Japan Geoscience Union Meeting 2019

Presentation information

[E] Poster

S (Solid Earth Sciences ) » S-CG Complex & General

[S-CG51] Role of volatiles on Earth and planetary dynamics

Thu. May 30, 2019 5:15 PM - 6:30 PM Poster Hall (International Exhibition Hall8, Makuhari Messe)

convener:Takayuki Ishii(Bavarian Research Institute, University of Bayreuth), Eiji Ohtani(Department of Earth and Planetary Materials Science, Graduate School of Science, Tohoku University), Michihiko Nakamura(Division of Earth and Planetary Materials Science, Department of Earth Science, Graduate School of Science, Tohoku University), Bjorn Mysen(Geophysical Laboratory, Carnegie Inst. Washington)

[SCG51-P03] Pargasite stability in the upper mantle at H2O-undersaturated conditions

*Marija Putak Juricek1, Hans Keppler1 (1.BGI)

Keywords:pargasite, stability, upper, mantle

Amphibole is the most commonly occurring hydrous mineral in mantle xenoliths and the majority of upper mantle amphiboles have compositions close to pargasite end member. Previous experimental work has demonstrated that pargasite stability is sensitive to variations in bulk alkali and water content of the starting material. While it has been successfully shown that addition of more alkali leads to stabilization of pargasite to higher P and T, the effect of water on amphibole stability is not so straightforward.

At H2O-saturated conditions, alkali elements partition strongly into aqueous fluid and become less available for crystallization of amphibole. As a result, amphibole seemingly becomes stable at higher P and T when less H2O is added to the starting material. This observation led to the hypothesis that pargasite may be a ubiquitous phase in the upper mantle, which is believed to contain only several hundred ppm of H2O. However, most of the previous studies determined pargasite stability at H2O-saturated conditions where solid phases coexist with an aqueous fluid during experiments. Since such environments are not representative of the nearly dry upper mantle, the purpose of this study is to investigate pargasite stability at H2O-undersaturated conditions.

Because of the high specific surface area of powders, adsorption of atmospheric H2O on fine grained starting materials prevents attaining H2O-undersaturation by adding small amounts of water to the sample. Instead, H2O-undersaturated conditions may be precisely controlled by diluting the water in the sample with another component. Under such conditions, solid phases are expected to coexist with a low water activity fluid throughout the experiment. As molecular nitrogen is mostly insoluble in upper mantle minerals, it can be used to adjust the water activity of the fluid to a desired low value.

All experiments were performed in a piston cylinder apparatus. The samples for H2O-undersaturated experiments were prepared by adding silver azide (AgN3) and brucite to a synthetic peridotite mixture. In several experiments water was added directly as H2O(l). The amount of water was kept constant in all experiments (1 wt. percent) and the molar fraction of water in the N2-H2O fluid was adjusted to the desired value by weighing an appropriate amount of AgN3, which decomposed to metallic silver and molecular nitrogen during heating. In H2O-saturated experiments, no silver azide was added to the peridotite mixture. Experimental products were analyzed using JEOL JXA-8200 electron microprobe.

Preliminary results show that pargasite is not stable in equilibrium with fluids diluted to XH2O=0.1 at pressures above 1 GPa. At H2O-saturated conditions pargasite stability field can extend to pressures of 4 GPa and above 1100 ℃, depending on bulk alkali and water contents. Results from experiments at XH2O=0.1 and XH2O=1 suggest that pargasite can exist only in regions of the upper mantle which have been affected by high water activity fluids, such as areas in the vicinity of subduction zones. The ambient nominally anhydrous upper mantle, better represented by experiments at H2O-undersaturated conditions, is not expected to contain pargasite.