Experimental Illitization of Smectites in K-rich Solutions Monitored by a Multitechnique Analysis of Coexisting I/S and Solutions

Régine Mosser-Ruck (regine.ruck@g2r.u-nancy.fr)¹, Jacques Pironon (jacques.pironon@g2r.u-nancy.fr)¹, Michel Cathelineau (michel.cathelineau @g2r.u-nancy.fr)¹ & Alain Trouiller²

¹ G2R(UMR 7566), Faculté des Sciences, UHP, BP-239, 54506, Vandoeuvre, France

² ANDRA, 1-7 rue Jean Monnet, PArc de la Croix blanche, 92298 Châtenay-Malabry, France

The natural or experimental illitization of smectite is studied since several decades because of its great interest for the understanding of the diagenetic transformations in sedimentary basins and of the thermal stability of smectites or bentonites under radioactive waste disposals conditions. Until now, the mechanisms of illitization are not completely elucidated, and the nature of 10 Å layers (true illite or dehydrated smectite) in I/S remains debated. The aim of this study is to examine if dehydrated smectite layers are present in I/S minerals when experimental fluids are enriched in potassium (KCl 0.5 molal) e.g. in solutions favouring the formation of true illite layers. Analytical investigations concern both experimental solids (XRD, microprobe, FTIR spectroscopy) and solutions (ICPAES/ICPMS). Main results are the followings: (i) discrepancies between measured Si content in mixed layer clays and Si content estimated from illite layers determined by XRD, IR spectroscopy (on NH₄⁺ saturated experimental products) and index variable of dissolution rate of smectite (for instance evolution with time of Mg concentration in the experimental solutions, remarkable lack of by-products like quartz and the relatively low Si concentration in solution) confirm the existence of dehydrated K-smectite layers as already shown by a few previous studies (Ransom and Helgeson 1994, Mosser-Ruck et al. 1999), (ii) infrared spectroscopy specifies that some of these dehydrated layers have interlayer K⁺ ions not fixed and consequently still exchangeable whereas others are definitively collapsed; this behaviour is probably related to the positive charge deficiency location in the structure of smectite, (iii) with time, dissolution rate, illite layer and dehydrated smectite layer fractions both increase in newly formed ordered I/S. Therefore, high concentration of K⁺ in the experimental solution tends to favour the formation of dehydrated K-smectite instead of illite in short time (less than 6 months) experiments.

Ransom B & Helgeson HC, Am. J. Sci., 294, 449-484, (1994).

Mosser-Ruck R, Cathelineau M, Baronnet A & Trouiller A, Clay Min., 34, 275-290, (1999).

Interferometric Method for Velocity Measurements of Ultrasonic Compressional and Shear Waves in Piston-Cylinder and Multi-Anvil High Pressure Cells

Hans Joachim Mueller (hjmuel@gfz-potsdam.de)¹, Joern Lauterjung¹, Frank Schilling¹ & Georg Nover²

¹ GeoForschungsZentrum Potsdam, Telegraphenberg, D-14473 Potsdam, Germany

² Petrologisch-mineralogisches Institut, Universitaet Bonn, Poppelsdorfer Schloss, D-53115 Bonn, Germany

The progress of global seismology in general and of the tomographic method in particular illustrated a lot of exciting structural details of Earth's deep interior. Consequently the measurement of elastic wave velocities under experimental simulated mantle conditions should play a similar role in the interdisciplinary approach of high pressure mineral physics, mineralogy and petrology to the interpretation of those data. The paper describes adaptation, optimization and first results of ultrasonic interferometry in two different high pressure cell systems.

The piston-cylinder experiments were performed in a system with an inner vessel diameter of 22 mm. The pistons were designed for optimum ultrasound transmission regardless of the resulting limitations of maximum pressure. Each piston was equipped with 3 lithiumniobate transducers of 16.3 MHz natural frequency. Two of them have the same cut, i.e. generate/receive p- or s-waves, the third have the alternative cut for receiving the transmission signal of the transducer from the opposite piston. The basic design of the experimental setup followed the publication of Knoche, Webb & Rubie (1997), i.e. two cylindrical platinum reflectors with the sample between them. Aluminium, San Carlos olivine single crystals and anorthite polycrystal samples were measured up to 1 GPa pressure in several runs. The frequency sweep steps were between 25 and 125 kHz. The results show no profound signal quality dependence on wave type, pressure or sample material. The only difference of the experimental setup for multi-anvil cells is the cube-shaped outer pressure transmitting material. Because of the still non-optimized transducer frames resulting in growing anvil load the maximum pressure of the first multi-anvil runs was limited to 3 GPa. The results do not show any significant negative effect of outer shape or scattering inside the tungsten carbide anvils to the method.

Chen G, Li B & Liebermann RC, Science, 272, 979-980, (1996).

Knoche R, Webb SL & Rubie DC, Che. Earth, 22, 125-130, (1997).

Li B, Chen G, Gwanmesia GD & Liebermann RC, Geophys. Mon, 101, 41-61, (1998).

Li B, Jackson I, Gasparik T & Liebermann RC, Phys. Earth Plan. Int, 98, 79-91, (1996).

McSkimin HJ, J. Acoust. Soc. Am, 22, 413, (1950).

Niesler H & Jackson I, J. Acoust. Soc. Am, 86, 1573-1585, (1989).

Experimental Constraints on the Genesis of Pyroxenites in Subduction-Related Magmatic Arcs

Othmar Muentener (othmar@erdw.ethz.ch)¹, Peter B. Kelemen (peterk@whoi.edu)² & Tim L. Grove (tlgrove@mit.edu)³

¹ Dep. of Geology and Geophysics, WHOI, Woods Hole Ma, 02543, present address: Dep of Earth science, ETH Zuerich, Switzerland

² Dep. of Geology and Geophysics, WHOI, Woods Hole Ma, 02543, USA

³ EAPS, MIT, Cambridge MA, 02139, USA

Igneous pyroxenites are common in the crust/mantle transition zone in exposed sections of accreted island arc terrains [1,2] and continental basement [3]. They contain potentially important clues regarding the primary magmas which pass from the mantle to form continental crust. Among their characteristics are high Mg numbers, high Cr, but relatively low Al contents. Here we discuss the effects of temperature, melt composition and H₂O content on pyroxene composition using crystallization experiments on primitive mantle derived melts [4]. Experiments were performed on natural compositions by melting mixtures of H₂O saturated basaltic glass and natural rocks in a piston cylinder apparatus. Hydrous basaltic glasses have been pre-synthesized at 0.2 GPa in TZM cold seal pressure vessels. The initial content varied from 2.5 to about 5 wt% H₂O in the melt. To minimize Fe-loss, pre-saturated Au₈₀Pd₂₀ inner capsules were used for the hydrous piston cylinder experiments. The capsule was placed into a graphite sleeve which then was placed into a Pt outer capsule. Run times varied between 20 and 30 hours. Our results at 1.2 GPa combined with a compilation of data from the literature reveals that clinopyroxene/melt major element partition coefficients vary as a function of temperature and melt composition. For example, for a given bulk composition Al correlates positively with Al in pyroxene. Extrapolation of this correlation to natural pyroxenites [1,2] indicates that pyroxenes with low Ca-Tschermaks component represent near-liquidus phases of primitive hydrous (Si-rich?) magmas, consistent with their high Cr content. At lower temperatures spinel websterites, garnet and/or garnet-amphibole websterite formed. The results show that the combination of high pressure and high magmatic H₂O combine to suppress plagioclase. Calculated densities of these cumulate rocks are in part significantly higher than densities of peridotitic rocks, creating a potentially unstable density inversion between ultramafic plutonic rocks and (depleted) underlying mantle.

Debari SM & Coleman RG, JGR, 94, 4373-4391, (1989).

Miller DJ & Christensen NI, JGR, 99, 11623-11642, (1994).

Muentener O & Hermann J, EOS, 79, 46, (1998).

Baker MB, Grove TL & Pryce R, CMP, 118, 111-129, (1994).