2003/10/01 by D. M. Jenkins, David M. Jenkins · 2 citations
Earth and Planetary Sciences · Materials Science · #Geological and Geochemical Analysis #Crystal Structures and Properties #Nuclear materials and radiation effects
paper · doi:10.2475/ajs.303.8.723
A re-investigation of the synthesis, characterization, and upper-thermal stability of ferro-actinolite has been performed in the system CaO-FeO-SiO<sub>2</sub>-H<sub>2</sub>O. Synthesis experiments were done over the range of 2 to 10 kbar and 420° to 600°C for several bulk compositions along the join Ca<sub>2</sub>Fe<sub>5</sub>Si<sub>8</sub>O<sub>22</sub>(OH)<sub>2</sub> - Fe<sub>7</sub>Si<sub>8</sub>O<sub>22</sub>(OH)<sub>2</sub> and at oxygen fugacities (f<sub>O<sub>2</sub></sub>) defined by the vessel (∼Ni-NiO), Co-CoO, and iron-magnetite (IM) buffers. The highest yields of actinolite were obtained at 2 kbar, 420°C, and at the relatively oxidizing conditions of the Ni-NiO buffer. Grinding and retreatment of the sample was found to be important for increasing the yield of amphibole up to a maximum of about 70 weight percent; however, complete yields of actinolite were never obtained indicating some physical barrier or incorrect chemical (f<sub>O<sub>2</sub></sub>) conditions preventing complete reaction. The highest-yield actinolite syntheses were characterized using a high-resolution transmission electron microscope (TEM) equipped with an energy-dispersive spectrometer. A survey of representative grains using the TEM revealed the presence of very few chain-multiplicity faults (A′(2) = 0.98 –0.99); however, a rather large spread of compositions was observed, with crystals having Ca contents that range from as little as 0.5 up to 2.0 Ca atoms per formula unit (apfu). There was no clear correlation between actinolite crystal composition and the bulk composition of the starting material. By combining these analyses with three other techniques for determining the composition of the synthetic actinolite (that is, volume-composition relationships, Rietveld refinement of Ca and Fe on the M4 site, and mass balancing involving bulk-composition and mineral-proportions), the average composition of the synthetic actinolite was determined to be Ca<sub>1.67</sub>Fe<sub>5.33</sub>Si<sub>8</sub>O<sub>22</sub>(OH)<sub>2</sub>. Experimental reversals were obtained on reaction (1): 2 ferro-actinolite = 4 hedenbergite + 3 fayalite + 5 quartz + 2 water over the range of 1 to 5 kbar and 500° to 525°C with the f<sub>O<sub>2</sub></sub> defined by the Co-CoO buffer. A parallel set of experiments on reaction (1) was done in the presence of grunerite and the iron-magnetite/wüstite-magnetite (IM/WM) buffers. These latter experiments allowed the direct comparison of the upper-thermal stability of actinolite via reaction (1) and of Ca-saturated grunerite via reaction (2): 2 grunerite = 7 fayalite + 9 quartz + 2 water. Reaction (2) was found to lie 20° to 60°C above reaction (1). Modeling the miscibility gap between actinolite and grunerite as a regular solution with W = 15.3 kJ, a thermodynamic analysis of the internal consistency of the experimental reversals for reaction (1) was performed, from which the best-fit values of 696.4 J/K · mol and –10,518.2 kJ/mol were derived for the <i>S</i>° and Δ<i>H</i><sub>f</sub><sup>o</sup> of pure ferro-actinolite, respectively. Calculation of the invariant array of curves involving the phases ferro-actinolite, grunerite, hedenbergite, fayalite, quartz, and water indicate that reaction (1) is metastable at all geologically relevant pressures and that the stable reaction defining the stability limit of ferro-actinolite is 7 ferro-actinolite = 14 hedenbergite + 3 grunerite + 4 quartz + 4 water.