High-temperature properties of silicate liquids: applications to the equilibration and ascent of basic magma

Abstract

High-temperature heat content measurements have been made on a series of silicate liquids, which in conjunction with published data, are used to derive partial molar heat capacities of SiO₂, TiO₂, Al₂O₃, Fe₂O₃, FeO, MgO, CaO, Na₂O and K₂O in the temperature range 1200-1650 K. Only Fe₂O₃appears to be compositionally dependent, and the best evidence suggests that there is no excess heat capacity (CPi i = C°). In combination with calorimetric data and the effect of pressure on the fusion temperature of solid compounds, a consistent set of enthalpy, entropy and volume data have been derived for the liquid compounds CaMgSi₂O₆, NaAlSi₃O₈, KAlSi₃O₈, Fe₂SiO₄and TiO₂. By using activities (relative to a liquid standard state) calculated at 1 bar for a range of lavas, the equilibration pressures and temperatures of lavas with a lherzolitic source material are calculated, and for basanites indicate 22-26 kbar and 1310-1360 °G. The regular solution formulation used in these calculations gives an estimated error of 40 °G and 5.7 kbar when compared to experimental equilibria. It is suggested that one of the thermal responses of ascending alkali basalt magma to engulfing cooler lherzolitic nodules could be the precipitation of megacrysts, and the calculated equilibration pressures and temperatures of the megacryst assemblage (16-20 kbar, 1220-1240 °G) is in accord with this. The importance of viewing volcanic eruptions as the last stage in a sequence of chemical and thermomechanical instabilities is pointed out. Equations expressing the conservation of energy, mass and momentum on a macroscopic scale are given. The high Rayleigh numbers appropriate for even the relatively small magma volumes of erupted alkali basalts indicate turbulent flow-regimes with characteristic thermal convection velocities of the same order as nodule settling velocities. There is a significant partial melting effect in the mantle surrounding an ascending diapir if buoyancy is a significant force acting to drive the magma upwards. The effect of latent heat and convective heat losses on the thermal budget of a rising diapir has been calculated and shows the assumption of adiabaticity is often unwarranted - even for rapidly ascending magma. Finally, mass transfer rates due to convective diffusion have been calculated for all the major components in a basic silicate liquid. Integral mass exchange depends inversely on the ascent rate and is quite small for the rapidly ascending alkalic basalts.