Chemical and isotope compositions of rocks from altered ocean crust at DSDP/ODP Sites 417/418

Large-scale compositional domains at DSDP/ODP drill sites 417A, 417D and 418A were analyzed for O, Sr and Nd isotope ratios, and REE, U, K, Rb and Sr abundances, to constrain the bulk chemical composition of the oceanic crust that is recycled at subduction zones. The combination of the three sites gives the composition of the upper oceanic crust in this region over a distance of about 8 km. The d18O(SMOW) and 87Sr/86Sr(meas) of compositional domains 10-100 m in size correlate well, with a range of 7.7-19.2 and 0.70364-0.70744, and mean of 9.96 and 0.70475, respectively. The Rb inventory of the upper crust increases by about an order of magnitude, while Sr contents remain constant. U abundances increase moderately under oxidizing alteration conditions and nearly triple in the commonly reducing alteration environments of the upper oceanic crust. REEs are influenced by alteration only to a small extent, and recycled oceanic crust is similar to MORB with respect to 143Nd/144Nd. Even though the average composition of the upper oceanic crust is well defined, the large scale composition varies widely. Highly altered compositional domains may not have a large impact on the average composition of the oceanic crust, but they may preferentially contribute to fluids or partial melts derived from the crust by prograde metamorphic reactions.

Mineral an chemical composition of basalts from DSDP Sites 417 and 418, West Atlantic Ocean

Major element composition ranges of closely associated basalt glass-whole rock pairs from individual small cooling units approach the total known range of basalt glass and whole rock compositions at IPOD sites 417 and 418. The whole rock samples fall into two groups: one is depleted in MgO and distinctly enriched in plagioclase but has lost some olivine and/or pyroxene relative to its corresponding glass; and the other is enriched in MgO and in phenocrysts of olivine and pyroxene as well as plagioclase compared to its corresponding glass. By analogy with observed phenocryst distributions in lava pillows, tubes, and dikes, and with some theoretical studies, we infer that bulk rock compositions are strongly affected by phenocryst redistribution due to gravity settling, flotation, and dynamic sorting after eruption, although specific models are not well constrained by the one-dimensional geometry of drill core. Compositional trends or groupings in whole rock data resulting from such late-stage processes should not be confused with more fundamental compositional effects produced in deep chambers or during partial melting.