The Crystal Structure of Synthetic TlAlSi₃O₈ and RbAlSi₃O₈: The Influence of Inert Pair Effect of Thallium (I) on the Feldspar Structure

Atsushi Kyono (kyono@arsia.geo.tsukuba.ac.jp) & Mitsuyoshi Kimata (kimata@arsia.geo.tsukuba.ac.jp)

Institute of Geoscience, University of Tsukuba, 1-1-1 Tennodai, Tsukuba-shi, Ibaraki, 305-8571, Japan

The crystal structures of synthetic Tl-feldspar, TlAlSi₃O₈, and Rb-feldspar, RbAlSi₃O₈, were determined based on single crystal X-ray diffraction data; monoclinic, a = 8.882(3), b = 13.048(2), c = 7.202(2)Å, ß = 116.88(1)°, V = 744.5(4)ų, Z = 4, space group C2/m (R index of 9.8% for 720 observed reflections), and a = 8.839(2), b = 13.035(2), c = 7.175(2)Å, ß = 116.11(1)°, V = 742.3(3)ų. Z = 4, space group C2/m (R index of 10.7% for 648 observed reflections). The underlying frameworks for both structures can be regarded as isotypic with sanidine. Both Tl and Rb at the unique extraframework sites are coordinated by nine oxygen atoms: Tl-O distances range from 3.03 to 3.21Å, forming a unique expanded polyhedron as compared with the Rb-polyhedron of Rb-O distances ranging from 2.96 to 3.18Å, in spite of the small ion radius of Tl⁺ (1.59Å) relative to Rb⁺ cation (1.61Å). Penetration of the crystal structures completely incorporating Tl and Rb cations discerns the evident difference of feldspar conformation imposed by the tetrahedral framework. The inert pair effect of Tl atom on the coordination polyhedra, differing morphologically from the corresponding Rb-polyhedron in Rb-feldspar, seems to be suppressed by environmental hard oxygen anions to accommodate the distortion which results from the stereochemical activity of inert pair electrons. Inasmuch as the direction of displacement of the Tl atom from the best center of the ligand arrangement is assumed to be opposite to the orientation of the lone pair on Tl atom, the inert pair electrons may be identified as parallel to [101] along large channels of the feldspar structure. This orientation is consistent with the VSEPR model: a nonbonding pair occupies more space on the "surface" of the central atom than a bonding pair.

Homogeneous Bubble Nucleation in a Rhyolitic Liquid: The Effect of Magma Ascent Rate

Didier Laporte (laporte@opgc.univ-bpclermont.fr)¹, Catherine Mourtada-Bonnefoi (c.mourtada-bonnefoi@bris.ac.uk)² & Philippe Cacault (cacault@opgc.univ-bpclermont.fr)¹

¹ Laboratoire Magmas et Volcans, OPGC, CNRS and Univ. B. Pascal, 5 rue Kessler, F-63038 Clermont-Ferrand, France

² Department of Earth Sciences, University of Bristol, BS8 1RJ, United Kingdom

We recently completed a series of isothermal decompression experiments aimed at characterizing the conditions of homogeneous bubble nucleation (HBN) in rhyolitic liquids with different volatile compositions (4.5-8 wt% H₂O, 0-800 ppm CO₂; Mourtada-Bonnefoi and Laporte, 1999). The experiments were performed in a rapid-quench, externally-heated pressure vessel at 800°C: the rhyolitic liquid was first subjected to a pressure of 200-295 MPa depending on the volatile content; pressure was then decreased at 1-10 MPa/s to a pressure P_N lower than the volatile saturation pressure, P_Sat, to produce bubble nucleation; the sample was quenched after 2 min at P_N. Using this procedure, we were able to measure the supersaturation pressures P required for HBN and the resulting bubble number densities (P is the difference between P_Sat and the maximum value of P_N at which HBN was observed). Our latest experiments also revealed a strong effect of volatile composition on HBN: for 7.5 wt% water, P increased by 50-100 MPa with CO₂ increasing from 10 to 600 ppm; for low CO₂ contents (0-100 ppm), P decreased by more than 50 MPa with water content increasing from 4.5 wt% to 8 wt%.

A major aspect not addressed in our previous experiments was the effect of the decompression rate (magma ascent rate) on HBN. To study this effect, we connected to the pressure vessel a device composed of a set of air-operated valves and allowing to decrease the pressure from 200-300 MPa to near atmospheric pressure by steps of about 0.1 MPa. The system is controlled by a PC computer and is devised to produce decompression rates of 1-100 MPa/hour. The principle of the decompression apparatus, its technical performances and the first experimental results will be presented.

Mourtada-Bonnefoi CC & Laporte D, Geophys. Res. Lett, 26, 3505-3508, (1999).

Determination of Fe/Mg and Fe²⁺/Fe³⁺ Partitioning in Silicate Perovskite/Magnesiowüstite Assemblages Using Electron Energy Loss Spectroscopy (EELS)

Stefan Lauterbach (stefan.lauterbach @uni-bayreuth.de)¹, Catherine A. McCammon¹, Falko Langenhorst¹, Friedrich Seifert¹ & Peter van Aken²

¹ Bayerisches Geoinstitut, Universitätsstrasse 30, D-95440 Bayreuth, Germany

² Institut für Mineralogie, Schnittspahnstraße 9, D-64287 Darmstadt, Germany

The oxidation state of iron in lower mantle perovskite and magnesiowüstite can affect many properties, including electrical conductivity, diffusivity and rheology. As determined by Mössbauer spectroscopy, the Fe³⁺/(sum)Fe ratio in the perovskite phase is relatively high and is controlled primarily by the aluminium concentration (Lauterbach et al. 1999). In contrast, similar studies have shown that the Fe³⁺/(sum)Fe ratio is relatively low in single-phase magnesiowüstite. While study of the single-phase systems is important to establish the individual crystal chemistry of these phases, equilibrium studies of the two-phase assemblages are essential to characterise geophysically-important variables such as partitioning of Fe²⁺ and Fe³⁺ between co-existing phases. We have used EELS as outlined in the method of van Aken et al. (1998) to determine the relative Fe³⁺ content in both coexisting phases with high spatial resolution.

The perovskite/magnesiowüstite assemblages were synthesised in Re-capsules at high pressure and temperature from pure oxide mixtures with a high initial Fe³⁺/(sum)Fe ratio. To achieve equilibrium in a multi-anvil press at conditions relevant to the lower mantle (26 GPa, ~1700°C), the runtime was extended to 22-26 h. The resulting run products were examined using X-ray diffraction, electron microprobe, and transmission electron microscopy (TEM), and Fe³⁺/(sum)Fe was determined using EELS.

We have determined the partition coefficients of Fe²⁺ and Fe³⁺ between the two phases. Preliminary results from (Mg,Fe)(Si,Al)O3 perovskite-(Mg,Fe)O assemblages confirm that Fe³⁺/(sum)Fe is highest in the perovskite phase, although energy-dispersive X-ray analysis performed on the same grains indicate that Fe²⁺ remains concentrated in (Mg,Fe)O phase. Further experiments are underway to characterise Mg/Fe and Fe²⁺/Fe³⁺ partitioning between the two phases as a function of bulk Al concentration.

Lauterbach S, McCammon CA, van Aken P, Langenhorst F & Seifert F, Contrib Mineral Petrol, (in press).

van Aken P, Liebscher B, Styrsa VJ, Phys Chem Minerals, 25, 323-327, (1998).