Enthalpies of formation at 970 K of compounds in the system MgO-Al 2O 3-SiO 2 from high temperature solution calorimetry

Abstract

The enthalpies of solution of petrologically important phases in the system MgO-Al 2O 3-SiO 2 were measured in a melt of composition 2PbO · B 2O 3 at 970 ± 2 K. The substances investigated included synthetic and natural (meteoritic) enstatite (MgSiO 3), synthetic aluminous enstatite (MgSiO 3 0.9Al 2O 3 0.1), synthetic and natural cordierite (Mg 2Al 4Si 5O 18), synthetic and natural sapphirine (approx. 7MgO·9Al 2O 3 · 3SiO 2), synthetic spinel (MgAl 2O 4), natural sillimanite (Al 2SiO 5), synthetic forsterite (Mg 2SiO 4), synthetic pyrope (Mg 3Al 2Si 3O 12), natural quartz (SiO 2), synthetic periclase (MgO) and corundum (Al 2O 3). Improvement in standardization of the calorimeter solvent made possible greater precision in this study than obtainable in former work in this laboratory on some of the same substances. The enthalpies of formation of enstatite, synthetic cordierite, forsterite and spinel are in reasonable agreement with values previously determined by solution calorimetry. The enthalpy of formation of enstatite is about 0.7 kcal less negative than the value for clinoenstatite resulting from the HF calorimetry of Torgesen and Sahama ( J. Amer. Chem. Soc. 70. 2156–2160, 1948), and is in accord with predictions based on analysis of published pyroxene equilibrium work. Aluminous enstatite with 10 wt.% Al 2O 3 shows an enthalpy of solution markedly lower than pure MgSiO 3: the measurements lead to an estimate of the enthalpy of formation at 970 K for MgAl 2SiO 6 (Mg-Tschermak) orthopyroxene of + 9.4 ± 1.5 kcal/mole from MgSiO 3 and Al 2O 3. Comparison of the enthalpies of formation of synthetic cordierite and anhydrous natural low-iron cordierite shows that they are energetically quite similar and that the synthetic cordierite is not likely to have large amounts of (Al, Si) tetrahedral disorder. Comparison of the enthalpies of formation of synthetic sapphirine and natural low-iron sapphirine shows, on the other hand, that the former is not a good stability model for the latter. The lower enthalpy of formation of the high-temperature synthetic sample is undoubtedly a consequence of cation disordering. The enthalpy of formation of natural sillimanite is considerably less negative than given by the tables of Robie and Waldbaum ( U.S. Geol. Surv. Bull. 1259 1968). The measured enthalpy of formation of synthetic pyrope is consistent with that deduced from published equilibrium diagrams in conjunction with the present measured enthalpy of formation of aluminous enstatite. Calculation of the entropy of synthetic pyrope from the present data yields surprisingly high values and suggests that synthetic pyrope is not a good stability model for natural pyrope-rich garnets. Hence, considerable doubt exists about the direct quantitative application of experimental diagrams involving pyropic garnet to discussions of the garnet stability field in the Earth's outer regions.