Deep beneath the Sun’s visible surface, a narrow layer of extra heavy elements may hold the clue to why standard models struggle to reproduce the star’s interior. A new set of solar models suggests that the material could have been left behind when the young Sun swallowed a rocky planet of roughly five to 10 Earth masses.
The Sun is one of the best-observed stars, and helioseismology provides unusually detailed information about its interior. Among the strongest constraints are the speed of sound through much of the Sun, the depth of its outer convection zone and the chemical composition of its surface.
Yet standard solar models have a persistent problem. Just beneath the convection zone, the models generally predict a lower sound speed than helioseismic measurements indicate, while placing the bottom of the convection zone farther from the center of the Sun than observations require. Surface chemical abundances can also disagree with observations, depending on the model’s assumed composition.
The study by astronomer Mutlu Yıldız explores whether the problem could be addressed by changing the Sun’s chemical structure early in its history.
The central idea is unusual: some of the material that now sits beneath the Sun’s convection zone may have come from a planet that was swallowed by the young star.
Rather than treating the planet as a fully intact object that simply disappears into the Sun, the models represent its contribution as metal-rich material deposited beneath the convection zone. There, the added heavy elements change the local composition, opacity and internal stratification.
The calculations indicate that this arrangement can substantially reduce the mismatch between the modeled and observed solar sound-speed profiles.
A chemical layer left behind by planetary material
The proposed scenario starts during the Sun’s early evolution, when its surrounding protoplanetary disc was still supplying material to the young star.
Planet formation separates material chemically. Heavy elements become concentrated in planetesimals and planets, while some of the remaining hydrogen- and helium-rich gas can become relatively poor in heavy elements. The study models this history with two broad episodes: metal-rich material associated with planetary engulfment, followed by accretion of metal-poor gas.
That sequence matters because simply adding heavy elements to the Sun’s surface would not reproduce the required structure. The models need a localized enhancement beneath the convection zone.
In the best-fitting models, the enhanced-metallicity region lies around 0.67–0.70 times the Sun’s radius, corresponding to a mass coordinate of roughly 0.96–0.973 solar masses. The added heavy elements increase the opacity in that region and alter the stratification in a way that brings the modeled sound speed closer to the helioseismic result.
The result is not simply a better match to one measurement.
The strongest models simultaneously reproduce the solar sound-speed profile, the depth of the convection zone and the observed surface helium and heavy-element abundances much more closely than the standard models considered in the study.
The planet’s mass emerges from the solar model
Once the model contains a localized excess of heavy elements, the author can estimate how much planetary material would be needed to produce it.
For an earlier model, the metal-rich region extended across a mass interval of about 0.947–0.974 solar masses. Integrating the excess heavy elements gave 16.457 Earth masses of additional heavy material. Using the study’s adopted relation between planetary core mass and total planetary mass produced an estimated engulfed planet of 28.4 Earth masses for that particular model.
But the inferred mass changes substantially with the assumed internal solar composition and with the details of the model.
The models that best match the newer, higher solar metallicity constraint generally require much less planetary material. Model NCz195, for example, gives a dissolved planetary mass of about one Neptune mass and an estimated total engulfed-planet mass of about 1.83 Neptune masses.
Other closely matching models produce still lower values.
The strongest overall solution, Model DD1020, points to an engulfed planet with a mass of about 5.6 Earth masses, or about 0.33 Neptune masses. Related models that also fit the solar constraints favor planets between roughly 4.6 and 5.8 Earth masses. Across the broader set of best-fitting models, the preferred range is approximately 5–10 Earth masses, placing the proposed object in the super-Earth regime.
The author calls the hypothetical planet Dev Dilek.
The exact mass should not be treated as a precise measurement. The study shows that the inferred planetary mass depends sensitively on the assumed deep solar metallicity and on how the metal enhancement is distributed beneath the convection zone. Increasing the initial metallicity generally lowers the inferred dissolved planetary mass.
Other ways of changing the Sun’s model
The study also tests whether the improvement could be produced without planetary engulfment.
One possibility is turbulent mixing below the convection zone. The calculations show that turbulent mixing can improve the sound-speed agreement over some depths, but the improvement is followed by larger discrepancies deeper down. On its own, it does not reproduce all the observational constraints.
Thermohaline mixing performs less well in the tested models. Neither of the two adopted thermohaline prescriptions produces a significant improvement in agreement with the helioseismic constraints.
The mixing-length parameter, which controls how convection is represented in the stellar models, can also be varied with time. Several such models improve the agreement, and the favored engulfed-planet masses in these calculations are around eight Earth masses.
The author then combines planetary engulfment with turbulent mixing.
That combination produces the closest matches among the models explored. Three models with time-dependent turbulent-mixing depths give sound-speed differences below 0.0008 across the solar interior, while their other structural parameters also agree closely with the helioseismic constraints. The corresponding engulfed-planet masses range from 7.2 to 10.5 Earth masses.
A further set of models starts with a solar mass of 0.97 times the present solar mass and uses two accretion episodes without mass loss. These models achieve root-mean-square sound-speed differences of about 0.00034–0.00036. They consistently favor an engulfed rocky planet of approximately 4.6–5.8 Earth masses.
The lithium problem adds another constraint
The Sun contains far less lithium at its surface today than expected from its inferred initial abundance. The study uses this as another test of the proposed accretion history.
A standard solar model in the calculation produces a present-day lithium abundance of A(Li) = 2.22, compared with the observed value of about 1.1. When the model’s accreted material is assumed to be lithium-free, the predicted abundance falls substantially, although not all the way to the observed value in every model.
The low-mass, two-accretion models perform better. With lithium-free accretion, Models DD1020 and DD1020p reach A(Li) of about 1.25. Adjusting the initial lithium abundance downward by about 0.15 dex brings the predicted present-day abundance to approximately 1.1.
Models with a different division between the two accretion episodes give present-day lithium abundances of 1.22 and 1.21, while continuing to match the helioseismic constraints closely. The calculation still requires the accreted material to be essentially lithium-free.
The lithium calculation also places a constraint on mixing.
The models contain a sharp lithium decline immediately below the convection zone, where lithium is transported into deeper, hotter layers and destroyed. But if mixing were extended too deeply, it would also smear out the heavy-element enhancement produced by the engulfed planet, removing the very structure that improves the solar sound-speed model.
The improvement is not explained only by extra parameters
Because the engulfment models have many adjustable parameters, the study tests whether their success could simply result from having more freedom to fit the observations.
The best engulfment model, DD1020, has 10 adjustable parameters. A control model called PureENGF uses five, while another model without engulfment, TurMixaV, uses six and includes turbulent mixing and a time-dependent mixing-length parameter. The comparison uses the observed surface heavy-element abundance, surface helium abundance, convection-zone depth and sound speed at 71 locations inside the Sun.
The resulting total χ² values are 5,066.01 for DD1020, 21,816.17 for PureENGF and 24,139.65 for TurMixaV. Their corresponding Bayesian Information Criterion values are 5,109.09, 21,837.71 and 24,165.49.
The absolute χ² values remain large because helioseismic sound-speed uncertainties are extremely small, so even modest residual differences contribute substantially. The study therefore emphasizes the relative comparison between models fitted with the same observational constraints and procedure.
Within that comparison, the engulfment models perform substantially better than the optimized non-engulfment model.
Could a rocky planet actually reach that deep?
A successful solar model is not enough if the proposed planet could not physically survive its journey into the young Sun.
The study therefore examines several stages of that journey, beginning with the planet’s Roche lobe, the region around it within which its own gravity dominates its material.
Observed super-Earths and Neptune-like planets generally have radii larger than their corresponding Roche lobe radii under the conditions examined, implying that volatile-rich planets could undergo substantial mass loss before entering the stellar envelope. But the study’s compact, core-dominated rocky planet models have radii close to the Roche-lobe limit across a broad range of planetary masses.
That suggests that a rocky planet, or the stripped core of a once more volatile-rich planet, could approach the stellar surface without catastrophic Roche-lobe overflow.
Once inside the Sun, however, the Roche-lobe calculation no longer applies. Pressure and density rise rapidly with depth, changing the planet’s structure and its interaction with the surrounding stellar material.
For the best-fitting solar model, the gas pressure between the surface and the base of the convection zone increases by nearly nine orders of magnitude, while the density rises by more than six orders of magnitude. The study models how that pressure compresses rocky planets with masses of 6, 8 and 10 Earth masses.
The planets shrink as they descend. The 10-Earth-mass model becomes about 26 percent smaller in radius by the base of the convection zone, while the 6-Earth-mass model contracts by about 29 percent. The shrinking reduces the planet’s cross-sectional area, which in turn lowers the drag and classical ablation rates.
The calculation does not include thermal expansion from frictional heating, shocks or heat exchange with the surrounding plasma. The author therefore treats the pressure-induced contraction as a lower limit on the planet’s radius under engulfment conditions.
Classical ablation would not destroy the planet on the way down
The final test considers aerodynamic drag and ablation as the planet moves through the convective envelope.
For a reference calculation, the planet has a mass of 10 Earth masses and an initial velocity of 447 kilometers per second. The calculation uses a drag coefficient of 1, a heat-transfer coefficient of 0.05 and a specific ablation energy of 3 × 10¹¹ erg per gram. The planetary radius changes with pressure according to the interior models.
Under those assumptions, the calculated mass loss is negligible compared with the planet’s original mass.
The planet crosses the convective envelope in only a few hours, and its pressure-compressed radius reduces its effective cross-sectional area as it descends. Even with parameters chosen to favor efficient aerodynamic erosion, the calculation removes only a negligible fraction of the planet before it reaches the base of the convection zone.
That does not establish exactly how the planet would ultimately dissolve. Fragmentation, shocks, thermal expansion, hydrodynamic instabilities and deeper-envelope interactions are not modeled in full. The study explicitly notes that a complete treatment would require a coupled calculation of orbital dynamics, planetary structure, thermal effects and hydrodynamics.
The result is narrower: classical drag and ablation alone do not appear sufficient to destroy a compact rocky planet before it reaches the base of the solar convection zone.
Taken together with the stellar models, the calculations make a specific case for planetary engulfment as a possible source of the Sun’s localized heavy-element layer. The strongest model combines an early planetary engulfment with turbulent mixing and favors a super-Earth of about 5.6 Earth masses, while related models favor roughly 5–10 Earth masses. The proposed planetary material settles below the convection zone, where its heavy elements alter the opacity and internal stratification in a way that improves the match to helioseismic observations.
The study does not provide a detailed dynamical history of such an engulfment. Its authors instead identify the chemical structure needed in solar models and test whether a compact rocky planet could plausibly survive far enough into the young Sun to produce it. A full hydrodynamic treatment of the planet’s orbital evolution, structural changes, shocks, thermal effects and final dissolution remains outside the calculation.
The study was published in Monthly Notices of the Royal Astronomical Society.
