Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.


Robotic sample return reveals lunar secrets

The wait is over for more news from the Moon1. Three studies in this issue, by Tian et al.2, Hu et al.3 and Li et al.4, together with one in Science by Che et al.5, report data on the lunar samples brought back by China’s robotic Chang’e-5 mission — the first to return samples since the Soviet Union’s Luna 24 mission in 1976. These data shed light on volcanic eruptions that occurred more than one billion years more recently than those known about previously, and provide information on the cause of the volcanism that cannot be obtained from orbit. The results raise questions about the structure and thermal evolution of the lunar interior, and could help to improve methods for estimating the age of planetary surfaces throughout the inner Solar System.

In December 2020, the Chang’e-5 lander set down in the Rümker region near the northwest corner of Oceanus Procellarum on the side of the Moon closest to Earth (Fig. 1). Like the sites visited by Luna and by NASA’s Apollo missions, the Rümker region consists predominantly of a magnesium-rich volcanic rock known as basalt, but the difference from previous missions is that the Rümker basalts are potentially as young as 1.2 billion to 2.3 billion years old, which makes the Chang’e-5 samples the youngest taken from the Moon so far6.

Figure 1

Figure 1 | Mission sites on the Moon. Tian et al.2, Hu et al.3, Li et al.4 and Che et al.5 report analyses of samples returned from the Chang’e-5 mission, which landed in the Rümker region of the Moon, away from the landing sites of NASA’s Apollo missions and the former Soviet Union’s Luna missions. Data taken from orbit are shown here for thorium, a radioactive element that is often used as an indicator of a type of rock known as KREEP. The concentration of thorium is high in a region on the near side of the Moon (left) called the Procellarum KREEP Terrane, and low on the far side of the Moon (right). Although these orbital data imply that the Rümker region is rich in KREEP, samples from Chang’e-5 suggest otherwise. p.p.m., parts per million.Credit: NASA; Adapted from Extended Data Fig. 1 of ref. 3 and Plate 1b of ref. 7

The Chang’e-5 landing site is in an area known as the Procellarum KREEP Terrane7, where KREEP is an acronym for a rock that is rich in potassium (chemical symbol K), the rare-earth elements and phosphorus (chemical symbol P), together with a number of other elements, including the radioactive elements uranium and thorium. All of these elements are called ‘incompatible’ because they do not readily fit into the group of minerals that crystallize from a magma of composition similar to that of the interior of the Moon.

KREEP features prominently in models of lunar evolution that suggest the Moon was initially mostly molten, existing as a lunar magma ocean8. In this scenario, KREEP is thought to have been the last liquid left during the final stages of crystallization of this ocean. Radioisotope studies show that its chemical characteristics formed around 4.4 billion years ago, at roughly the same time that the oldest rocks in the Moon’s crust were produced, as well as the source regions for younger lunar basalt magma9. That implies that this age might date the formation and crystallization of the magma ocean.

The high abundance of radioactive elements in KREEP also suggests that radioactive decay in KREEP is a key source of heat in the lunar interior. Orbital data for thorium concentrations (Fig. 1) show that KREEP is found mainly on the Moon’s Earth-facing side. This high concentration of radioactive heating might explain several differences between the lunar near and far sides. The younger age of volcanoes, different crater shapes10 and thinner crust11 on the near side all reflect the role of high temperatures in prolonging melting in the lunar interior and weakening of the overlying crust.

The Chang’e-5 mission brought back 1.731 kilograms of the lunar surface, consisting of basalt fragments and other surface material welded together by impacts into a rock type known as a breccia. Using a technique called secondary ion mass spectrometry, Li et al. found that the uranium-rich mineral phases in the basalt fragments came from eruptions that occurred between 2,026 million and 2,034 million years ago. Che et al. found that the samples were between 1,906 million and 2,020 million years old. Tian et al. and Che et al. determined that these basalt fragments have titanium concentrations that lie in the middle of the concentration range for the fragments sampled by Apollo. Hu et al. calculated that the concentration of water in the source of the Chang’e-5 basalts is 30 to 140 times lower than for the mantle sources of Earth’s driest lavas12. These low water concentrations are consistent with the fact that the Moon is generally depleted in volatile elements and compounds, which are chemical species that can readily vaporize. If the Moon formed from material ejected during a giant impact with Earth, such species would have been vaporized and lost at the high temperatures involved in this event13.

The concentrations of incompatible elements in the Chang’e-5 basalt samples are comparable to those typically found in KREEP, which is consistent with data from orbiting spacecraft. Tian et al. obtained an unexpected result, however, when they looked at the isotopic composition of strontium and neodymium — the isotopic abundances of these two elements are affected by radioactive decay. Their analysis suggests that the Chang’e-5 basalts derive from the melting of ancient sources in the lunar interior that are depleted in incompatible elements. These elements became enriched in the basalts because the magmas from which they crystallized came from minimal melting of the lunar interior and underwent large amounts of cooling and crystallization during their transit from source to eruption. Both the partial melting and the crystallization enriched the remaining magma in incompatible elements, but neither process affected the isotopic composition of the magma’s source, which is depleted in these elements.

The age and compositional characteristics of the Chang’e-5 samples have at least two implications for our understanding of lunar structure and evolution. First, even though regions that have high concentrations of incompatible elements are seen across most of Oceanus Procellarum (Fig. 1), the Chang’e-5 data show that not all of these are composed of KREEP. This means that KREEP might form a much smaller component of the lunar interior than was suspected from Apollo samples, and data taken from orbit. Second, the melting that led to lunar volcanic eruptions two billion years ago did not involve heating due to high concentrations of radioactive elements in the magma source regions. Nor did it occur as a result of high water concentrations that reduce the melting temperature of rock in the same way that salt reduces the melting temperature of ice. An alternative explanation must now be sought for how the lunar interior was hot enough to drive volcanic eruptions two billion years ago.

The Chang’e-5 results demonstrate that samples returned from previously unvisited regions of the lunar surface can prompt a revision of models of lunar evolution that were developed on the basis of the Apollo and Luna samples. This is not surprising, given that the combined Apollo and Luna missions sampled only a restricted portion of the Moon’s Earth-facing side. Much like Earth’s surface, the lunar surface is a mosaic of materials created over the past 4.5 billion years. Each piece of that mosaic provides different information about the history of the Moon. The Chang’e-5 results show that sample-return missions to previously unexplored portions of the lunar surface will help models of the evolution of our nearest planetary neighbour to eventually converge on reality.

Nature 600, 39-40 (2021)



  1. 1.

    Spudis, P. Nature Geosci. 2, 234–236 (2009).

    Article  Google Scholar 

  2. 2.

    Tian, H.-C. et al. Nature 600, 59–63 (2021).

    Article  Google Scholar 

  3. 3.

    Hu, S. et al. Nature 600, 49–53 (2021).

    Article  Google Scholar 

  4. 4.

    Li, Q.-L. et al. Nature 600, 54–58 (2021).

    Article  Google Scholar 

  5. 5.

    Che, X. et al. Science 374, 887–890 (2021).

    PubMed  Article  Google Scholar 

  6. 6.

    Qian, Y. Q. et al. J. Geophys. Res. Planets 123, 1407–1430 (2018).

    Article  Google Scholar 

  7. 7.

    Jolliff, B. L., Gillis, J. J., Haskin, L. A., Korotev, R. L. & Wieczorek, M. A. J. Geophys. Res. 105, 4197–4216 (2000).

    Article  Google Scholar 

  8. 8.

    Warren, P. H. Annu. Rev. Earth Planet. Sci. 13, 201–240 (1985).

    Article  Google Scholar 

  9. 9.

    Borg, L. E., Gaffney, A. M. & Shearer, C. K. Meteorit. Planet. Sci. 50, 715–732 (2015).

    Article  Google Scholar 

  10. 10.

    Miljković, K. et al. Science 342, 724–726 (2013).

    PubMed  Article  Google Scholar 

  11. 11.

    Laneuville, M., Wieczorek, M. A., Breuer, D. & Tosi, N. J. Geophys. Res. Planets 118, 1435–1452 (2013).

    Article  Google Scholar 

  12. 12.

    Saal, A. E., Hauri, E. H., Langmuir, C. H. & Perfit, M. R. Nature 419, 451–455 (2002).

    PubMed  Article  Google Scholar 

  13. 13.

    Stevenson, D. J. Annu. Rev. Earth Planet. Sci. 15, 271–315 (1987).

    Article  Google Scholar 

Download references

Competing Interests

The author declares no competing interests.


Nature Careers


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

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


Quick links