HR: 17:45h
AN: V42H-08 INVITED     [PDF]
TI: Magma Genesis in the Hawaiian Hot Spot: From melting experiments on basalt/peridotite hybrid source
AU: * Takahashi, E
EM: etakahas@geo.titech.ac.jp
AF: Magma Factory, Earth & Planetary Sciences, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguroku, Tokyo, 152-8551 Japan
AB: Melting mantle peridotite is one of the central themes in experimental petrology. Melting studies in CMAS, NCMAS and natural peridotites have extensively documented the magma genesis process at Mid Oceanic Ridges (e.g., Presnall et al., 1979). Magma genesis in OIBs and LIPs, on the other hand, has been poorly constrained by experiments. Evidences from isotope geochemistry indicate that the source materials for basalt magmas in these provinces are not peridotite alone. Based on a geological and geochemical reconstruction of 3 Ma old Koolau volcano, I proposed that the size of eclogite blocks in the Hawaiian plume would exceed 1000km3 (Takahashi and Nakajima, 2002) and therefore the melting interaction of eclogite blocks and the surrounding peridotite would play essential roles in magma genesis in the Hawaiian hot spot. Melting experiments on basalt/peridotite composite starting materials were carried out at 2.5 to 3.0 GPa at temperatures from the peridotite dry solidus to that of basalt for 20 to 100 hours. Three layered starting materials consisting of 1 basalt to 2 peridotite (in volume) were placed in graphite/Pt double capsules. Peridotite KLB-1 (Fo89.6) and two basalt-starting materials (CLG-46 and CRB72-31) were used as starting materials. In temperatures ca.50-100 degrees below the peridotite solidus, silica-rich partial melts are produced in the basalt zone and the boundaries between the basalt and peridotite are coated with a 10 to 50 micron thick opx reaction band. The chemical reactions between the basalt and peridotite domains are controlled by solid diffusions across the opx reaction band and are very slow. In temperatures within 50 degrees of the peridotite dry solidus, a time dependent reaction process takes place. The basalt/peridotite boundary gradually partial melts as the chemical reaction lowers the peridotite solidus locally. At 2.8 GPa and 1450-1470C after 50-100 hours, resultant melt in the basalt layer becomes saturated with oliv + opx + cpx + gar and has 13-15 wt% MgO. Compared with partial melt of KLB-1 at 1500-1525C (Hirose and Kushiro, 1993), the basalt/peridotite hybrid melts are much lower in Al2O3 and CaO but are distinctly higher in K2O and TiO2. Compared with partial melt of KLB-1+50%MORB at 1500-1525C (Kogiso et al., 1998), melt in this study are much higher in SiO2 but lower in Al2O3. Among the three sets of experiments at 2.8-3.0GPa, compositions of the basalt/peridotite hybrid melts are most similar to the primitive Hawaiian tholeiite. Various hypotheses have been proposed to explain the chemical character of the Hawaiian magma (e.g., involvement of a metasomatic component at the base of the Pacific plate, melting of a fertile lower mantle component, melting at much higher pressure, etc). Based on the above experiments, I propose that Hawaiian tholeiite magmas are produced in basalt-peridotite reaction domains in the plume. The systematic change in the SiO2 content of Hawaiian tholeiites (from lower to higher, Loihi, Kilauea and Mauna Loa) could be explained by lowering melting temperatures in magma feeding zones from Loihi to Mauna Loa. This model predicts that the PMT for the Hawaiian plume could be roughly 100C lower than the previous model (PMT=1558C, Watson \& McKenzie, 1991) and the involvement of eclogite component in magma genesis in the Hawaiian tholeiite would be 30 to 100%, which is much higher than the previous estimate (0-10%, Hauri, 1996).
DE: 1010 Chemical evolution
DE: 1025 Composition of the mantle
DE: 3630 Experimental mineralogy and petrology
DE: 3640 Igneous petrology
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting