New Experimental Data on Biotite + Magnetite + Sanidine Saturated Phonolitic Melts and Application to the Estimation of Magmatic Water Fugacity

Paul J. Rouse (p.rouse@bristol.ac.uk)¹ & Michael R. Carroll (carroll@campus.unicam.it)²

¹ Dept. Earth Sci., Univ. Bristol, Bristol BS8 1RJ, UK

² Dipt. Sci. Terra, Univ. Camerino, 62032 Camerino, Italy

Water is the dominant volatile species in many evolved magma types but during volcanic eruptions magmas typically exsolve most of their original water. However, estimation of pre-eruptive magmatic water fugacities is possible in samples which contain phase assemblages which are sensitive to water fugacity. One such assemblage involves biotite + magnetite + sanidine (Bt+Mt+San), for which water fugacity can be estimated if oxygen fugacity is know. We have conducted new experiments in which we crystallized Bt+Mt+San from a hydrous natural phonolitic melt at water fugacities of 39 to 152 MPa and oxygen fugacities from FMQ to NNO+1 (750-825°C). Experiments were done using standard hydrothermal techniques, with double capsule buffering for the FMQ experiments and the intrinsic oxidation state of the pressure vessel controlling the NNO+1 experiments (verified using Ni-Pd alloy measurements). Following microprobe analysis of all phases in the run products we calculate orthoclase activity in alkali feldspar following Waldbaum and Thompson (1969) and magnetite activity following Woodland and Wood (1994). Of the various possible ways of evaluating annite activity in biotite, we obtained the best results (calculated water fugacity close to experimental water fugacity) for the experimental assemblages when we use a simple ionic model in which annite activity is equal to the annite mole fraction cubed. Evaluation of the biotite compositional variation in the experiments suggests that the major substitution mechanism involves substitution of Ti + a vacancy for 2[MgFe]. This aids calculation of biotite structural formulae (Fe³⁺ typically less that 30%) and annite activities. Over the range of conditions investigated we find that experimental water fugacities can be calculated to within 10-15%. The success of this approach is helped considerably by the relatively simple substitutions encountered in the phonolitic system investigated.

Waldbaum DR, Thompson JB, Am. Mineral., 54, 1274-1298, (1969).

Woodland AB & Wood BJ, Eur. J. Mineral, 6, 23-37, (1994).

Kinetics of Equilibration between Liquid Metal and Magnesiowüstite at High Pressure

David C. Rubie (dave.rubie@uni-bayreuth.de)¹, Sophie Fortenfant (sophie.fortenfant@ uni-bayreuth.de)¹ & Christine K. Gessmann (christine.gessmann@bristol.ac.uk)²

¹ Bayerisches Geoinstitut, Universität Bayreuth, D-95440 Bayreuth, Germany

² Department of Earth Sciences, University of Bristol, Bristol BS8 1RJ, UK

Recent models of core formation suggest that liquid metal equilibrated with silicates at the base of a magma ocean at a pressure of 25-30 GPa. The liquid metal fraction then sank as diapirs through the lower mantle so that chemical reequilibration at higher pressures was limited. Such models can explain the present abundances of siderophile elements (e.g. Ni and Co) in the Earth's mantle based on experimental studies of element partitioning at high pressures. These models imply that limited disequilibration reactions must have occurred during transport of metal through the lower mantle. Furthermore the metallic core would not be in equilibrium with the lower mantle and disequilibrium reactions could still be occurring at the core-mantle boundary, depending on the kinetics. Such reactions might involve siderophile elements and also elements such as oxygen and/or silicon.

We are studying the kinetics of reactions between magnesiowüstite and Fe-rich liquid metal that is supersaturated in oxygen at 10 GPa and 1900-2100°C using a multianvil apparatus. Samples of Fe-Ni-O metal are contained in an MgO capsule. Reaction initially involves the growth of a FeO rim between the MgO and liquid metal, which consumes excess oxygen. With increasing time, this rim partially equilibrates with the MgO by Fe-Mg interdiffusion, and the composition of the oxide in contact with the liquid metal becomes progressively enriched in MgO. The oxygen fugacity, which controls the partitioning of Fe and Ni between the metal and oxide, decreases with time as the oxygen content of the metal is reduced by reaction. Further experiments are being performed to enable a reaction/diffusion model involving a moving interface and changing interface composition to be developed. The results will help to understand the nature and extent of reactions between liquid metal and silicates/oxides in the lower mantle during the early history of the Earth.

Study of Solvation and Cation-Anion Interactions in Aqueous Solutions Using Raman Spectroscopy

Fernando Rull Perez (rull@fmc.uva.es)¹ & Jean Dubessy (jean.dubessy@g2r.u-nancy.fr)²

¹ Cristalografia y Mineralogia, Facultad de Ciencias, 47005 Valladolid, Spain

² UM G2R (7566), Faculté des Sciences, UHP, BP 239 54506 Vandoeuvre, France

Thermodynamic properties of aqueous solutions are bulk properties which depend on molecular interactions. A good description of these interactions is a branch of fluid geochemistry research. Raman spectroscopy is sensitive to molecular interactions, and specially in the formation of solvated structures and ion-pair at the contact. A thorough and unambiguous interpretation of the Raman spectra must be based on the examination of all the bands of the spectra: i) the bands assigned to the intramolecular vibrations of the solvent, which are also affected by their environment; ii) the bands assigned to the intermolecular modes of the solvent, which are fingerprint of the hydrogen bonds; iv) the solvation bands of the cations, whose identification is dependent both on the cation and the anion; v) the internal vibrations bands related to the polyatomic anions. For the latter one, the evolution, as a function of the concentrations, of the band parameters (wavenumbers, integrated intensities and band profiles) gives information on anion solvation and cation-anion interactions. This type of analysis was applied to alkaline and earth alcaline solutions of chloride, sulfate and nitrate to compare these solutions from the point of view of the different molecular interactions identified by Raman spectroscopy.