The engulfment of a planet by its host star HD82943 (Animation)

What may at a first glance look like the recipe for a dramatic
science-fiction story is in fact the well-considered conclusion of a
serious scientific study, to be published by a group of astronomers in
Switzerland and Spain [1] in tomorrow’s issue of the international
research journal “Nature”.

10 May 2001: Evidence for planet engulfment by the star HD82943, Nature

Using the very efficient UVES high-resolution spectrograph at the ESO VLT
8.2-m KUEYEN telescope, they have convincingly detected the presence of
the rare isotope Lithium-6 (6Li; [2]) in this metal-rich, solar-type dwarf
star that is also known to possess a planetary system, cf. ESO Press
Release 13/00.

Unlike the Lithium-7 (7Li) isotope of this light element, any primordial
Lithium-6 would not survive the early evolutionary stages of a metal-rich
solar-type star. The Lithium-6 now seen in HD 82943 must therefore have
been added later, but from where? The astronomers believe that this
observation strongly suggests that the star has at some moment engulfed
one of its planets, whose Lithium-6 was then deposited in the star’s

This surprising discovery represents important observational evidence that
planets may fall into their host stars.

The engulfment of a planet by its host star HD82943

small |

medium |


HD 82943 and its planetary system

The last few years have seen the discovery of more than 60 new planetary
systems. One of the most prolific planet search programmes is being carried
out by the Geneva Extra-Solar Planet Search Group, by means of the CORALIE
spectrograph at the 1.2-m Leonard Euler Swiss Telescope at the ESO La Silla
Observatory (Chile).

One of the stars included in this programme is the dwarf star HD 82943 in
the constellation Hydra (The Water Snake). It is slightly hotter and larger
than the Sun and was recently found to harbour a planetary system with (at
least) two giant planets, cf. ESO Press Release 07/01.

Like most extra-solar planets (“exoplanets”) found to date, the orbits
of the objects orbiting HD 82943 are quite unlike those expected from
traditional theories of the formation and evolution of such systems [3].
Contrary to the giant planets in the Solar System, those at HD 82943 have
rather elongated orbits, and they are unsually close to the central star.

Astronomers believe that giant planets must form in comparatively cool
environments, as this was the case in the solar system. The existence of
systems in which the giant planets are much closer to the central star can
only be explained by certain dynamical processes, e.g. significant orbital
changes with time (“orbital migration”) or the effects of strong
gravitational interaction between several planets.

These processes can explain the short-period planetary systems found to
date, in which planets are very close to the central star, and also the
very elongated orbits found in some cases.

These theories also predict that it may be the fate of some planets to fall
into their host star.

The significance of Lithium

Unlike most other elements lighter than Iron, the light nuclei of Lithium
(both the Lithium-6 and Lithium-7 isotopes [2]), Beryllium and Boron are
not produced in significant amounts in the stellar spheres of fire.

In fact, Lithium-6 is extremely “fragile”, being easily destroyed by proton
collisions at a temperature of “only” 1.5 million degrees — by comparison,
the fusion of Hydrogen to Helium takes place at about 10 million degrees.
In the case of solar-like stars, any Lithium-6 atoms present in a newborn
star will be “burnt” during the early evolutionary stages. Strong internal
motions will thoroughly mix the outer (cooler) and inner (hotter) stellar
layers, and Lithium-6 will completely disappear in just a few million
years. We would therefore not expect to find any Lithium-6 in a developed
solar-type star.

However, during the later evolutionary stages the outer layers of a
solar-type star remain better “separated” from the hotter central parts.
Thus, if some Lithium-6 is now picked up from the outside, it is therefore
possible that it will be preserved in the upper, cooler regions for some
time, possibly billions of years.

Unlike stars, planets never reach temperatures that are high enough to burn
their initial content of Lithium-6. Consequently, planets will retain their
Lithium-6. So, if a planet happens to fall into a solar-type star like HD
82943, we may then be able to detect this isotope in the stellar spectrum.

In the case of “metal-poor” stars — that are less rich in metals than the
Sun — the mixing process in the early phase is less efficient and some
original Lithium-6 may actually survive.

The possible presence of Lithium-6 in a stellar atmosphere can be checked
by means of a detailed analysis of the star’s spectrum. For this, the
astronomers search for a very small asymmetry in the “stronger” absorption
line in the red spectral region that is caused by Lithium-7 atoms in the
stellar atmosphere.

However, this type of investigation is critically dependent on the
availability of very detailed and “clean” spectra (i.e., very high spectral
resolution and excellent signal-to-noise ratio). This is a great
observational challenge and to date, only about five stars are known to
display the signatures of Lithium-6 in their spectra, cf. ESO PR 08/00.
In all cases, the measured isotopic abundance ratio is very small, with
6Li/7Li less than about 0.05. All of these stars are metal-poor and may
have retained some of their initial Lithium-6, see above.

Until now, no convincing detection of Lithium-6 has ever been made in a
metal-rich, solar-type star.

The UVES spectrograph at the 8.2-m VLT KUEYEN telescope is perfectly suited
for this kind of study. Three high-resolution spectra of HD 82943 were
obtained in June 2000 that show a significant asymmetry in the Lithium-7
absorption line, cf. PR Photo 17/01. After a careful analysis, this
asymmetry is confirmed as the spectral signature of Lithium-6 atoms. The
observed abundance ratio is 6Li/7Li = 0.12. This is unusually high when
compared to the detections in metal-poor stars and is in fact more
compatible with the value of 0.08, observed in solar-system meteorites!

HD 82943 has swallowed a planet

The astronomers believe they know the answer: “The simplest and most
convincing way to explain this observation is that one or more planets,
or at least planetary material, have fallen into the star, sometime after
it passed through its early evolutionary stage”, says Nuno Santos of the
Geneva Observatory.

Garik Israelian of Instituto de Astrofisica de Canarias adds: “One may also
try to determine the quantity of material needed to explain the observed
isotopic ratio of 0.12. Based on the mass estimate of the star HD 82943 and
the known Lithium-6 content of meteorites, it appears that the star has
swallowed the equivalent of a giant planet with twice the mass of Jupiter”.
If the unlucky planet were of the terrestrial type, in which the relative
Lithium-6 content is higher, it would have had a mass of about three times
the mass of the Earth.

The observational search for Lithium-6 in other stars with planetary systems
now continues. In due time, it will permit to better understand the
formation and evolution of the newly discovered exoplanets. In particular,
it will demonstrate whether the fall of planets into their host stars is a
common process or not.

More information

Further detailed information is available in the research article (“Evidence
for planet engulfment by the star HD 82943”, by G. Israelian, N.C. Santos,
M. Mayor and R. Rebolo), published in the May 10, 2001, issue of the
international research journal Nature.


[1] The team consists of Garik Israelian and Rafael Rebolo (Instituto de
Astrofisica de Canarias, Spain), Nuno C. Santos and Michel Mayor (Geneva
Observatory, Switzerland).

[2] The nuclei of Lithium-6 (6Li) atoms consist of three protons and three
neutrons; those of Lithium-7 (7Li) have three protons and four neutrons.
Both isotopes were produced during the Big Bang and in spallation reactions
in the interstellar medium.

[3] According to the “traditional” view, giant planets like Jupiter would
be formed by rapidly accelerating (“runaway”) accretion of gas around an
initial, icy “planetesimal” with a mass of about 10 Earth masses. An
associated prediction was that giant planets would only be found at a
distance of at least 750 million kilometres (5 Astronomical Units; or five
times the distance between the Earth and the Sun) from their host stars and
that their orbits would be circular, like the orbits of the planets in the
Solar System.

Technical information about the photo

PR Photo 17/01 shows the spectrum of the V = 6.5 mag star HD 82943, as
obtained on June 7, 2001, with the UVES spectrograph at the 8.2-m VLT KUEYEN
telescope. It is based on three exposures made with Image Slicer 3 in
Director’s Discretionary Time in Service Mode, and each lasting 120 sec.
The spectral resolution is 110,000 and the final S/N-ratio is about 500.

Artist’s impression of the engulfment of a planet

[Artist’s Impression] [Planet Engulfment – Video Clip – MPEG – 1.3Mb]

“The engulfment of a planet by its host star”. Artist’s impression, prepared
by Gabi Perez (IAC). Credits to Gabi Perez and the IAC.


Garik Israelian and Rafael Rebolo

Instituto de Astrofisica der Canarias

La Laguna, Tenerife, Spain

Tel.: +34-922-605200

email: and

Nuno Santos and Michel Mayor

Observatoire de Geneve


Tel.: +41-22-755-2611

email: and

Text with all links and the photo is available on the ESO Website at URL: