HR: 10:35h
AN: V51K-02 INVITED     [PDF]
TI: Fractionation of Boron (and Lithium) Between Hydrous Fluid and Silicate Melt: Diffusion, Contamination, and Orphaned Experiments
AU: * Hervig, R L
EM: hervig@asu.edu
AF: Arizona State University, CSSS PO 871704, Tempe, AZ 85287 United States
AU: Moore, G
EM: gordon.moore@asu.edu
AF: Arizona State University, CSSS PO 871704, Tempe, AZ 85287 United States
AB: We have measured the isotopic fractionation of boron between hydrous fluid and basaltic and rhyolitic melt using boron rich (approx. 2000 ppm B) anhydrous starting glasses. The results showed that heavy boron was preferentially partitioned into the fluid, even at high temperatures. We are now attempting reversal experiments by equilibrating B-rich hydrous fluid with B-poor silicate melts. A rhyolitic run product most recently recovered (100MPa, 900C, 48 hours) was mounted in epoxy and analyzed by SIMS for H, Li, B, and B isotopes. At the fluid-melt interface, boron concentrations were influenced by B-rich phases on the meniscus presumably quenched from the fluid. At distances of 10 to 100 microns from this interface the boron concentration dropped from approx. 600 ppm to 9 ppm, respectively (the initial B content of the rhyolite) while the H content was constant (3.2 wt.%) throughout. The boron isotope ratios become lighter with decreasing B content. This experiment 1) shows that the diffusion of boron is slow ($<$1E-13m2/s), but is $>$10000x faster than in anhydrous systems, and 2) confirms that hydrous silicate liquids concentrate the light isotope of B compared to hydrous fluid. We have also examined older experiments (performed for other purposes) to see if the boron and/or lithium contents and isotope ratios of these melts (equilibrated with hydrous fluids) can help us understand the high-temperature behavior of these light elements. For example, our earlier experiments measuring boron isotope fractionation between Li-rich (3400 ppm Li) macusanite rhyolite and hydrous fluid also show significant Li loss to the fluid allowing the determination of a fluid melt partition coefficient (D(Li) fl/melt) of 1.7 at 500 MPa. However, experiments on MORB glass containing initially ~6 ppm Li reveal a gain of ~1000 ppm Li during the run. Similar contamination of piston cylinder and internally heated experiments have been observed for boron as well as lithium. For example, in a study of D and H diffusion in rhyolite melt by (1), a lithium profile was also observed, with an implied diffusion coefficient near 1E-11m2/s. In contrast to these experiments, the above study on rhyolite showed a slight decrease in Li content in the melt implying a D(Li) fl/melt of 0.5 at 100 MPa. This Au capsule was cleaned in HF followed by sonication in a degreasing solvent, thereby minimizing light element contamination. Capsules from earlier experiments (Au, AuPd, AgPd) that were not degreased all show extreme light element contamination. Complementary experiments on boron fractionation between andesite and hydrous fluid are in progress. (1) Stanton, T. (1990) PhD Thesis, ASU.
DE: 1000 GEOCHEMISTRY (New field, replaces Rock Chemistry)
DE: 1040 Isotopic composition/chemistry
DE: 1065 Trace elements (3670)
DE: 3630 Experimental mineralogy and petrology
DE: 3670 Minor and trace element composition
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting