Table of Contents
Fetching ...

Differentiation, the exception not the rule -- Evidence for full miscibility in sub-Neptune interiors

Edward D. Young, Aaron Werlen, Sarah P. Marcum, Lars Stixrude, Cornelis P. Dullemond

TL;DR

This work investigates whether sub-Neptune interiors can host distinct silicate mantles and iron cores or instead form fully miscible MgSiO$_3$-Fe-H$_2$ melts deep in their interiors. By extrapolating binary Gibbs free energy of mixing surfaces for MgSiO$_3$-H$_2$, MgSiO$_3$-Fe, and Fe-H$_2$ to a ternary framework and employing a convex-hull approach, the authors map spinodal and binodal surfaces across relevant pressures and temperatures. The key finding is that, above modest hydrogen fractions and at interior pressures, the MgSiO$_3$-Fe-H$_2$ system becomes completely miscible, yielding under-dense, single-phase cores and shifting the core-envelope boundary from a discrete layered structure to a unified phase. This has significant implications for exoplanet mass–radius inferences, core EOS research, and our understanding of hydrogen storage in sub-Neptune interiors, with potential extensions to Neptune-like planets.

Abstract

We investigate the consequences of non-ideal mixing between silicate, iron metal, and hydrogen for the structures of the cores of sub-Neptunes with implications for super-Earths, warm Neptunes, and ice giants. A method of extrapolating what we know about the miscibility in the three bounding binary systems MgSiO$_3$-H$_2$, MgSiO$_3$-Fe, and Fe-H$_2$ to the ternary composition space is used to deduce the phase equilibria of this system at relevant temperature and pressure conditions. We find that while separate silicate and metal phases can exist at shallow depths, the phases become entirely miscible deeper in the cores, thus altering the density structure of the cores. The assumption that the interiors of large rocky planets, either with extant magma oceans beneath H$_2$-rich envelopes, or evolved from such bodies, are composed of a differentiated metal core overlain by a silicate mantle is inconsistent with our understanding of the phase equilibria of these bodies.

Differentiation, the exception not the rule -- Evidence for full miscibility in sub-Neptune interiors

TL;DR

This work investigates whether sub-Neptune interiors can host distinct silicate mantles and iron cores or instead form fully miscible MgSiO-Fe-H melts deep in their interiors. By extrapolating binary Gibbs free energy of mixing surfaces for MgSiO-H, MgSiO-Fe, and Fe-H to a ternary framework and employing a convex-hull approach, the authors map spinodal and binodal surfaces across relevant pressures and temperatures. The key finding is that, above modest hydrogen fractions and at interior pressures, the MgSiO-Fe-H system becomes completely miscible, yielding under-dense, single-phase cores and shifting the core-envelope boundary from a discrete layered structure to a unified phase. This has significant implications for exoplanet mass–radius inferences, core EOS research, and our understanding of hydrogen storage in sub-Neptune interiors, with potential extensions to Neptune-like planets.

Abstract

We investigate the consequences of non-ideal mixing between silicate, iron metal, and hydrogen for the structures of the cores of sub-Neptunes with implications for super-Earths, warm Neptunes, and ice giants. A method of extrapolating what we know about the miscibility in the three bounding binary systems MgSiO-H, MgSiO-Fe, and Fe-H to the ternary composition space is used to deduce the phase equilibria of this system at relevant temperature and pressure conditions. We find that while separate silicate and metal phases can exist at shallow depths, the phases become entirely miscible deeper in the cores, thus altering the density structure of the cores. The assumption that the interiors of large rocky planets, either with extant magma oceans beneath H-rich envelopes, or evolved from such bodies, are composed of a differentiated metal core overlain by a silicate mantle is inconsistent with our understanding of the phase equilibria of these bodies.

Paper Structure

This paper contains 19 sections, 26 equations, 13 figures.

Figures (13)

  • Figure 1: The Gibbs free energy of mixing surface for the MgSiO$_3$-H$_2$ system at 3591 K and 4 GPa as formulated by Stixrude2025. The binodes for the melt and envelope phases at this $T$ and $P$, as well as the tie line between them, are shown for clarity. Also shown are the positions of the spinodes where the curvature is zero.
  • Figure 2: Isobaric phase diagram for the MgSiO$_3$-H$_2$ binary system as determined by Stixrude2025. The 1 and 2-phase regions outside and inside of the binodal are indicated. A region of metastable compositions for two coexisting phases exists between the binodal and spinodal curves. The compositions of two stable phases at approximately 3600 K are shown by way of example with a dashed tie line connecting them.
  • Figure 3: Images of initial conditions and final conditions for the H$_2$ - MgSiO$_3$ system illustrating complete miscibility at 4,000 K and $3.5$ GPa. The initial condition is composed of MgSiO$_3$ melt overlain by H$_2$ gas. The final state is after $13.5$ ps of model time. See Appendix B for details.
  • Figure 4: Images of initial conditions and final conditions for the MgSiO$_3$-H$_2$ system illustrating immiscibility at 3,000 K and $2.5$ GPa. The initial condition is composed of MgSiO$_3$ melt overlain by H$_2$ gas. The final state is after $16.5$ ps of model time. Example SiO and H$_2$O molecules in the gas phase that persist for significant intervals of model time are labelled.
  • Figure 5: Images of initial conditions and final conditions for the MgSiO$_3$-Fe system illustrating complete miscibility at 9,000 K and $60$ GPa. The final state is after 7 picoseconds of model time. Time steps were 0.5 fs. Model consists of 76 Fe atoms, 18 Mg atoms, 18 Si atoms, and 54 O atoms.
  • ...and 8 more figures