Soot Planets instead of Water Worlds

TL;DR

The paper argues that low-density sub-Neptunes need not be rocky cores plus water ice; instead, planets with substantial refractory carbon (soot) can yield similar low densities. It builds four-component planet models (rock, soot, water, hydrogen), treats soot as a fictive carbon-rich phase with a plausible density, and computes mass–radius relations under stratified and mixed interior limits using a Rose–Vinet equation of state. The findings show that soot-rich interiors can reproduce observed mass–radius trends previously attributed to rock+water compositions, with degeneracies relative to possible hydrogen envelopes and distinct atmospheric signatures such as methane-rich atmospheres and hydrocarbon hazes; JWST observations of carbon-bearing species in some sub-Neptunes align with these predictions. The study broadens the interpretation of exoplanet interiors, suggests soot-rich planets are likely common, and provides open-source tools for modeling M–R relations, while acknowledging uncertainties in soot chemistry and high-pressure carbon phases that warrant further investigation.

Abstract

Some low-density exoplanets are thought to be water-rich worlds that formed beyond the snow line of their protoplanetary disc, possibly accreting coequal portions of rock and water. However, the compositions of bodies within the Solar System and the stability of volatile-rich solids in accretionary disks suggest that a planet rich in water should also acquire as much as 40% refractory organic carbon (``soot''). This would reduce the water mass fraction well below 50%, making the composition of these planets similar to those of Solar System comets. Here we show that soot-rich planets, with or without water, can account for the low average densities of exoplanets that were previously attributed to a binary combination of rock and water. Formed in locations beyond the soot and/or snow lines in disks, these planets are likely common in our galaxy and already observed by JWST. The surfaces and interiors of soot-rich planets will be influenced by the chemical and physical properties of carbonaceous phases, and the atmospheres of such planets may contain plentiful methane and other hydrocarbons, with implications for photochemical haze generation and habitability.

Figures (8)

• Figure 1: Illustration of the soot line and water ice line structure of protoplanetary disks, and the three chemically distinct planet archetypes. The water ice line is also known as the water snow line or snow line. Compositions of the model rocky planet, soot planet, and soot-water worlds are shown as pie charts, with red, yellow, and blue segments representing the mass fractions of rock, soot, and water, respectively (Table \ref{['Tab:composition']}). For the soot-water worlds, the pie chart and numbers refer to the end-member dry case and the numbers for the end-member wet case are: 36% rock, 14% soot, and 50% water. The soot line is defined by its sublimation temperature of 500 K under disk conditions Li21. The water ice line is placed at 160 K Minissale22. The locations of the soot lines and ice lines vary with time and across stellar mass range. The soot line and ice line are drawn for illustrative purposes, Their actual shapes depend on the pressure distribution in the disk, as well as the intensity of viscous heating and the contribution from irradiation from the central star, which vary over time and with physical conditions in the disk. For this study, it is the relative locations that matter. Figure credit Ari Gea and Sayo Studio.
• Figure 2: Correlation between density at 100 kPa (1 bar) $\rho_0$ and mean atomic number $\overline{Z}$ for a range of plausible solid phases in rocky or icy planetary interiors (data sources in Table \ref{['Tab:rho_z']}). The gray band is a linear fit, $\rho_0$ = 0.327 $\overline{Z}$, for 11 selected phases that are thermodynamically stable at 100 kPa and 300 K (red open circles). Measured or fitted 100 kPa densities of high-pressure phases (black filled squares) mostly plot above the gray band, as expected. The cryogenic ice phases (blue open circles) may be considerably less dense and plot below the gray band. The solid line is linear fit, $\rho_0$ = 0.318 $\overline{Z}$, for 43 phases that are thermodynamically stable at 100 kPa and 300 K. The dashed line is linear fit, $\rho_0$ = 0.303 $\overline{Z}$, for all 48 phases. Fitting the 43 phases and all 48 phases yielded 1.26 g/cc and 1.33 g/cc for the fictitious soot component, respectively. The arrow points to soot (the red filled circle, C:H:O = 100:78:17 in atomic ratio, mean atomic number = 4.17), with an estimated density at 100 kbar and 300 K, $\rho_0$, of 1.32(6) g/cc.
• Figure 3: Mass-Radius relations for model Earth-like rocky planets (black curves), soot planets (gray bands), and soot-water worlds (blue bands). Panel A is for multi-layer planets and Panel B is for single-layer planets. The shaded bands for the soot planets and soot-water worlds encompass end-member cases where the soot component varies from highly incompressible (diamond) to highly compressible (water ice) (Fig. \ref{['fig:eos']}). Shown for comparison are curves for model planets consisting of 50% Earth-like material and 50% water ice (blue dotted curves) along with Earth-like rock planets with 1% by mass H$_2$ envelope (purple band), as well as exoplanets with up to 3 Earth radii and up to 10 Earth masses, and for which masses and radii are known to $<$20% and $<$10%, respectively (blue crosses, NASA Exoplanet Archive: https://exoplanetarchive.ipac.caltech.edu.)
• Figure A1: Histogram of stellar C/O ratios, (C/O)$_\star$, for planet host stars as compiled by the Hypatia Catalog Hypatia. The histogram shown in red (High Mass Planets) isolates systems where the most massive planet is $\ge$30 M$_\oplus$ and the one is blue (Low Mass Planets) isolates those with all planet mass below that value. The median error for each sample is shown with its magnitude comparable to the length of the line. The solar value is also show as reference using oxygen abundance estimates by Bergemann et al. 2021MNRAS.508.2236B and oxygen by Asplund et al. 2021AA...653A.141A.
• Figure A2: Correlation between density and mean atomic number at 100 kPa (1 bar).