The Ontong Java Plateau: Evidence of Sulfide Involvement Using Platinum Group Elements.

Ontong Java Plateau (OJP) basalts in the SW Pacific have been produced by large degrees of partial melting (Mahoney and Spencer, 1991, EPSL, 104:196-210; Neal et al., 1997, AGU Mono. 100:183-216), consistent with decompression melting of a surfacing plume head originating at the core-mantle boundary (CMB) (Coffin and Eldholm, 1993, Sci. Am., 269:42-49). Due to significant fractional crystallization (up to 50%, Neal et al., ibid.) and the resulting low Os abundances of the OJP basalts, Os isotope techniques used to demonstrate a CMB origin for other plumes (Walker et al., 1995, Science, 269:819-822 and 1997, GCA, 61:3145-3160) cannot be easily applied to the OJP. Geophysical modeling suggests chemical interactions across the CMB would occur (~3-6%) (e.g., Kellogg and King, 1993, GRL, 20:379-382; Boehler et al., 1995, Chem. Geol., 120:199-205), which would produce enriched siderophile signatures in the resulting lower mantle-core mixture. The high degree of partial melting (>20%)(Michael, 1999, Geochem. Geophys. Geosyst.) supports the view that any platinum group elements (PGEs) as well as sulfides would be partitioned into the OJP plume rather than remaining in the source (Barnes et al., 1985, Chem. Geol. 53, 303-323). PGE abundances (excluding Os) have been analyzed in OJP basalts from the islands of Malaita and Makira. Melting models using spinel peridotite demonstrate the PGE abundances in the OJP basalts cannot be generated from typical upper mantle. Melting models using primitive mantle mixed with a small amount of bulk core material can produce the PGE patterns and abundances present in OJP basalts (Ely and Neal, 1999, EOS, 80:F1103). Adjacent samples from separate flows indicate that magma chamber fractional crystallization had little effect on PGE abundances. PGE abundances are also remarkably uniform across sequences of flows, except for the variable removal of Pd by weathering. Primitive mantle-normalized patterns using PGEs and trace elements demonstrate that there may have been PGE removal by sulfides in some samples. However, any sulfide removal of PGEs argues for even higher PGE abundances in the source, assumed to be a small amount of bulk core material mixed into the OJP plume source.