A precision analysis of gases from Earth's mantle collected at a geologic formation in the U.S. Southwest points to a source for the gas that more closely resembles carbonaceous meteorites than it does the sun. If confirmed by further research, the new study would challenge a theoretical model for atmosphere formation in which Earth began with two reservoirs of solar gas captured during the planet's formation and youth—one surrounding the planet, the other buried beneath the surface.

Greg Holland, a postdoctoral researcher in isotope geochemistry at the University of Manchester in England, and his colleagues measured the amounts of various isotopes of noble gases in the Bravo Dome gas field in New Mexico, where magmatic gases—primarily carbon dioxide—that allow the mantle to be sampled are buried hundreds of meters below. (Isotopes are species of the same element, albeit with different numbers of neutrons and hence different atomic masses.) The prevalences and isotope ratios of the noble, or inert, gases, such as neon, argon, krypton and xenon, provide a valuable tracer of ancient processes, because they are chemically nonreactive and so do not change much over time. In the December 11 issue of Science, Holland's group outlines its findings and how they might rule out some theories of the way Earth formed its atmosphere as the planet coalesced some 4.5 billion years ago.

There are many possible sources for the components of Earth's primordial atmosphere, from the so-called solar nebula, a cloud of dust and gas leftover from the sun's formation, to comets and other impactors that may have delivered significant amounts of chemicals to Earth during or after the planet's formation. Similarly, there are myriad ways that planets can lose their atmospheres, through stripping by the solar wind, baking by the sun's radiation or catastrophic impacts by comets or asteroids.

By comparing Earth's present atmosphere with the composition of solar gases, previous researchers have developed a model by which a pair of distinct solar-acquired gas reservoirs would evolve into the present-day atmosphere. "If you look at the atmosphere today, it turns out that it is emphatically nonsolar in composition," says Robert Pepin, an emeritus physics professor at the University of Minnesota who did not contribute to the new research. "And the diagnostic clue is that all of the isotope ratios for neon, argon, krypton, xenon, nitrogen, what have you, look as if they have suffered an escape process in which the lightest isotope escaped preferentially relative to the heavier ones"—a process called fractionation. In other words, the noble gases of the atmosphere are isotopically heavier than those of the sun.

Modeling the turbulent processes—giant impacts and extreme ultraviolet radiation, for instance—that could drive xenon fractionation leads to what Pepin calls a "krypton problem." The same fractionation process that could turn a solar sample of xenon into a modern-day atmospheric sample of xenon would leave a deficit in light isotopes of krypton, compared with what is found in Earth's present atmosphere. But Pepin and his colleagues realized that if another source of solar gas were available—say, one trapped in Earth's mantle from the planet's early days—it could bring the elements back into balance as it outgassed into the atmosphere. "It so happens that if you bring a solar component in and mix it with that fractionated krypton, you get the present atmospheric composition," Pepin explains.

Holland can agree with this theory on its face, but his data fail to support it. Leftover gas from the formation of the sun may have persisted into the era of planetary precursors, Holland and his co-authors note, so the dual capture of solar gas in and around planets is plausible enough. "It's perhaps a natural assumption to say that it's inside the mantle," Holland says. But the trace noble gases inside the Bravo Dome field do not reflect a solar origin. "We don't see any solar noble gas in the mantle," he says.

Krypton has several stable isotopes, the relative abundances of which can be used to parse the gas's source. Solar-derived krypton is isotopically light—it has relatively low ratios of the heavier isotopes krypton 84 and 86 to the lighter krypton 82. Krypton in today's atmosphere is somewhat heavier than solar krypton, and the krypton embedded in meteorites known as carbonaceous chondrites is even heavier than that.

In the samples of mantle gas taken by Holland and his colleagues, the krypton measurements were heavy, producing "something that looks rather like gases that are trapped in primitive chondrites today," Pepin says. The implication is that the gases in the mantle came not from the sun but from the early accretion of rocky material. With such a makeup of mantle gases, Holland says, "you can't outgas the interior to make the atmosphere—the atmosphere has to come from somewhere else." One possibility called out by the authors would be icy comets bombarding Earth after formation and delivering their own distinct cocktail of compounds and gases to an existing atmosphere evolved from a solar source.

Pepin says that the new work challenges his own but calls it "a wonderful idea, backed up by data." But he notes that his model of a planet acquiring and then outgassing solar material is not dead yet. For one thing, he says, Holland and his colleagues need to sample other sources to see if their result is typical of the mantle or simply a quirk of the New Mexico site. What is more, Pepin notes, solar gas once embedded in Earth's mantle may have been exhausted long ago. "They assume that since they don't see solar-type krypton today in this one reservoir, that it was never there," Pepin says. "The Earth was such an active thing in its early history that it could have been there and degassed to the point that you can't see it anymore."

Holland acknowledges the limitations in the new data set: "These samples are just from one specific area of the southwest U.S.A." But testing other mantle sources, he adds, is "our next line of work."