摘要

It has long been customary to assume that in the bulk composition of the Earth, all refractory-lithophile elements (including major oxides Al(2)O(3) and CaO, all of the REE, and the heat-producing elements Th and U) occur in chondritic, bulk solar system, proportion to one another. Recently, however, Nd-isotopic studies (most notably Boyet M. and Carlson R. W. (2006) A new geochemical model for the Earth's mantle inferred from (146)Sm-(142)Nd systematics. Earth Planet. Sci. Lett. 250, 254-268) have suggested that at least the outer portion of the planet features a Nd/Sm ratio depleted to similar to 0.93 times the chondritic ratio. The primary reaction to this type of evidence has been to invoke a "hidden" reservoir of enriched matter, sequestered into the deepest mantle as a consequence of primordial differentiation. I propose a hypothesis that potentially explains the evidence for Nd/Sm depletion in a very different way. Among the handful of major types of differentiated asteroidal meteorites, two (ureilites and aubrites) are ultramafic restites so consistently devoid of plagioclase that meteoriticists were once mystified as to how all the complementary plagioclase-rich matter (basalt) was lost. The explanation appears to be basalt loss by graphite-fueled explosive volcanism on roughly 100-km sized planetesimals; with the dispersiveness of the process dramatically enhanced, relative to terrestrial experience, because the pyroclastic gases expand into vacuous space (Wilson L. and Keil K. (1991) Consequences of explosive eruptions on small Solar System bodies: the case of the missing basalts on the aubrite parent body. Earth Planet. Sci. Lett. 104, 505-512). By analogy with lunar pyroclastic products, the typical size of pyroclastic melt/glass droplets under these circumstances will be roughly 0.1 mm. Once separated from an asteroidal or planetesimal gravitational field, droplets of this size will generally spiral toward the Sun, rather than reaccrete, because drag forces such the Poynting-Robertson effect quickly modify their orbits (the semimajor axis, in a typical scenario, is reduced by several hundred km during the first trip around the Sun). Assuming a similar process occurred on many of the Earth's precursor planetesimals while they were still roughly 100 km in diameter, the net effect would be a depleted composition for the final Earth. I have modeled the process of trace-element depletion in the planetesimal mantles, assuming the partial melting was nonmodal and either batch or dynamic in terms of the melt-removal style. Assuming the process is moderately efficient, typical final-Earth Nd/Sm ratios are 0.93-0.96 times chondritic. Depletion is enhanced by a relatively low assumed residual porosity in batch-melting scenarios, but dampened by a relatively high value for "continuous" residue porosity in dynamic melting scenarios. Pigeonite in the source matter has a dampening effect on depletion. There are important side effects to the Nd/Sm depletion. The heat-producing elements, Th, U and K, might be severely depleted. The Eu/Eu* ratio of the planet is unlikely to remain precisely chondritic. One of the most inevitable side effects, depletion of the Al/Ca ratio, is consistent with an otherwise puzzling aspect of the composition of the upper mantle. A perfectly undepleted composition for the bulk Earth is dubious.

  • 出版日期2008-4-15