ABSTRACT

We have carried out noble gas analysis for gases released by a high-resolution stepwise crushing on basaltic glass and olivine separated from a new set of pillow basalts collected form the Loihi seamount, Hawaii. One basaltic glass sample yielded quite consistent ³He/⁴He ratios of about 35 Ra (Ra denotes the atmospheric ratio) for helium released over six successive crushing steps. This ratio is the highest among those so far reported for Loihi samples, suggesting that this sample represents the most primitive Loihi magma in terms of helium isotope composition. In contrast, the argon isotope ratios (⁴⁰Ar/³⁶Ar) observed in this sample appeared to be low (<400) compared with the ratio suggested for the source of Loihi magma (i.e., 8000 by Trieloff et al., 2000), implying a significant contribution from the atmospheric argon. A clear binary mixing trend was defined between ³He/³⁶Ar and ⁴⁰Ar/³⁶Ar ratios measured during successive crushing steps, which is extending from an atmospheric component to a mantle-derived component. This mixing trend permits us to estimate ⁴⁰Ar/³⁶Ar ratios of the mantle by assuming the mantle ³He/³⁶Ar ratio of 0.7 (Moreira et al., 1998). For the Loihi basaltic glass sample with 35 Ra, the ⁴⁰Ar/³⁶Ar ratio without atmospheric contribution is estimated to be 3000, whereas the ratio estimated for the olivine separates with ³He/⁴He ratio of 27 Ra appeared to be 11000. We found that ³He/⁴He and ⁴⁰Ar/³⁶Ar (corrected for air addition) ratios determined for the present samples, as well as for other plumes (Reunion and Iceland) and MORB-popping rock, plot on a rough negative trend, suggesting that helium and argon isotope ratios in plume-derived samples might be controlled by a variable contribution from the MORB-type component to the plume endmember. Extrapolation of this trend to ³He/⁴He of 50 Ra as found in the Baffin Island picrite (Stuart et al., 2003) yielded the plume source ⁴⁰Ar/³⁶Ar ratio of about 1000, suggesting that the source region of the high ³He/⁴He mantle plume might have ⁴⁰Ar/³⁶Ar ratio being significantly less radiogenic than has usually been anticipated.