Direct Determination of the Volumes and Expansivities of Relaxed Silicate Melts Just Above the Glass Transition

Michael Toplis (mtoplis@crpg.cnrs-nancy.fr)¹ & Pascal Richet (richet@ipgp.jussieu.fr)²

¹ CRPG-CNRS, BP20, F-54501, Vandoeuvre-les-Nancy, FRANCE

² IPGP, 4, place Jussieu, F-75272, Paris, France

The density of silicate melts controls many aspects of the physical evolution of crustal magma chambers, such as the extent of crystal settling or convection. Most experimental determinations of silicate melt density have been at super-liquidus temperatures, well above those of natural melts. Extrapolation of the experimental results to lower temperature generally assumes a constant expansivity of the melt phase. However, this assumption is currently the source of controversy, in particular for diopside melt.

We have therefore developed a new dilatometric technique to measure equilibrium volumes of silicate melts in and just above the glass transition range. Starting with a glass of known density at room temperature, the glass transition is first observed as a break in the slope of the thermal expansion curve. On both sides of this temperature, the length of the sample is then measured as a function of time at various constant temperatures spanning an interval of 50-60 K. Attainment of equilibrium is demonstrated by the fact that the same length is measured starting from fictive temperatures higher and lower than the run temperature. Despite the small temperature range, the precision of the dilatometric measurements is sufficient to calculate expansivities to better than 10%.

Measurements have been performed in the system anorthite-diopside. The results show that melt expansivity at the glass transition decrease linearly from diopside to anorthite. Measured expansivities are considerably greater than those determined for the same compositions at high temperature, demonstrating that expansivities of silicate melts may decrease markedly with increasing temperature. Care must therefore be taken when applying existing models for melt density outside of the temperature range of the measurements.

The C2/c -P2₁/c Phase Transition in Intermediate Clinopyroxenes Along the Join Diopside-Enstatite (Di-En, CaMgSi₂O₆-Mg₂Si₂O₆)

Mario Tribaudino (triba@dsmp.unito.it) & Mauro Prencipe (prencipe@dsmp.unito.it)

Dipartimento di Scienze Mineralogiche e Petrologiche, Università di Torino - Via Valperga Caluso, 35, 10125 Torino, Italy

The C2/c -P2₁/c phase transition was investigated by means of transmission electron microscopy at room and high temperature and by X-ray powder diffraction at high pressure using synchrotron radiation.

Samples of composition Di₇₀En₃₀, Di₆₀En₄₀ and Di₅₀En₅₀ were synthesized at pressures between 1.3 and 1.8 GPa and temperatures between 1430 and 1550°C. Combined TEM/EDS observation showed that reflections with h+k odd are always present for clinopyroxenes with Ca content lower than 0.6 Ca atoms p.f.u, indicating that the C2/c -P2₁/c phase transition occurs at room temperature at Di₆₀En₄₀. Antiphase domains of size between 100 and 1000 Å were observed, elongated and almost parallel to the c axis. The domain size decreases with increasing Ca content. A mottled texture was present in grains with Ca lower than 0.65 a.p.f.u. HTTEM investigation showed that in samples with composition between Di₆₀En₄₀ and Di₅₀En₅₀ the critical h + k odd reflections are rather sharp up to 400-500°C and become broader at higher temperature, disappearing completely at temperature higher than 700-800°C.

High pressure investigation on a sample of composition Di₅₀En₅₀ P2₁/c at room conditions showed a phase transition at pressures between 3.7 and 5.9 GPa marked by a decrease in 0.7% in volume and by a strong decrease in the ß and c parameters. The disappearance of the critical (231) reflection, monitored with increasing pressure and the observed changes in the cell parameters indicate that the observed transition corresponds to the HP P2₁/c - C2/c transition described in clinoenstatite by Angel et al. (1992). In this work a definite transition point could not be determined possibly due to compositional inhomogeneities. Moreover enlargement of the (231) reflection is observed before the transition.

The growth of antiphase boundaries with C2/c symmetry at the expense of primitive antiphase domains can account the above observations.

Angel RJ, Chopelas A & Ross, NL, Nature, 358, 322-324, (1992).

A New Pressure Vessel for Magma-H₂O Interaction Studies: First Results from Isobaric vs. Impulsive Injected Water Experiments

Raffaello Trigila (trigila@axrma.uniroma1.it)¹ & Luigi Castiglione²

¹ Dipartimento di Scienze della Terra, Università di Roma La Sapienza, P.le A.Moro 5, 00185 Roma, Italy

² V.Baldissera, 66, 00159 Roma, Italy

As already recognized on volcanoes, water under pressure coming in contact with magma bodies or hot rocks may interact explosively originating phreatomagmatic or phreatic eruptions. With the aim to understand the boundary conditions causing the explosive interaction, it has been designed by R.T. and realized by LECO Co. (U.S.A.) an experimental apparatus able to perform interaction experiments with water (or other suitable fluids) up to Pmax of 200 MPa and Tmax of 1200°C. The interaction cell is contained within a two liters vertical IHPV (Internally Heated Pressure Vessel) and is equipped with a membrane allowing the volume variations due to H₂O liquid-vapor transition. The water has a separate pressurization system up to 200 Mpa controlled by a motorized valve. In situ measurements of the cell bottom and top temperatures, pressures and volume are taken during the experiments by a computer assisted data logger. Samples of Onano Eruption Unit (Vulsini Volcanes) in form of: i) finely grounded powder from juvenile spatter, ii) scoria grains 1 mm in size, have been treated at temperatures between 800° and 1000°C and confining pressures up to 250 bar. Water was injected in the reaction cell at pressures between 700 and 1400 bar in amounts corresponding to Melt/Water (wt) = 3. Preliminary results, concerning the effect of melt temperature, different pressures of water injection and material different grain-sizes, as indicated by the cell expansion and the amount of transported products, show these parameters are not intrinsically correlated. In the isobaric H₂O injected experiments the cell expansion at a reference temperature (V_(1000°C) = 12-18%) is essentially controlled by H₂O injection rate and amount while transport (as average in the range of 1%wt) appears to be controlled instead by the rheological properties of the sample which in turn depend on grain size, melt composition and temperature. In the impulsive injected water experiments, cell expansion under the same conditions doesn't change but transport of the granular sample at near solidus conditions reaches 9wt%.