New analysis of Apollo samples find that the Moon's interior may contain 100 times more water than earlier measurements suggested, comparable to the water content of magma in Earth's upper mantle, which has implications for the giant impact theory that is currently accepted to describe its formation.

Seven melt inclusions were sampled from the Apollo 17 rock (sample 74220, and the site of the famous "orange soil"), which contained volcanic glasses erupted 3.6 billion years ago, thus providing a window into the Moon's past. Melt inclusions are different from typical magma or lava samples because they are extremely small – the samples in this particular study were at most 30 micrometres across, smaller than the width of a human hair – and are totally encased within solid crystals when they are trapped within the magma inside the Moon. Probing the water content of pre-eruption magmas in this way is a well-used technique on Earth, too.

"The boiling point of water is at a much lower temperature than magma temperatures, so when a lava erupts, all the water is lost as a gas phase (steam) that escapes into the atmosphere," explains Carnegie Institute for Science's Erik Hauri, who lead the research. "But because melt inclusions are trapped within solid crystals (usually a mineral called olivine), the water cannot escape the magma during eruption, so it remains trapped and we can measure it using a micro-analysis instrument called NanoSIMS. We expected the water contents to be higher, but it was somewhat surprising that we found water contents 100 times higher than in the volcanic glasses from the same sample that we reported in our 2008 paper."

The melt inclusions were originally found by a freshman undergraduate student at Brown University, Tom Weinreich who is the second author on the paper published by Science Express this week.

The previously reported low abundance of water and volatile elements was used in strong favour of giant impact model for the Moon's formation, in which a Mars-size object collides with the Earth, vaporizing away the water from the debris that eventually coalesced to make the Moon. But the impact theory could still work because other observations – such as the angular momentum of Earth-Moon system, evidence for the presence of a lunar magma ocean during the Moon's formation, and the probabilities of giant impact collisions in planet formation models – almost require an impact to account for them.

"But the giant impact does not predict such a high lunar water content, because the impact is thought to result in almost total melting of the material that goes into orbit around the Earth post-impact, and such hot material in the vacuum of space will result in total dehydration," Hauri tells Astronomy Now. "There are two ways out of this. One is that we might be over-estimating the amount of energy in a giant impact, and perhaps some part of the Earth was incorporated into the Moon in a solid unmelted state, and that's why we observe such high water contents and similarly Earth-like amounts of sulfur, fluorine and chlorine. The other is that we might be under-estimating giant impact energy, and maybe the material that was ejected was so hot that it formed an atmosphere of silicate vapor that enclosed both the Earth and proto-Moon, which would be very dense and might be enough to trap water and other volatiles for a long enough period of time that the Earth exchanged volatile material with the proto-Moon."

Despite the uncertainty Hauri thinks that their findings can at least rule out some assumptions about the Earth getting much of its oceans and atmosphere from water-rich comets that struck our planet after the impact. "This material would of course also hit the Moon, but the Earth, by virtue of its larger diameter and higher gravity, would incorporate 20-40 times more of this stuff than the Moon. So this idea would predict a Moon with 20-40 times lower water than the Earth and not the same amount, as we have estimated," he says.

The study also puts a new twist on the origin of water ice detected in craters at the lunar poles by several recent NASA missions, and which have also been attributed to delivery by comets and meteoroid impacts. Hauri notes that the Cabeus crater, the focus of the LCROSS mission that impacted a probe into the crater, resulted in an estimate of that crater alone containing one billion gallons of water – about 1,500 Olympic swimming pools worth.

"In contrast, if the entire Moon has as much water as the source of this volcanic sample we studied, the bulk Moon could have contained one billion times more than this," says Hauri. "Even if only one percent of this water were released by lunar volcanism and trapped on the lunar surface, it would still be tens of millions of Cabeus craters-worth of water. Lunar volcanic eruptions could have produced much more water than delivery of comets or asteroids to the Moon's surface."

The type of sample studied to extract the melt inclusions is an explosive volcanic deposit, which have also been mapped on Mars, Venus and Jupiter's volcanic moon Io, and are now also being mapped by the MESSENGER mission orbiting Mercury.  "It is no longer a mystery where these deposits can be found, we know where to go to recover these samples from other planets," says Hauri. "These samples are the only way to accurately estimate how much water any planet might have within, and how much water can potentially be released to form ice deposits or atmospheres or oceans. Water plays a critical role in determining the tectonic behavior of planetary surfaces, the melting point of planetary interiors, and the location and eruptive style of planetary volcanoes. We cannot conceive of a more important type of sample to return to Earth than these explosive volcanic deposits, and we think they should be a top priority of future sample return missions."