News & Views | Published:

Solar system

When the dust unsettles

Two attempts to measure the isotopic composition of oxygen in the Sun from particles trapped in lunar soils give very different results. A rethink of why the Solar System is as it is might be required.

On page 776 of this issue, Ireland et al.1 report investigations of lunar soil from which they infer that, compared with other Solar System bodies, the Sun is depleted in the naturally most plentiful oxygen isotope, 16O. Just a year ago, another study2 of soils on the Moon concluded exactly the reverse. So who is right?

Oxygen is the third most abundant element in the Solar System (the Sun's huge stores of hydrogen and helium claim first and second place), and a principal constituent of the rocks, ices and atmospheres that make up the planets. Its three naturally occurring isotopes — 16O, 17O and 18O — are found in relative abundances of about 2,700:1:5, although the physical and chemical processes occurring in different environments of the Solar System can cause shifts in these abundances of up to a few per cent.

Such deviations can be depicted on an oxygen three-isotope plot (Fig. 1, overleaf). Here, all samples from Earth fall along a single line with a slope of about 0.5, indicating that the ratio 18O/16O shifts by twice as much as the ratio 17O/16O. This is mass-dependent behaviour, as the difference in mass between 18O and 16O is twice that between 17O and 16O. Oxygen from other planets or asteroids — represented on Earth by different classes of meteorite — show similar mass-dependent fractionations, but as the underlying oxygen composition of each body is different, the data fall on different lines on the three-isotope plot. Oxygen isotopic composition has therefore become a crucial parameter in classifying meteorites.

Figure 1: Oxygen in the Solar System.
figure1

The quantities δ17O and δ18O give deviations in the ratios 17O/16O and 18O/16O relative to the same ratios in standard mean ocean water (SMOW), with δ17O=[(17O/16O)sample/(17O/16O)SMOW−1]×1,000, and similarly for δ18O. a, Solar System material lies within the small box at the intersection of two lines: the terrestrial mass fractionation line, TFL, with slope 0.5 and along which all samples from Earth lie; and the 16O fractionation line, 16OFL, with gradient 1. Samples with enhanced or depleted 16O content compared with the SMOW value lie farther up or down the 16OFL line, respectively. The star represents the ratios found in Ca–Al-rich inclusions (CAIs) from meteorites. Hashizume and Chaussidon's results from lunar soils2 lie within the grey area; Ireland and colleagues' new findings1 are in the red disc. b, An expansion of the intersect area, showing the lines on which the oxygen isotopic compositions for samples from Mars, the Moon and various meteorite parent bodies sit. The line CCAM refers to an array of measurements of minerals in CAIs.

How this oxygen-isotope variability arose, however, is not understood. Does it represent a heterogeneity inherited from the raw materials that made up the Solar System? Or is it the result of physical or chemical processes in the early Solar System3,4,5? The initial oxygen isotopic composition of the dust and gas from which our Solar System formed is not known. The Sun contains most of the matter in the Solar System, so its oxygen isotopic composition effectively defines the Solar System's oxygen composition. Models that generate the compositions of other bodies from an 16O-rich composition are very different from those that start with an 16O-poor composition. But measuring the chemical and isotopic composition of the Sun directly is difficult; the best data come from measurements of the stream of charged particles emitted by the Sun known as the solar wind.

One of the best places to measure the solar wind is in lunar soils. The Moon has no atmosphere and no magnetic field, so solar-wind particles are implanted directly into its surface. Measurements of matter from the solar wind have so far concentrated on elements such as the noble gases, nitrogen and carbon, which are not themselves significant components of minerals in lunar soil. Oxygen, by contrast, is found in many minerals. Therefore, solar-wind oxygen in lunar soils must be studied in minerals such as iron metal, which are by nature oxygen-free.

Hashizume and Chaussidon2 were the first to separate iron-metal particles from lunar soils and measure their oxygen content using an ion microprobe. They found that oxygen in a few grains of their soil, which had been exposed to the solar wind between one billion and two billion years ago, was enriched in 16O compared with the oxygen on Earth and that from most other Solar System bodies (Fig. 1). This implied that the Sun itself is 16O-rich, just like the calcium–aluminium-rich inclusions (CAIs) that are found in meteorites and are believed to be among the oldest solid bodies in the Solar System. Now, Ireland et al.1 report results from a contemporary lunar soil that support the opposite conclusion.

The main difference between the two studies is the choice of sample. Hashizume and Chaussidon2 used a sample of ancient lunar regolith that they had previously shown contained carbon enriched in 13C (ref. 6) and nitrogen depleted in 15N (ref. 7), results they ascribed to the effect of the solar wind. Their metal grains contained a thick oxide layer that compromised the normal solar-wind implantation profile, but the authors identified 16O-rich oxygen they found inside the grains as so-called solar energetic particles. These particles travel at much higher speeds than normal solar-wind particles, and are therefore much more deeply embedded in the grains.

Ireland and colleagues1 used a sample of lunar soil only recently exposed to the solar wind, but which had one of the highest exposures known. Their metal grains had only a very thin oxide layer, and the 16O-poor oxygen was found at a depth consistent with normal implanted solar wind. The two studies1,2 thus measured solar particles of different ages and, apparently, from different energy regimes. Yet even taking these facts into account, such divergent results are hard to understand.

Processes within the Sun, and those that accelerate solar-wind and solar energetic particles away from its surface, would be expected to affect isotopic composition in a mass-dependent manner. Thus, samples would simply move up or down a line of slope 0.5 on the three-isotope plot. To move away from such a line, either oxygen of a different composition must be added or a process that is mass independent must be invoked. Although the outer layer of the Sun could have changed through the addition of new material over two billion years, there is no known reservoir of material in the Solar System that has a composition extreme enough or a mass great enough to shift the composition of this layer by the required amount.

A process called self-shielding can also alter oxygen isotopic composition. When certain compounds such as carbon monoxide are exposed to intense ultraviolet radiation, molecules containing the highly abundant 16O can exhaust the supply of photons with the correct energy to break them up, although there are still plenty of photons that can disrupt molecules containing 17O and 18O. A reservoir of oxygen with a composition different from the starting carbon monoxide can be created if the released oxygen becomes trapped in a different molecule, such as water5. But there is no obvious way to apply this mechanism to solar-wind or solar energetic particles.

We are therefore left with an intriguing dilemma. The data both of Hashizume and Chaussidon2 and of Ireland and colleagues1 are good. Their results cannot, however, be explained within our current understanding of oxygen isotopes and the structure of the Sun. More data should help: a major goal of NASA's Genesis mission, launched in 2001 to capture the solar wind, was to determine the oxygen isotopic composition of the Sun. But the difficulty inherent in measuring oxygen in the Genesis collectors, and the complications introduced when the spacecraft crashed into the Utah desert on its return to Earth in 2004, mean that it will be some time before we have further data. Depending on the results, some new ideas might be required.

References

  1. 1

    Ireland, T. R., Holden, P., Norman, M. D. & Clark, J. Nature 440, 776–778 (2006).

  2. 2

    Hashizume, K. & Chaussidon, M. Nature 434, 619–622 (2005).

  3. 3

    Clayton, R. N., Grossman, L. & Mayeda, T. K. Science 182, 485–488 (1973).

  4. 4

    Thiemens, M. H. & Heidenreich, J. E. III Science 219, 1073–1075 (1983).

  5. 5

    Clayton, R. N. Nature 415, 460–461 (2002).

  6. 6

    Hashizume, K. et al. Astrophys. J. 600, 480–484 (2004).

  7. 7

    Hashizume, K., Marty, B. & Wieler, R. Earth Planet. Sci. Lett. 202, 201–216 (2002).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading

Comments

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.