HR: 17:30h
AN: V22H-07    [PDF]
TI: Dissolution kinetics of oceanic lower-crustal cumulate-minerals and the potential effect of the melts on ascending magmas
AU: * Kvassnes, A J
EM: astri@mit.edu
AF: WHOI, MS 8, Woods Hole, MA 02543 United States
AU: * Kvassnes, A J
EM: astri@mit.edu
AF: MIT, EAPS 54-1220, Cambridge, MA 02139 United States
AU: Grove, T L
EM: tlgrove@mit.edu
AF: MIT, EAPS 54-1220, Cambridge, MA 02139 United States
AU: Dick, H J
EM: hdick@whoi.edu
AF: WHOI, MS 8, Woods Hole, MA 02543 United States
AB: The most primitive Atlantis Bank (SWIR) olivine-gabbros have augite oikocrysts surrounding more evolved plagioclase chadacrysts. In addition, this coarse type of augite commonly shows reverse zoning. The observations motivated an experimental study. We investigated the kinetics of melting of the grain-boundaries between mineral-pairs commonly found in lower ocean-crust, and discovered rapid melting rates and melt-compositions that may explain the phenomenon. \\ In our study, An$_{62}$ and An$_{54}$ plagioclase were melted together with Fo$_{73}$ or Fo$_{82}$ and with Mg\#86 augite. The experiments were distributed over a melting interval of 1240-1330$\deg$C and 1180-1300$\deg$C respectively. No melting was observed below 1210$\deg$C (augite - An$_{54}$) and 1255$\deg$C (Fo$_{82}$ - An$_{54}$). Plagioclase is buoyant in the melt, therefore the minerals were melted with the heavier mineral on top, to preserve a short distance between them. The duration of the experiments varied from 30 minutes to 24 hours. The solidus for the plagioclase-clinopyroxene-olivine system was determined to be 1150$\deg$C, and the solidus for augite-plagioclase and olivine-plagioclase was inferred to be 5$\deg$C and 40$deg$C higher, respectively, on the basis of previous studies.\\ Olivine, the mineral that experiences the fastest internal solid-state diffusion, has very narrow (tens of $\mu$ms) or no observable diffusion gradient along the actively melting surface, indicating that the melting rate is similar to or faster than the diffusion rate for Fe/Mg in olivine. Some recrystallization occurred in the melt close to olivine, away from the most active melt interface.\\ Plagioclase and clinopyroxene grains melted without internal diffusion of major elements in the crystals. Augite starts disintegrating internally at the highest temperatures, but does not show any sign of preferential melting of exsolution lamellae or preferential melting of different crystal faces. Plagioclase show a very narrow (10$\mu$m) jagged reaction zone, but no significant anisotropy of melting is apparent.\\ The melting rates for augite and plagioclase appear to be dependent on the $\Delta$T above solidus for the grain-boundary in question. Therefore, at a given temperature, plagioclase melts slower when in contact with an olivine than when in contact with an augite. For instance, An$_{54}$ plagioclase at 1290$\deg$C melts 0.54-mm/hr when in contact with augite, but only 0.17-mm/hr in contact with olivine. The minerals of the augite-plagioclase pairs melt at comparable rates to the plagioclase of the plagioclase-olivine pair, relative to the solidus. However, the olivine of the pair melts at a slightly slower rate.\\ The results indicate that the grain-boundaries of normal gabbros from ocean ridges will melt very efficiently at a temperature of 1240 or higher. Stoping of xenoliths is the most efficient method of reheating wall-rock material by ascending melts. A 20cm xenolith may be thermally reequilibrated in 8 hours. Heat produced during crystallization will balance that consumed in the melting of existing cumulates. The latter melts will have higher Mg\#'s than expected expected from equilibrium conditions i.e. appear to be more primitive melts than the ones that produced the cumulates. Mixing of these high Mg\# melts and the xenocrysts with the enclosing magma will make the resulting magma appear more primitive than the original. This process may have caused the large augites at Atlantis Bank to have higher Mg\# than what equilibrium-conditions would suggest.
DE: 1020 Composition of the crust
DE: 3035 Midocean ridge processes
DE: 3630 Experimental mineralogy and petrology
DE: 3640 Igneous petrology
DE: 3655 Major element composition
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting