basaltic melt cannot reach the most differentiated compositions of the PVD products (i.e. Campanian Ignimbrite trachytes) with a single-step isobaric equilibrium (or fractional) crystallization process. Indeed geophysical and petrological/petrographical features of natural products, according with those highlighted by the experimental data, confirm the hypothesis of a multilevel polybaric magmatic system feeding the PVD, with a high pressure differentiation step fractionating clinopyroxene +/- olivine. The possibility of a deep liquidus crystallization of clinopyroxene (alone or in co-saturation with olivine) is also supported by the major element compositional trends defined by the MIs from Minopoli I and Fondo Riccio, along with a portion of those from Solchiaro. Thus these results display the great relevance of clinopyroxene, not only as leading phase in driving the overall evolution of the residual melts, but also as witness of the first steps of the deep differentiation process in the PVD. The low pressure (400 and 200 MPa) experiments always show olivine as liquidus phase, thus implying a strong abundance of this phase also in natural products. However olivine does not occur in the most evolved PVD products, such as the trachytes form Campanian Ignimbrite. Once again a possible explanation could come from the multilevel magmatic system. In this light the supposed deep differentiation step at the base of the crust may provide a level in which olivines and perhaps clinopyroxenes could be segregated. The experimental glasses trends allow us to hypothesize that the difference in the initial differentiation path between Procida and Campi Flegrei may be due to a different amount of water content in the initial steps of differentiation. Interestingly, the good match with natural products is gained without the addition of a limestone component. Carbonate would indeed shift the compositions toward a strong silica undersaturation. As a result, the experiments from this work show that there’s no need to consider limestone assimilation in order to explain the compositional features of the PVD magmas. A last finding regards natural samples showing calcic amphibole in their paragenesis. They are extremely rare and only occur in more differentiated compositions. This allow to exclude from the natural system the conditions in which amphibole experimentally crystallizes, which are P=200 MPa; T<1050°C, H2O ~7 wt%, giving an experimental constraint on the maximum amount of water content in the melt at low pressure.