High Resolution Microprobe Analysis of Individual Fluid Inclusion Using Synchrotron X-Ray Fluorescence (SXRF)

Bénédicte Ménez (menez@ccr.jussieu.fr)¹, Pascal Philippot (philippo@ccr.jussieu.fr)¹, Alexandre Simionovici (simion@esrf.fr)², Michael Drakopoulos (drako@esrf.fr)², Irina Snigireva (irina@esrf.fr)² & Anatoly Snigirev (snigirev@esrf.fr)²

¹ CNRS-ESA 7058, Laboratoire de Petrologie, T26-E3, Universite Paris 6, 4 place Jussieu, 75252 Paris cedex 05, France

² ESRF, BP 220, 38043 Grenoble, France

Geological fluids are of crucial interest, linking the chemistry of all rock types and the concentration of economic and energetic resources. Massive fluid fluxes occur in the Earth's crust where the solar energy and heat from the Earth's interior meet and force hydrothermal waters to migrate over km-scale distances. The only relic testimony of past movements of hydrothermal waters occur as fluid inclusions containing minute amounts of paleo-fluids. Most natural minerals, however, contain several generation of fluid inclusions of distinctly different compositions. Because information relevant to the evolution of hydrothermal systems is lost when bulk-leach analysis are determined, quantitative analysis of single fluid inclusions are required.

Owing to its high sensitivity, non-destructive character and high spatial resolution, the Synchrotron X-Ray Fluorescence (SXRF) technique is well suited for individual fluid inclusion analysis. In order to establish absorption correction and calibration procedures required for quantification, an experimental protocol has been carried out on a variety of synthetic and natural fluid inclusions using the ESRF (beamline ID22, micro-FID) synchrotron facility.

Results indicates that elemental concentration estimates can be achieved without a precise knowledge of the analytical volume and the inclusion 3D geometry. The SXRF technique have proved to be particularly well adapted for tracking a large variety of ionic species in single fluid inclusions (down to the ppm and possibly tenths of ppb level). The concentration and spatial distribution of major (Cl, K, Ca, Mn, Fe) and trace (Ti, Cr, Ni, Cu, Zn, As, Br, Rb, Sr, Zr, Ba, Pb, Th, U) elements were determined in individual fluid inclusions from a variety of ore deposits worldwide. Among others, results gives an insight into metal and halogen partitioning between the vapor, liquid and solid phases of the fluid inclusion.

Accurate Determination of the Degree of Order in Orthopyroxene: Leverage Analysis in Structure Refinement

Marcello Merli (merli@crystal.unipv.it), M. Chiara Domeneghetti & Vittorio Tazzoli

Dip.to Scienze Terra, Via Ferrata 1, 27100 Pavia, Italy

Study of the kinetic and thermodynamic aspects of the ordering-disordering process in orthopyroxene requires precise and accurate determination of site occupancies, that can be affected by the choice of refinement strategies (range of (theta), weighting scheme, selection of the "observed" reflections, adequate corrections for absorption and secondary extinction, use of constraints and/or restraints etc.). Therefore a check of the influence of each reflection on the estimate of each refined variable proves to be very useful in structure refinement to avoid potentially dangerous truncation and to obtain reliable results. This check can be done using leverage analysis (Prince & Nicholson, 1985), already applied by Merli et al., 2000 to garnet, amphiboles and olivines. This statistical method, when applied in orthopyroxene to M1 and M2 occupancies, shows that: i) the pattern of the leverage distribution as a function of sin(theta)/(lambda) is similar for the two sites; ii) the highest leverage reflections are at a low-medium (theta) value for both sites; iii) the incorrect measurement of even only one high leverage reflection yields a significant variation in the degree of order. Leverage analysis also shows that reflections that have a high leverage on site occupancies of M1 and M2 sites have a very low leverage on its atomic coordinates and vice versa. This means that in structure refinement the estimate of site occupancies can be considered as independent from that of atomic coordinates and therefore the combined use of interatomic distances and site occupancies in the calculation of the site population does improve the accuracy of this procedure.

Prince E & Nicholson WL, Structure and Statistics in crystallography. Adenine Press - Guiderland NY USA, 183-195, (1985).

Merli M, Ungaretti L & Oberti R, American Mineralogist, in press, (2000).

Hydrothermal Synthesis and Geochemistry of Wolframite

Silvano Mignardi (mignardi@uniroma1.it)¹, Carlo Aurisicchio (aurisicchio@uniroma1.it)¹ & Vincenzo Ferrini (ferrini@uniroma1.it)²

¹ C.N.R. - Centro di Studio per gli Equilibri Sperimentali in Minerali e Rocce, c/o Dip.to di Scienze della Terra - Universita` La Sapienza - Roma, P.le A. Moro, 5 - 00185 Roma, Italy

² Dipartimento di Scienze della Terra, Universita` La Sapienza - Roma, P.le A. Moro, 5 - 00185 Roma

Wolframite crystals were synthesized hydrothermally at 0.5 kbar between temperatures of 300° and 500°C for studying the distribution of Fe, Mn and Cu between solution and solid. Synthesis was carried out in the range of 10-90 mole% FeWO₄ in steps of 10 mole% in "cold seal" vessels. As starting substances for the synthesis, stoichiometric mixtures of chlorides of bivalent iron and manganese, sodium tungstate and water were used. The pH values of the solutions have been varied in one series of runs between the weakly acid to weakly alkaline range. Cu-bearing wolframites were also synthesised from solutions containing 2% Cu besides Fe, Mn and W. The synthetic crystals were analysed by stereomicroscope, SEM, XRD and EMPA. All the crystals have variable colour from pale yellow to dark brown, depending on their Fe content. SEM analyses have shown that they are prismatic, ranging from about 5 to 800-1000 µm in length. Hubnerite crystals of > 70 mole% MnWO₄ are more elongated than ferberite crystals, that are slightly squatter, and show a much wider variability of size. The crystals synthesized from 2% Cu solutions range from about 2 to 150 µm in length. XRD analyses have shown that in all experiments only wolframite crystals were synthesized. EMPA analyses have shown that the crystals are compositionally homogeneous and a good correlation between Fe and Mn contents in both the solutions and crystals exists. On the contrary, a negative correlation exists between Cu contents in the solid and Fe concentrations in the solutions. The influence of the investigated parameters on the wolframite composition is discussed.