Did the terrestrial planets of the Solar System form by pebble accretion?

The dominant accretion process leading to the formation of the terrestrial planets of the Solar System is a subject of intense scientific debate. Two radically different scenarios have been proposed. The classic scenario starts from a disk of planetesimals which, by mutual collisions, produce a set of Moon to Mars-mass planetary embryos. After the removal of gas from the disk, the embryos experience mutual giant impacts which, together with the accretion of additional planetesimals, lead to the formation of the terrestrial planets on a timescale of tens of millions of years. In the alternative, pebble accretion scenario, the terrestrial planets grow by accreting sunward-drifting mm-cm sized particles from the outer disk. The planets all form within the lifetime of the disk, with the sole exception of Earth, which undergoes a single post-disk giant impact to form the Moon. To distinguish between these two scenarios, we revisit all available constraints: dynamical, chronological, and compositional (chemical and in terms of the nucleosynthetic isotope dichotomy between non-carbonaceous (NC) and carbonaceous chondrite (CC) type materials, representing the inner and outer disk, respectively). We find that, taken together, these constraints argue against a simple model of terrestrial planet formation by pebble accretion. In particular we find that the necessary substantial flux of pebbles from the outer protoplanetary disk necessary in this model is inconsistent with (i) the lack of a clear temporal trend of the isotopic composition of the NC reservoir towards more CC-rich compositions, (ii) the isotopic composition of Earth in the framework of the NC–CC dichotomy of the Solar System, (iii) the close isotopic similarity between Earth and some meteorites (i.e. aubrites and enstatite chondrites), and (iv) the lack of a CC contribution to Mars together with the overall low CC fraction in Earth. Together, these observations indicate that CC dust from the outer disk never penetrated in significant amount into the terrestrial planet-forming inner disk. Thus, if planets accreted from pebbles, they did so from a reservoir of local dust, presumably trapped in a pressure bump of the disk. However, even this scenario faces difficulties. The gradual depletion of the Earth in volatile elements of decreasing condensation temperature appears to be difficult to reconcile with pebble accretion, which should lead to a quasi-stepwise depletion pattern (i.e. full retention/depletion of elements with condensation temperature above/below a certain value). Moreover, the rapid formation of the proto-Earth within the lifetime of the disk, required in the pebble accretion model, would have inevitably led to a much more radiogenic ¹⁸²W composition of Earth’s mantle than observed. A single late giant impact, as is commonly invoked for the origin of the Moon, is insufficient for removing this radiogenic ¹⁸²W to Earth’s core because of the small degree of equilibration between the impactor core and Earth’s mantle expected for giant impacts. We thus conclude that the pebble accretion scenario is unable to match the available dynamical, chronological, and compositional constraints in a self-consistent manner, unlike the classic scenario.