Until about five years ago, the moon was thought to be bone dry. Practically all of volatiles, especially water, were thought to have been lost when the moon formed from debris after the impact of a Mars-size body with the proto Earth, the favored moon formation scenario. Shortly after its formation, the moon (or at least the outer portion) was thought to be molten, forming a lunar magma ocean. As solidification of the magma ocean proceeded, different minerals crystallized from the magma. The minerals denser than the magma (e.g., olivine and pyroxene) sank to form the lunar mantle, while plagioclase, a less dense mineral, floated to form plagioclase-dominated (anorthositic) highlands rocks and the original lunar crust. Such an energetic origin for the moon is consistent with its volatile depleted nature.
However, the anhydrous view of the moon has been challenged by detections of water on the lunar surface and in rocks and soil returned by the Apollo missions. While a significant part of this water is thought to have been brought to the moon through solar wind implantation and the impacts of volatile-rich comets long after the primary lunar crust formed from the cooling magma ocean, the detection of water in volcanic rocks derived from the lunar mantle is both exciting and puzzling at the same time. We have been fortunate that some lunar anorthosites were returned to Earth during the Apollo missions, as we have samples of the primary lunar crust. Plagioclase grains in these anorthosites were analyzed using Fourier transform infrared spectroscopy, and approximately 0.0006 percent water by weight was detected in the crystal. This is a significant amount because it is known that water is not readily included in the plagioclase structure, so the magma it crystallized from would have contained much higher abundances. Therefore, the implication of our results is that the initial water content of the lunar magma ocean would have contained approximately 0.032 percent by weight; water accumulating in the final residuum of the lunar magma ocean could have reached 1.4 percent by weight, an amount sufficient to explain water contents measured in lunar volcanic rocks. While the presence of water in the early moon needs to be reconciled its supposed energetic origin and the isotope results for Chlorine and Zinc (both of which are volatile and support a "dry moon"), the presence of water in rocks from the primary lunar crust demonstrates the early moon contained significant amounts of volatiles.