Valley et al. (2002, Geology, 30, p. 351) suggest that a constant range of values of oxygen isotope ratios in zircons implies somewhat temperate conditions on Earth throughout most of the Archean Eon (4.4-2.6 Ga). They question whether the Giant-Impact model for the origin of the Earth-Moon system is compatible with this information because the putative Earth-shattering impact which results in the formation of the Moon must happen and the Earth must cool to a condition where oceans can form before 4.45 Ga. They then suggest that maybe a planetoid capture model should be considered for the origin of the Earth-Moon system.

We have been working on such a planetoid capture model (Malcuit et al., 1992, Proc. Vol., 3rd Int. Archean Symp., p. 223) and we think that it is compatible with the "Cool Early Earth" model. In our model the capture episode occurs about 3.9 Ga. To explain the dehydrated nature of lunar rocks, we suggest an ultimate birthplace for the body of the Moon, and sibling planetoids, inside the orbit of Mercury. Such a place of origin is compatible with chemical condensation models which feature the condensation of anhydrous Ca-Mg-Al-Ti-rich refractory minerals of about lunar density inside the orbit of Mercury.

In this capture model the Earth would be moonless from the end of the accretion process until the sharply defined capture event at about 3.9 Ga. The Earth, then, could develop a crust with some felsic component (containing zircons) as early as 4.45 Ga. During capture the planetoid undergoes a "major thermal episode" featuring remelting of a large portion of the magma ocean zone of the lunar body. In contrast, the planet absorbs only a few percent of the energy for lunar capture and would remain "cool" during the capture episode and subsequent geocentric orbit circularization era.