One Star of a Twin Pair Ate More Than 11 Earths' Worth of Rock. Beryllium Gave It Away.

Two stars born from the same cloud of gas should be chemical twins. HD 129171 and HD 129209 are exactly that kind of pair — two sun-like stars in a binary system, matched in size, temperature and magnetic activity — and an international team led by the University of São Paulo has found that they are not chemically identical at all. One of them, the researchers say, has been eating its own planets.

"The star HD 129171 is enriched in refractory elements — that is, elements that typically condense in the solid state and make up rocky planets. That strongly suggests that it has engulfed planetary material throughout its evolution," said Anne Rathsam, a doctoral student at the Institute of Astronomy, Geophysics and Atmospheric Sciences at USP and lead author of the study, published in Astronomy & Astrophysics.

Astronomers have suspected for years that stars sometimes swallow their planets. What makes this result new is the tracer. Almost every chemical element in the universe comes either from the first minutes after the Big Bang or from fusion inside stars. Lithium, beryllium and boron are the exceptions — they are made mainly by cosmic spallation, in which high-energy particles smash heavier nuclei like carbon, nitrogen and oxygen into lighter fragments, as study adviser Jorge Luis Melendez Moreno explained. A star does not manufacture beryllium in its core. So finding a beryllium signature in a star's light is a flag that something from outside got in, long after the star formed.

Lithium had been tried as an indicator before and proved unreliable. "Lithium had already been used as a possible indicator of planetary engulfment, but it's destroyed relatively easily," Rathsam said. "Beryllium is more resistant, and its chemical signature can last longer." The team used the UVES spectrograph on the European Southern Observatory's Very Large Telescope in Chile, which splits starlight finely enough to pick out extremely subtle chemical signatures.

HD 129171 turned out to carry noticeably more iron, magnesium, silicon, calcium and titanium than its companion, plus an excess of both lithium and beryllium. Run through the models, that pattern corresponds to swallowing rocky material equal to more than 11 times the mass of Earth. Whether that was one large planet or a series of smaller bodies is not recoverable. "In the case of sun-like stars, internal mixing is so efficient that the final chemical signature doesn't allow us to distinguish between those scenarios," Rathsam said.

The mechanisms that drive a planet into its star are well catalogued: gravitational tugging between planets, perturbations from a companion star, orbital migration. Any of these can stretch an orbit into a long ellipse and destabilize it until the planet is thrown out of the system, collides with a sibling, or falls in. Roughly half the stars in the Milky Way have a gravitational companion, which means the raw ingredients for that kind of chaos are extremely common.

The uncomfortable implication is about our own arrangement. "When we bring together evidence from dynamical simulations, exoplanet observations, and chemical studies of binary stars, a consistent picture emerges, indicating that systems similar to the solar system may be less common than we imagined," Melendez said. Computer models of planet formation rarely produce what we have — giant planets on nearly circular outer orbits, rocky planets stable on the inside — and surveys of sun-like stars have found few Jupiter analogues on Jupiter-like orbits.

That bears directly on the odds for life. "Life wouldn't just need billions of years to emerge and evolve," Rathsam said. "The planet would also have to remain in a sufficiently stable orbit to survive significant gravitational perturbations." The work also touches a technique called chemical tagging, which reconstructs the Milky Way's history from stellar compositions and assumes that twin stars start identical. The team's results support planetary ingestion as the cause of the mismatch rather than lumpiness in the original cloud — which is the answer that leaves current star-formation models standing. Researchers from the Polish Academy of Sciences, the Chinese Academy of Sciences, Monash University and Italian observatories took part.