The CO2-He-Ar-H2O systematics of the Manus back-arc basin: Resolving source composition from degassing and contamination effects

We report new stepped heating He, Ar, CO2 and water data on a petrogenetically diverse suite of lavas from the Manus back-arc basin, where a plume component has previously been identified. The aim of this study is to evaluate the superimposed effects of degassing and contamination in order to identify mantle source characteristics. CO2 abundances and carbon isotopes in both the vesicle ([CO2] up to 180ppm; delta(13)C as low as -33.6 parts per thousand) and glass ([CO2] up to 270ppm; delta(13)C as low as -34.3parts per thousand) phases reveal that samples have been modified by varying degrees of degassing. High water concentration samples (back-arc basin basalts (BABB) and arc type samples) show the highest degrees of degassing (i.e. lower delta(13)C values and lower CO2 contents). The results are modelled for both the glass and vesicle phases using batch and fractional degassing models. Parental melt compositions can be constrained to show the following CO2 concentration trend: are-type > BABB s.r. (southern rift) > MORB-2, E-MORB, X-BABB (extreme BABB), BABB > MORB-1 and MORB-smt. He-4/Ar-40* ratios of samples (14.6-1100) are consistent with residual volatiles from a degassed source. Variations in CO2/(3) He values are likely due to degassing, followed by contamination from a crustal source (either the subducting Solomon Sea Plate or the pre-existing crust through which the lavas erupt), as evidenced by high K2O/TiO2 ratios and low delta(13)C. The CO2/(3) He of the Manus plume is best estimated by the MORB-smt and MORB-1 samples at 3.1 +/- 0.6 X 10(9). This value is similar to previous estimates of plume CO2/(3) He values, which are either equal to or slightly greater than the upper mantle average of 2 X 10(9). Copyright (C) 2004 Elsevier Ltd.