Solar model favors an engulfed super-Earth, but direct evidence is still missing

The Sun is measured precisely enough to reveal small shortcomings in models of its evolution. Standard calculations struggle to reproduce at once the sound-speed profile just below the outer convective layer, the depth of that layer, and some surface chemical abundances. A new study asked whether part of the mismatch could preserve the memory of a heavy-element-rich planet that fell into the young Sun.

Mutlu Yıldız of Ege University used the MESA stellar-evolution code to simulate two stages of material exchange. Heavy-element-rich matter, interpreted as a planet, would arrive first; gas depleted in those elements would then accrete from the planet-forming disk. He varied these accretion histories, turbulent mixing, and other parameters, comparing each simulated Sun with helioseismic measurements—stellar vibrations that probe the interior—and observed surface abundances.

Within the tested family, the best fit combined engulfment with turbulent mixing. The preferred model corresponded to a planet of about 5.6 Earth masses, while related variants favored 5 to 10 Earth masses. In the calculation, dissolved material forms a heavy-element-enriched region below the convective layer. This changes opacity and internal stratification, bringing the simulated sound-speed profile and the base of convection closer to values inferred for the Sun.

The study also required a specific condition to reproduce the observed scarcity of surface lithium: the accreted material would need to be lithium-poor, and the best models for this indicator used 4.6 to 5.8 Earth masses. Controls without engulfment but with adjustable mixing parameters improved some discrepancies. A comparison using the Bayesian Information Criterion, which penalizes added model flexibility, favored the engulfment solution; that test ranks the implemented alternatives rather than excluding every possible piece of physics.

A second calculation asked whether a compact rocky planet could cross the solar convective layer. With pressure-dependent planetary-structure models and only classical aerodynamic drag and ablation, calculated mass loss was small. That feasibility calculation does not describe the full event: shocks, fragmentation, instabilities, and three-dimensional dissolution were outside the one-dimensional model. Surviving this simplified calculation makes the scenario possible, but does not show that it happened.

The hypothesis links three present-day clues—helioseismic structure, surface composition, and lithium—to a shared chemical history early in the solar system. Its next test must be independent: search for the enriched layer or another predicted signature without selecting it to improve the same fit. Until then, 5 to 10 Earth masses should be read as the mass range that works best in the simulations, not as a measurement of a lost planet.