Stability and chemical equilibrium of amphibole in calc-alkaline magmas: an overview, new thermobarometric formulations and application to subduction-related volcanoes

This work focuses on a rigorous analysis of the physical–chemical, compositional and textural relationships of amphibole stability and the development of new thermobarometric formulations for amphibole-bearing calc-alkaline products of subduction-related systems. Literature experimental results (550–1,120°C, <1,200 MPa, −1 ≤ ΔNNO ≤ +5), H₂O–CO₂ solubility models, a multitude of amphibole-bearing calc-alkaline products (whole-rocks and glasses, representing 38 volcanoes worldwide), crustal and high-P (1–3 GPa) mantle amphibole compositions have been used. Calcic amphiboles of basalt-rhyolite volcanic products display tschermakitic pargasite (37%), magnesiohastingsite (32%) and magnesiohornblende (31%) compositions with aluminium number (i.e. Al# = ^6]Al/Al_T) ≤ 0.21. A few volcanic amphiboles (~1%) show high Al# (>0.21) and are inferred to represent xenocrysts of crustal or mantle materials. Most experimental results on calc-alkaline suites have been found to be unsuitable for using in thermobarometric calibrations due to the high Al# (>0.21) of amphiboles and high Al₂O₃/SiO₂ ratios of the coexisting melts. The pre-eruptive crystallization of consistent amphiboles is confined to relatively narrow physical–chemical ranges, next to their dehydration curves. The widespread occurrence of amphiboles with dehydration (breakdown) rims made of anhydrous phases and/or glass, related to sub-volcanic processes such as magma mixing and/or slow ascent during extrusion, confirms that crystal destabilization occurs with relatively low T–P shifts. At the stability curves, the variance of the system decreases so that amphibole composition and physical–chemical conditions are strictly linked to each other. This allowed us to retrieve some empirical thermobarometric formulations which work independently with different compositional components (i.e. Si^*, Al_T, Mg^*, ^6]Al^*) of a single phase (amphibole), and are therefore easily applicable to all types of calc-alkaline volcanic products (including hybrid andesites). The Si^*-sensitive thermometer and the fO₂–Mg^* equation account for accuracies of ±22°C (σ_est) and 0.4 log units (maximum error), respectively. The uncertainties of the Al_T-sensitive barometer increase with pressure and decrease with temperature. Near the P–T stability curve, the error is <11% whereas for crystal-rich (porphyritic index i.e. PI > 35%) and lower-T magmas, the uncertainty increases up to 24%, consistent with depth uncertainties of 0.4 km, at 90 MPa (~3.4 km), and 7.9 km, at 800 MPa (~30 km), respectively. For magnesiohornblendes, the ^6]Al^*-sensitive hygrometer has an accuracy of 0.4 wt% (σ_est) whereas for magnesiohastingsite and tschermakitic pargasite species, H₂O_melt uncertainties can be as high as 15% relative. The thermobarometric results obtained with the application of these equations to calc-alkaline amphibole-bearing products were finally, and successfully, crosschecked on several subduction-related volcanoes, through complementary methodologies such as pre-eruptive seismicity (volcano-tectonic earthquake locations and frequency), seismic tomography, Fe–Ti oxides, amphibole–plagioclase, plagioclase–liquid equilibria thermobarometry and melt inclusion studies. A user-friendly spreadsheet (i.e. AMP-TB.xls) to calculate the physical–chemical conditions of amphibole crystallization is also provided.