A dash of deep nebula on the rocks

The cocktail of noble-gas isotopes in an Icelandic rock suggests that the upper mantle does not, and never did, receive gas from a deeper mantle reservoir. This challenges ideas of deep Earth's behaviour and formation. See Letter p.101

The pattern of isotopic abundances of inert and rare noble gases, trapped in small bubbles in volcanic rock, act as a 'fingerprint' of how and where our planet first acquired its gas. Furthermore, the type of volcanic setting, and the way that parts of the fingerprint change with time, offer insight into the workings of deep Earth. Squeezing out this information from lava derived from the deepest parts of our planet — possibly some 2,900 kilometres beneath our feet — is a challenge. Yet this is precisely what S. Mukhopadhyay1 has done, and in spectacular fashion. On page 101 of this issue, he reports the long-awaited detailed secrets of the planet's deepest gases, based on an isotope analysis performed using a new-generation mass spectrometer.

Helium is light enough to be lost from the atmosphere to space, and so its atmospheric concentration is very low. Its isotopic composition in mantle rocks (measured as the ratio of the gas's two isotopes, 4He and 3He) is therefore the easiest to ascertain of all the noble gases, because measurements are not swamped by background 'noise' from contaminating atmospheric helium. The 4He/3He ratios at mid-ocean ridges — the 65,000km of interconnected underwater volcanic systems that spew magma from the uppermost mantle to build new ocean crust — are almost constant around the globe. But lower ratios have been measured in rocks produced by certain 'hotspot' volcanoes, such as those in Hawaii and Iceland, which are thought to tap the deepest mantle.

The existence of different 4He/3He ratios underpins the idea that there are at least two geochemical reservoirs in the mantle2: a deep reservoir rich in gases and volatile compounds feeds material into an upper reservoir, which is the convecting part of the mantle that supplies magma to mid-ocean ridges (Fig. 1). Although ideas about the depth, size and nature of the deepest reservoir have changed substantially, the two-reservoir model has dominated attempts to explain observations of mantle geochemistry for the past 30 years.

Figure 1: Mantle movement.

Magma from the upper part of the convecting mantle erupts at mid-ocean ridges, whereas that from a deep reservoir is thought to erupt at 'hotspot' volcanoes. Subduction processes transfer material from the ocean crust back into the convecting mantle, and possibly also into the deep mantle. The isotopic composition and amount of helium in the upper mantle suggest a flux of 3He through this region13, entering from the deep reservoir and exiting to the oceans and atmosphere. Mukhopadhyay reports1 that the isotopic and elemental compositions of noble gases from an Iceland mantle plume that is thought to originate from the deep mantle are fundamentally different from those in the convecting mantle. If the Iceland plume is representative of the deep-mantle reservoir, this rules out the possibility of a large transfer of noble gases from the deep to the convecting mantle (red crosses), overturning a long-standing model of mantle-gas geochemistry. Yellow arrows, crust movement; pink arrows, gas transport.

Obtaining robust information about the isotopic composition of the heavy noble gases in the mantle (neon, argon, krypton and xenon) has been far harder to do than it was for helium. This is because the atmospheric concentrations of these gases are much higher than that of helium, greatly increasing the background noise caused by air contamination of mantle samples. (No magma can erupt into the ocean or the atmosphere without the resulting basalt rock becoming contaminated by air — it would be like jumping into a swimming pool and expecting to stay dry.) Our detailed understanding of the upper mantle's heavy noble gases has therefore come almost entirely from only two rare samples in which such contamination is minimized: a single gas-rich basalt dredged from the mid-Atlantic ridge3; and volcanic gas trapped in a deep carbon dioxide gas field4,5 in New Mexico.

Even fewer traces of noble gases are found in basalts produced by hotspot volcanism than in those produced at mid-ocean ridges6, making hotspot rocks highly susceptible to air contamination. However, in one area of Iceland, a basalt has been found7 in which more mantle gas is known to have been preserved than in most basalts, in part because it erupted under an ice cap. Mukhopadhyay1 has now re-examined this basalt by applying a technique that allows large samples of the rock to be crushed under vacuum (to protect the isotopic fingerprint of gases in the rock from air contamination), and then analysing the released gases using one of a new generation of mass spectrometers that greatly increases the precision with which isotope ratios are measured5. In this way, he has teased out a veritable cornucopia of fresh information.

The isotopic composition of neon in the basalt suggests that the deep Iceland mantle gases originated from the solar nebula — the cloud of dust and gas from which the planets of the Solar System formed. Such gas was around only for the first few tens of millions of years of the Solar System's formation, after which it was either pulled into the Sun or blasted out of the Solar System when the Sun ignited. By contrast, the noble gases in the upper mantle came from meteorites4,5. Mukhopadhyay's findings, together with those of others, allow us to appreciate the true complexity of gas delivery to our planet from different sources at different stages of Earth's infancy1,4,5,8.

But the questions of how gas from the solar nebula was trapped in the solid parts of growing planets, and how the gas was preserved through early accretionary events, will certainly test our models of accretion. Some of the noble-gas isotopes from the Icelandic deep mantle came from long-dead radioactive isotopes of iodine and plutonium that were present in the early Solar System. Mukhopadhyay1 compared these noble-gas isotopes with those in the convecting mantle9, and concluded that the Iceland deep mantle formed in a drier environment, and preserved a higher proportion of its plutonium-decay gases, than did the convecting mantle. This chimes with the idea of a process or location in the deep mantle that has preserved the earliest geochemical signals of accretion exceptionally well. Mukhopadyay's findings may also help to connect theories of how a planet starts to obtain its gas with evidence10 from other isotope systems that also points to the very early formation of reservoirs hidden in the deep mantle.

Although many geochemists have argued that Earth contains a deep, gas-rich reservoir, they have struggled to pinpoint where it should be. Ever since it became apparent from seismic tomography that Earth's mantle was not nicely layered11, the location or processes that could prevent such a deep reservoir from mixing into the convecting mantle and disappearing completely have remained enigmatic. Wherever this reservoir might be, it has survived the cataclysmic Moon-forming event (in which Earth was struck by a Mars-sized body)12; avoided mixing with volatile compounds brought to Earth by meteorites; and withstood continual removal of material by mantle plumes.

One result from Mukhopadhyay's work is touched on only lightly by the author, but might have the greatest impact on how we think the mantle behaves. If the isotopic composition of the basalt analysed by Mukhopadhyay — and therefore of the Iceland plume from which this hotspot rock is derived — is indeed representative of a deep mantle reservoir, then this reservoir cannot also be the source of 3He needed to explain the 4He/3He ratio in the upper mantle, because the heavy noble gases in the basalt don't match those in the upper mantle. The two-reservoir mantle model must therefore be modified. Mukhopadhyay's data about the cocktail of mantle noble gases, however, will endure.


  1. 1

    Mukhopadhyay, S. Nature 486, 101–104 (2012).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  2. 2

    Kurz, M. D., Jenkins, W. J. & Hart, S. R. Nature 297, 43–47 (1982).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Moreira, M., Kunz, J. & Allegre, C. J. Science 279, 1178–1181 (1998).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Ballentine, C. J., Marty, B., Sherwood Lollar, B. & Cassidy, M. Nature 433, 33–38 (2005).

    ADS  CAS  Article  Google Scholar 

  5. 5

    Holland, G., Cassidy, M. & Ballentine, C. J. Science 326, 1522–1525 (2009).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Gonnermann, H. M. & Mukhopadhyay, S. Nature 449, 1037–1040 (2007).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Harrison, D., Burnard, P. & Turner, G. Earth Planet. Sci. Lett. 171, 199–207 (1999).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Marty, B. Earth Planet. Sci. Lett. 313, 56–66 (2012).

    ADS  Article  Google Scholar 

  9. 9

    Pepin, R. O. & Porcelli, D. Earth Planet. Sci. Lett. 250, 470–485 (2006).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Boyet, M. & Carlson, R. W. Science 309, 576–581 (2005).

    ADS  CAS  Article  Google Scholar 

  11. 11

    van der Hilst, R. D., Widiyantoro, S. & Engdahl, E. R. Nature 386, 578–584 (1997).

    ADS  CAS  Article  Google Scholar 

  12. 12

    Canup, R. M. & Asphaug, E. Nature 412, 708–712 (2001).

    ADS  CAS  Article  Google Scholar 

  13. 13

    Kellogg, L. H. & Wasserburg, G. J. Earth Planet. Sci. Lett. 99, 276–289 (1990).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Chris J. Ballentine.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Ballentine, C. A dash of deep nebula on the rocks. Nature 486, 40–41 (2012).

Download citation


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.