News & Views | Published:


The noble art of recycling

Naturevolume 441pages169170 (2006) | Download Citation


Xenon trapped beneath Earth's crust provides clues to how our planet evolved, but quantifying atmospheric contamination has been impossible. The latest analysis surmounts a barrier to our understanding.

Because of their scarcity, chemical stability and presence as many distinguishable isotopes, noble gases in Earth's mantle — the solid layer between its outer crust and liquid core — provide constraints on the origin, structure and evolution of Earth and its atmosphere. On page 186 of this issue1, Holland and Ballentine use investigations of well gases from the upper mantle to challenge an established tenet concerning noble-gas abundance: the existence of a ‘subduction barrier’ that prevents the noble gases from recycling into the mantle through tectonic activity. In fact, the authors conclude that some 80% of xenon in the mantle comes from sea water introduced by just such a process.

The subduction barrier was described in a classic paper2 showing that at subduction zones — where one of Earth's tectonic plates dives under another — the noble gases in subducting materials are returned to the surface through volcanic activity, leaving the composition of the mantle unaffected. The fact that noble gases of the same isotopic composition as Earth's atmosphere are found in nearly all analyses of samples of rock extruded from the mantle has been attributed, quite reasonably, to contamination from the atmosphere during, or after, the eruption of magma3,4. With the exception of certain mantle-derived rocks that acquire fluid at subduction-zone settings5,6, the recycling of noble gases into the mantle reservoirs by subduction has never been proved.

Holland and Ballentine1 analyse high-quality isotope data from well gases in New Mexico to demonstrate that xenon of atmospheric nature is intrinsic to the mantle. They study the abundances of three primordial isotopes of xenon — 124Xe, 126Xe, 128Xe, which are not produced in any reaction inside Earth — in relation to that of a fourth primordial isotope, 130Xe, and compare the ratios obtained with those found in xenon in the Sun and in Earth's atmosphere (Fig. 1a). Their results are consistent with a xenon content of 90% atmospheric and 10% solar origin.

Figure 1: Differing isotopic ratios of xenon.
Figure 1

a, The ratios 124Xe/130Xe and 126Xe/130Xe in the Sun and in Earth's atmosphere are well known (large circles). Holland and Ballentine's data1 from gas wells in New Mexico (blue squares) fit a straight line drawn between these two points, indicating samples with variable components from solar- and air-like sources. The fraction, f, of air 130Xe in each sample can be estimated from these data by solving a mixing equation (scale bar). Part of the required air-like xenon (red on the scale bar) cannot be accounted for by local air contamination. b, The ratios 129Xe/130Xe and 136Xe/130Xe of New Mexico well gas (blue squares) and MORB rock (grey circles)9,12 plot on different mixing lines: whereas MORB data extend from air composition, suggesting that they contain purely atmospheric contamination, data from New Mexico well gases extend from a point consistent with pre-mixed air–crust gases, which are enriched in 136Xe from crustal uranium fission. The intersection of the two lines defines the isotopic ratios for the mutual source of the two samples — most probably the convecting upper mantle.

The authors address the origin of the air-like xenon component using the ratios of two further isotopes, 129Xe and 136Xe, to 130Xe. These are radiogenic isotopes, respectively produced by radioactive decay from a now-extinct iodine isotope (129I), and by fission of a now-extinct plutonium isotope (244Pu) and long-lived uranium-238. Non-atmospheric 129Xe/130Xe and 136Xe/130Xe ratios have already been established in data from extruded mantle rocks known as mid-ocean-ridge basalts (MORBs). These higher ratios can probably be explained by assuming that significant amounts of primordial 130Xe were removed from the mantle source by early catastrophic degassing. Quantifying the ‘pristine’ mantle composition will therefore constrain models of Earth's evolution from its early stages to the development of distinct mantle reservoirs7. But the pristine composition cannot be identified from MORB data alone, as the degree of atmospheric contamination after the rocks had been extruded is unknown.

The authors1 find that, on a three-isotope diagram, 129Xe/130Xe and 136Xe/130Xe ratios of the New Mexico well gases form a straight line that does not extend from the point that represents those ratios in air (Fig. 1b). Instead, the data lie on a line that extends from a point consistent with a hybrid of atmospheric xenon and xenon of crustal origin. The point of intersection of this line with that of the MORB data will represent the xenon composition of the common source of MORBs and well gases, most probably the upper mantle.

The results from the radiogenic isotopes indicate that the air–crust hybrid component, which is likely to be added to well gases through local contamination during their storage in the continental crust1,8, can account for only about half of the 90% atmospheric xenon component revealed by the non-radiogenic xenon isotopes. The rest of this air xenon must be intrinsic to the mantle reservoir, and the most likely process to introduce it into the mantle source region would be subduction.

Thus it seems that the noble-gas subduction barrier is not as effective as had been presumed. The survival of only 2% of pore water in subducting oceanic plates would be sufficient to account for the quantity of isotopically air-like noble gases found in the upper mantle1. But can these gases be recycled to the deeper mantle regions? Rocks known as oceanic island basalts (OIBs), which are presumed to be of lower-mantle origin, yield noble-gas isotopic ratios that are less radiogenic (and thus more air-like) than MORBs. The difference is normally explained by assuming that MORBs originate from a degassed reservoir, from which the primordial gas has largely been removed to form the present atmosphere.

Holland and Ballentine speculate, however, that the difference in noble-gas isotope signatures between the MORBs and OIBs might in fact result from different degrees of regassing by subduction. Such a model requires that noble gases in the OIB source should be more significantly modified by the recycled component than are those in the MORB source, but this is quite consistent with results from other geochemical tracers such as neodymium, strontium and lead isotopes. Recent attempts to determine precisely the ratios of primordial xenon isotopes in OIBs and MORBs could also support the idea that subduction adds an intrinsic air component to the source regions of these rocks9,10,11.

Holland and Ballentine explore the details of this air influx by showing that the recycled air found in the New Mexico well gases has relative noble-gas abundance ratios similar to those of sea water. But although sea water is undoubtedly a ready source of air-like noble gases in subducting materials, it is unclear how the seawater signature could survive subduction, storage in the mantle or crustal reservoirs and emanation at the continental gas field, without changes to its relative noble-gas abundances. Even the least-contaminated MORBs do not9 contain solar-like xenon, so the underlying assumption that MORBs and the New Mexico samples share the same uniform mantle reservoir needs further clarification. There are also many unresolved issues about the feasibility of deep air recycling — for example, the presence of a carrier in subducting materials other than pore water that can relay air-like noble gas into the deeper mantle regions.

Nevertheless, supported by an independent determination of xenon isotopic composition in the upper mantle12, Holland and Ballentine's conclusion1 that some air-like xenon in the New Mexico well-gas samples is intrinsic to their mantle source seems robust. However, extending the recycling hypothesis to other samples, such as MORBs and OIBs, with their uncertain air contamination, will require further work.


  1. 1

    Holland, G. & Ballentine, C. J. Nature 441, 186–191 (2006).

  2. 2

    Staudacher, T. & Allègre, C. J. Earth Planet. Sci. Lett. 89, 173–183 (1988).

  3. 3

    Patterson, D. B., Honda, M. & McDougall, I. Geophys. Res. Lett. 17, 705–708 (1990).

  4. 4

    Ballentine, C. J. & Barfod, D. N. Earth Planet. Sci. Lett. 180, 39–48 (2000).

  5. 5

    Matsumoto, T., Chen, Y. & Matsuda, J. Earth Planet. Sci. Lett. 185, 35–47 (2001).

  6. 6

    Matsumoto, T. et al. Earth Planet. Sci. Lett. 238, 130–145 (2005).

  7. 7

    Trieloff, M. & Kunz, J. Phys. Earth Planet. Inter. 148, 13–38 (2005).

  8. 8

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

  9. 9

    Kunz, J., Staudacher, T. & Allègre, C. J. Science 280, 877–880 (1998).

  10. 10

    Trieloff, M., Kunz, J. & Allègre, C. J. Earth Planet. Sci. Lett. 200, 297–313 (2002).

  11. 11

    Trieloff, M. et al. Science 288, 1036–1038 (2000).

  12. 12

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

Download references

Author information


  1. the Department of Earth and Space Science, Graduate School of Science, Osaka University, Toyonaka, 560-0043, Japan

    • Takuya Matsumoto


  1. Search for Takuya Matsumoto in:

About this article

Publication history


Issue Date



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.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing