Although melting of planetary interiors is ubiquitous, our understanding of melting and its effects on deep mantle chemistry of Earth — the planet we know the most about — remains contentious. One key problem is the so-called missing argon paradox. The isotope ⁴⁰Ar is produced as a radioactive decay product of the potassium isotope ⁴⁰K, whose abundance is estimated by assuming a bulk silicate Earth (BSE) composition. K is preferentially partitioned into melts induced by convective upwelling. K is thus removed from the deep interior and emplaced into the crust and uppermost mantle. ⁴⁰Ar is outgassed to the atmosphere from these layers. However, the combined abundances of ⁴⁰Ar in the atmosphere and the upper mantle/crust underpredict the ⁴⁰Ar budget when compared to the estimated BSE. Conventionally this has been explained either by concluding that the BSE assumption is wrong or by invoking the presence of an isolated deep primordial reservoir that has been unmelted and unmixed since planetary accretion and thus sequesters K. However, a chemically layered mantle is inconsistent with seismological observations and geodynamic evidence of convection.

To address this paradox, Jonathan Tucker (Carnegie/Smithsonian Institutions) and colleagues conducted numerical experiments of mantle convection to test whether the recycling of oceanic crust is a viable explanation to account for the missing argon. They show that subducted oceanic crust rich in ⁴⁰K releases ⁴⁰Ar back into the deep mantle rather than to the surface/atmosphere. Including this reservoir of subducted oceanic crust accounts for the totality of Earth’s ⁴⁰Ar based on BSE. These results do not require a significant deviation from the BSE assumption for Earth’s K budget. Neither do they require the deep mantle to be unmelted and chemically isolated from the surface over Earth’s history. Recycling of oceanic crust may explain both the outgassing and thermal history of Earth and matches key observations from seismology and geodynamics. READ MORE