Signs that Jupiter was mixed by a giant impact

Simulations suggest that Jupiter’s dilute core might be the result of a collision between the planet and a Uranus-mass planetary embryo. This finding indicates that giant impacts could be common during planet formation.
Tristan Guillot is in the Université Côte d’Azur, Laboratoire Lagrange, Observatoire de la Côte d’Azur, CNRS UMR 7293, 06304 Nice Cedex 4, France.

Search for this author in:

In the past couple of years, NASA’s Juno spacecraft has measured Jupiter’s gravitational field with exquisite accuracy1,2. The results have revealed that the planet’s fluid hydrogen–helium envelope does not have a uniform composition: the inner part contains more heavy elements than the outer part3,4. Writing in Nature, Liu et al.5 propose that this asymmetry resulted from a head-on collision between the young Jupiter and a planetary embryo that had a mass about ten times that of Earth. The authors suggest that the primordial cores of the planet and of the embryo would have merged and then partially mixed with Jupiter’s envelope, explaining the structure of the planet seen today.

Scars of impacts abound on rocky planetary bodies. For example, the Moon is covered in craters, and was formed by a collision that occurred 4.5 billion years ago between Earth and a massive body6. Although impacts leave no direct imprint on the surfaces of fluid planets, the tilts of the rotational axes of Saturn (27°), Uranus (98°) and Neptune (30°) might indicate that violent collisions occurred in the past7. After all, it is known that massive planetary embryos on the order of ten Earth masses must have been present in the early Solar System8, in addition to the planets that are still here. Jupiter, with its small tilt (3°), seems to have escaped unscathed7. But according to Liu and colleagues, this was not the case.

Jupiter is mostly made of hydrogen and helium. However, observations of its atmospheric composition9 and gravitational field show that it contains a non-negligible proportion of heavier elements in the form of a central core and in the hydrogen–helium envelope. This envelope is fluid and is expected to be largely convective10, so it was surprising when Juno revealed that the envelope’s composition is not uniform. Instead, the core seems to be partially diluted in the envelope, extending to almost half of the planet’s radius3,4 (Fig. 1).

Figure 1 | Three phases of Jupiter. Liu et al.5 propose that the present-day internal structure of Jupiter is the result of a giant impact between the young planet and a planetary embryo that had roughly the mass of Uranus. a, In the authors’ model, before the impact, both Jupiter and the embryo contained a dense central core of heavy elements and a hydrogen–helium envelope. The colours represent the density of material, ranging from low (white) to high (dark orange). b, Just after the impact, the two cores merged and partially mixed with the planet’s envelope to produce a dilute core. c, After subsequent evolution, the dilute core remained, but was partially eroded into the envelope, causing the envelope to be enriched in heavy elements.

Producing this internal structure directly would require the delivery (accretion) of 10–20 Earth masses3,4 of heavy elements to the young Jupiter after the core had formed and during the first half of the growth of the envelope. The accretion of this material would need to have stopped after the planet had grown to about half of its present mass.

Formation models indicate that this hypothesis is unlikely. In these models, when Jupiter reaches about 30 Earth masses, the growth of the envelope by accretion is fast11, and the planet efficiently pushes away any dust particle that is millimetre-sized or larger12. As a result, the envelope should be poor in heavy elements. Any subsequent delivery of heavy elements by planetesimals (the asteroid-sized precursors of planets) or small planets is inefficient and cannot explain a heavy-element abundance that would increase with depth, as is observed. Erosion of the core into the envelope is possible10,13, but simulations show that this process tends to remove any small composition gradients that exist in the envelope, rather than increase them14.

The solution proposed by Liu et al. is simple. In their model, a planetary embryo that has a dense core of heavy elements collides with the forming Jupiter. The cores of the two bodies then merge and become partially mixed with Jupiter’s envelope. This explanation requires a massive embryo (of about ten Earth masses) and an impact that is somewhat head-on, but these two requirements seem reasonably likely. The authors show that cooling and subsequent convective mixing of the outer part of the envelope mixes only some of the heavy elements, leaving the planet’s dilute core relatively unaffected (Fig. 1). In one fell swoop, this picture might therefore explain the dilute core detected by Juno3,4 and the global abundance of heavy elements in Jupiter’s atmosphere9.

Liu and colleagues’ model should now be refined. In particular, it needs to be coupled to realistic scenarios for the formation of the Solar System8. Moreover, the mixing of heavy elements in the model should take into account heat and element diffusion — a process known as diffusive convection13. The results should also be compared quantitatively with constraints on Jupiter’s gravitational field from Juno1,2 and on the planet’s atmospheric composition obtained from spectroscopy10.

The authors’ model indicates that giant impacts might frequently occur during planet formation. This possibility could account for the tilts of the planets in the Solar System. It might also explain how some giant exoplanets, known as hot Jupiters, have accreted more than 100 Earth masses of heavy elements15,16 — a feature that is extremely difficult to obtain from conventional formation models. Hot Jupiters are situated close to their host stars, in regions in which the gravitational pull of the star is extremely strong. As a result, these exoplanets might be able to collect planetary embryos efficiently through a series of giant impacts, rather than ejecting them, and thus increase their heavy-element content.

Although giant planets have a fluid surface that cannot record traces of impact events, such planets hold clues to a violent past that led to the planetary systems observed today. The model proposed by Liu et al. enables present-day observations to be linked to the early days of the formation of the Solar System. Progress will come from an extension of studies such as this one to giant planets around the Sun and other stars. A continued exploration of the Solar System is crucial, particularly of Uranus and Neptune, which might be thought of as leftovers from a large population of massive planetary embryos in the early Solar System.

Nature 572, 315-317 (2019)

doi: 10.1038/d41586-019-02401-1


  1. 1.

    Folkner, W. M. et al. Geophys. Res. Lett. 44, 4694–4700 (2017).

  2. 2.

    Iess, L. et al. Nature 555, 220–222 (2018).

  3. 3.

    Wahl, S. M. et al. Geophys. Res. Lett. 44, 4649–4659 (2017).

  4. 4.

    Debras, F. & Chabrier, G. Astrophys. J. 872, 100 (2019).

  5. 5.

    Liu, S.-F. et al. Nature 572, 355–357 (2019).

  6. 6.

    Hartmann, W. & Davis, D. Icarus 24, 504–515 (1975).

  7. 7.

    Chambers, J. & Mitton, J. in From Dust to Life: The Origin and Evolution of Our Solar System 216 (Princeton Univ. Press, 2017).

  8. 8.

    Izidoro, A., Morbidelli, A., Raymond, S. N., Hersant, F. & Pierens, A. Astron. Astrophys. 582, A99 (2015).

  9. 9.

    Wong, M. H., Mahaffy, P. R., Atreya, S. K., Niemann, H. B. & Owen, T. C. Icarus 171, 153–170 (2004).

  10. 10.

    Guillot, T., Stevenson, D. J., Hubbard, W. B. & Saumon, D. in Jupiter: The Planet, Satellites and Magnetosphere (eds Bagenal, F., Dowling, T. E. & McKinnon, W. B.) 35–57 (Cambridge Univ. Press, 2004).

  11. 11.

    Mordasini, C., Alibert, Y., Klahr, H. & Henning, T. Astron. Astrophys. 547, A111 (2012).

  12. 12.

    Paardekooper, S.-J. & Mellema, G. Astron. Astrophys. 425, L9–L12 (2004).

  13. 13.

    Moll, R., Garaud, P., Mankovich, C. & Fortney, J. J. Astrophys. J. 849, 24 (2017).

  14. 14.

    Vazan, A., Helled, R. & Guillot, T. Astron. Astrophys. 610, L14 (2018).

  15. 15.

    Moutou, C. et al. Icarus 226, 1625–1634 (2013).

  16. 16.

    Thorngren, D. P., Fortney, J. J., Murray-Clay, R. A. & Lopez, E. D. Astrophys. J. 831, 64 (2016).

Download references

Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.