Experimental constraints on amphibole stability in primitive alkaline and calc-alkaline magmas

Equilibrium crystallization experiments were carried out on two primitive basaltic rocks (APR16: Na2O+K2O=4.40 wt%; CM42: Na2O+K2O=2.59 wt%) with the aim to investigate the amphibole stability in the differentiation processes at deep crustal level, of primitive alkaline (APR16) and calc-alkaline (CM42) magmas. The experiments were performed with different initial H2O contents (0-5 wt%), at pressure of 800 MPa, in the temperature range of 975-1225 degrees C. For the explored conditions, amphibole crystallization occurs in both compositions at H2O in the melt >7wt% while the temperature of their occurrence is lower in the alkaline composition (<1050 degrees C in APR16 and >= 1050 degrees C in CM42). Moreover, amphibole crystallization seems to be influenced by the Na2O/K2O ratio rather than the absolute Na2O content in the melt. This is evident when experimental results on the APR16 and CM42 are compared with experimental data obtained from a primitive ultrapotassic composition (leucite-basanite: Na2O+K2O=4.58 wt%) and with thermodynamic modelling by the Rhyolite-MELTS algorithm. The comparison shows that amphibole never saturates the leucite-basanite at any of the investigated/modelled conditions, even when an extended crystallization increases the Na2O of melts up to contents like those of calc-alkaline experimental glasses. We conclude that, at pressure of 800 MPa and hydrous conditions, only primitive liquids with Na2O/K2O ratio >= 0.9 are more prone to crystallize amphibole.