SpringerLink
Forum Springer Astron. Astrophys.
Forum Whats New Search Orders


Astron. Astrophys. 339, 904-916 (1998)

Previous Section Next Section Title Page Table of Contents

3. Synthesis and analytical characterization of the samples

Iron-free Mg silicate minerals are relatively sparse in nature. Therefore, we synthesized the pure end members MgSiO3 and Mg2SiO4 in the laboratory. This was done by melting of SiO2 and MgCO3 in the right stoichiometric ratios. The melts were kept one hour at [FORMULA]C and then slowly cooled down (103 K/h) to room temperature. They crystallized very rapidly, so that a microcrystalline solid was formed.

Natural minerals often contain inclusions with completely different compositions. In a natural fayalite sample (Indonesia) we found a significant amount of inclusions consisting of sulfides and oxides of Fe, Cr, Ti and some other elements as well as separated phases of the iron silicate itself. For this reason, we prefered synthetic material to a natural mineral for spectroscopy. The synthetic fayalite was produced by melting a mixture of SiO2 (silicon dioxide) and FeC2O4[FORMULA]6H2O (iron oxalate) in the right stoichiometric ratio in an arc discharge furnace and cooling down the melt slowly to room temperature (103 K/h). Because of the tendency of the Fe[FORMULA] to be oxidized to Fe[FORMULA] during the melting and cooling, both processes were performed in an argon atmosphere. The synthesis of mixed Fe/Mg silicates of olivine or pyroxene composition in the way described for the pure magnesium and iron silicates leads to phase-separated (= inhomogeneous) materials.

Finally, we selected 8 samples of natural and synthetic pyroxenes and olivines, covering a wide range of Mg/Fe ratios (see Table 1). The analytical results are given in mass % of MgO, FeO, and SiO2, a usual way to represent mass fractions of the single components present in minerals or synthetic materials. The use of the oxides does not mean that these are present as separate phases in the minerals. Only in the case of the inclusions are the oxides present as a separate phase. Among the samples five natural minerals, olivine, hortonolite, enstatite, bronzite, and hypersthene were studied. The actual composition of the natural minerals and synthetic materials and the homogeneity of the samples was proved by Scanning Electron Analysis (SEM) and Energy Dispersive X-ray analysis (EDX) of polished samples embedded in epofix resin. The EDX results have been confirmed by wet-chemical analyses of the olivine and forsterite samples. The determination of the Fe[FORMULA] content in the minerals was based on the procedure described in former papers (Jäger et al. 1994, Dorschner et al. 1995). The SEM investigations of the natural minerals have shown that they also contain minor amounts of additional oxides like CaO or MnO2 and that most of them contain differently composed inclusions, which are given in the last column of Table 1. In all of our selected materials the inclusions are only minor components, so that all samples chosen for spectroscopic measurement are considered to be pure materials. In the case of the natural pyroxenes we detected a small amount ([FORMULA] 5%) of weathered silicate material in the form of Mg3[(OH)2/Si4O10] (talc) which must be homogeneously distributed in the samples. Additionally, the spectral influence of the inclusions was investigated by comparison of reflectance measurements of the inclusions by the microscope with the reflectance of the whole sample. We verified that they are minor components only and do not significantly influence the bulk spectra. However, we stress that in any case, it is important to check the spectral influence of the minor components in minerals used as laboratory analogues for cosmic dust.


[TABLE]

Table 1. Analytical results for the pure synthetic and natural silicate materials determined by EDX analysis; the symbol s means synthetic material, n stands for natural mineral. The last column contains the amount of minor oxide components (in mass %) and the kinds of inclusions.


The achievement of the thermodynamical equilibrium in the dust condensation zones of stars is not very probable (Nuth 1996, Frenklach 1997). Therefore, it is also interesting to investigate phase-separated materials in the olivine and pyroxene systems that also experienced non-equilibrium processes. Furthermore, new aspects of chemical processing of already formed silicates by heating and cooling processes in the interstellar medium and star-forming regions can be found by investigations of phase separations. Because of these astrophysical implications we also performed an extensive analytical characterization of such inhomogeneous samples. Fig. 1 shows the SEM micrographs of two selected materials of this type and Table 2 contains the analytical phase characterizations of them.

[FIGURE] Fig. 1a and b. Scanning electron microscope images of the inhomogeneous olivine (a ) and pyroxene (b ). Explanation see text and Table 2.


[TABLE]

Table 2. Analytical compositions of the inhomogenous silicate materials, determined by EDX analysis. The terms "dark", "gray", and "light" refer to the appearance of the corresponding phases in Fig. 1.


We found that the phase separations are much more frequent in synthetic pyroxene materials than in olivines. This is caused by the incongruent melting of pyroxene. This means, that for example a MgSiO3 melt at liquidus temperature begins to crystallize to forsterite, and a SiO2-rich melt remains, according to the chemical equilibrium (Matthes 1990)

[EQUATION]

At [FORMULA]C, the so-called peritectic reaction between forsterite and SiO2 begins, which means that the chemical equilibrium is shifted to the left side of Eq. (1), and all forsterite will be consumed. If the cooling is too rapid, the peritectic reaction cannot be completed. In this case forsterite, enstatite and cristobalite coexist in the crystallized mineral. As a matter of fact, these 3 minerals were found in our phase-separated synthetic pyroxenes. In contrast to the pyroxenes, olivines melt congruently, i.e. the melt and the crystallized phase have the same composition. Cooling an olivine melt too rapidly could lead to the zone crystallization in a Mg-rich and a Fe-rich olivine. This explains the occurence of only 2 phases. Additionally, the Fe[FORMULA] in the fayalite-rich component can be partially oxidized, leading to the formation of a rather Mg-rich olivine and to FeO and Fe2O3.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1998

Online publication: October 22, 1998
helpdesk.link@springer.de