It would be useful to have a better handle on how and when water appeared on the early Earth. We know that comets and asteroids can bring water from beyond the ‘snowline,’ that zone demarcated by temperatures beyond which volatiles like water, ammonia or carbon dioxide are cold enough to condense into ice grains. For our Solar System, that distance in our era is 5 AU, roughly the orbital distance of Jupiter, although the snowline would have been somewhat closer to the Sun during the period of planet formation. So we have a mechanism to bring ices into the inner Solar System but don’t know just how large a role incoming ices played in Earth’s development.

Knowing more about the emergence of volatiles on Earth would help us frame what we see in other stellar systems, as we evaluate whether or not a given planet may be habitable. Usefully, there are ways to study our planet’s formation that can drill down to its accretion from the materials in the original circumstellar disk. A new study from Caltech goes to work on the magmas that emerge from the planetary interior, finding that water could only have arrived later in the history of Earth’s formation.

Published in Science Advances, the paper involves an international team working in laboratories at Caltech as well as the University of the Chinese Academy of Sciences, with Caltech grad student Weiyi Liu as first author. When I think about studying magma, zircon comes first to mind. It appears in crystalline form as magma cools and solidifies. I’m no geologist, but I’m told that the chemistry of melt inclusions can identify factors such as volatile content and broader chemical composition of the original magma itself. Feldspar crystals are likewise useful, and the isotopic analysis of a variety of rocks and minerals can tell us much about their origin.

So it’s no surprise to learn that the Caltech paper uses isotopes, in this case the changing ratio of isotopes of xenon (Xe) as found in mid-ocean ridge basalt vs. ocean island basalt. Specifically, 129Xe* comes from the radioactive decay of the extinct volatile 129I, whose half-life is 15.7 million years, while 136Xe*Pu comes from the extinction of 244Pu, with a halflife of 80 million years. So the 129Xe*/136Xe*Pu ratio is a useful tool. As the paper notes, this ratio:

…evolves as a function of both time and reservoirs compositions (i.e., I/Pu ratio) early in Earth’s history. Hence, the study of the 129Xe*/136Xe*Pu in silicate reservoirs of Earth has the potential to place strong constraints on Earth’s accretion and evolution.

The ocean island basalt samples, originating as far down as the core/mantle boundary, reveal this ratio to be low by a factor of 2.8 as compared to mid-ocean ridge basalts, which have their origin in the upper mantle. Using computationally intensive simulations drawing on what is known as first-principles molecular dynamics (FPMD), the authors find that the low I/Pu levels were established in the first 80 to 100 million years of the Solar System (thus before 129I extinction), and have been preserved for the past 4.45 billion years. Their calculations assess the I/Pu findings under different accretion scenarios, drawing on simulated magmas from the lower mantle, which runs from 680 kilometers below the surface, to the core-mantle boundary (2,900 kilometers), and also from the upper mantle beginning at 15 kilometers and extending downward to 680 kilometers.

The result: The lower mantle reveals an early Earth composed primarily of dry, rocky materials, with a distinct lack of volatiles, with the later-forming upper mantle numbers showing three times the amount of volatiles found below. The volatiles essential for life seem to have emerged only within the last 15 percent, and perhaps less, of Earth’s formation. In the caption below, the italics are mine.

Image: This is Figure 4 from the paper. Caption: Schematic representation of the heterogeneous accretion history of Earth that is consistent with the more siderophile behavior of I and Pu at high P-T [pressure-temperature] conditions (this work). As core formation alone does not result in I/Pu fractionations sufficient to explain the ~3 times lower 129Xe*/136Xe*Pu ratio observed in OIBs [ocean island basalt] compared to MORBs [mid-ocean ridge basalt], a scenario of heterogeneous accretion has to be invoked in which volatile-depleted differentiated planetesimals constitute the main building blocks of Earth for most of its accretion history (phase 1), before addition of, comparatively, volatile-rich undifferentiated materials (chondrite and possibly comet) during the last stages of accretion (phase 2).Isolation and preservation, at the CMB [core mantle boundary], of a small portion of the proto-Earth’s mantle before addition of volatile-rich material would explain the lower I/Pu ratio of plume mantle, while the mantle involved in the last stages of the accretion would have higher, MORB-like, I/Pu ratios. Because the low I/Pu mantle would also have an inherently lower Mg/Si, its higher viscosity could help to be preserved at the CMB until today. Credit: Liu et al.

We’re a long way from knowing in just what proportions Earth’s water has derived from incoming materials from beyond the snowline. But we’re making progress:

…our model sheds light on the origin of Earth’s water, as it requires that chondrites represent the main material delivered to Earth in the last 1 to 15% of its accretion. Independent constraints from Mo [molybdenum] nucleosynthetic anomalies require these late accreted materials to come from the carbonaceous supergroup. Together, these results indicate that carbonaceous chondrites [the most primitive class of meoteorites, containing a high proportion of carbon along with water and minerals] must have represented a non-negligible fraction of the volatile-enriched materials in phase 2 and, thus, play a substantial role in the water delivery to Earth.

All this from the observation that mid-ocean ridge basalts had roughly three times higher iodine/plutonium ratios (inferred from xenon isotopes) as compared to ocean island basalts. The key to this paper, though, is the demonstration that the ratio difference is likely from a history of accretion that began with dry planetesimals followed by a secondary accretion phase driven by infalling materials rich in volatiles.

Thus Earth presents us with a model of planet formation from dry, rocky materials, one that presumably would apply to other terrestrial worlds, though we’d like to know more. To push the inquiry forward, Caltech’s Francois Tissot, a co-author on the paper, advocates looking at rocky worlds within our own Solar System:

“Space exploration to the outer planets is really important because a water world is probably the best place to look for extraterrestrial life. But the inner solar system shouldn’t be forgotten. There hasn’t been a mission that’s touched Venus’ surface for nearly 40 years, and there has never been a mission to the surface of Mercury. We need to be able to study those worlds to better understand how terrestrial planets such as Earth formed.”

And indeed, to better measure the impact of ices brought from far beyond the snowline to the infant worlds of the inner system. Tissot’s work demonstrates how deeply we are now delving into the transition between planetary nebulae and fully formed planets. working across the entire spectrum of what he calls ‘geochemical problematics,’ which includes studying the isotopic makeup of meteorites and their inclusions, the reconstruction of the earliest redox conditions in the Earth’s ocean and atmosphere, and the analysis of isotopes to investigate ancient magmas. At Caltech, he has created the Isotoparium, a state-of-the-art facility for high-precision isotope studies.

That we are now probing our planet’s very accretion is likely not news to many of my readers, but it stuns me as another example of extraordinary methodologies driving theory forward through simulation and laboratory work. And as we don’t often consider work on the geological front in these pages, it seems a good time to point this out.

The paper is Weiyi Liu et al., “I/Pu reveals Earth mainly accreted from volatile-poor differentiated planetesimals,” Science Advances Vol. 9, No. 27 (5 July 2023) (full text).