The formation of circumbinary planets through disc fragmentation

Dimitris Stamatellos; Matthew Teasdale

arxiv: 2604.24420 · v1 · submitted 2026-04-27 · 🌌 astro-ph.SR · astro-ph.EP

The formation of circumbinary planets through disc fragmentation

Matthew Teasdale , Dimitris Stamatellos This is my paper

Pith reviewed 2026-05-08 01:44 UTC · model grok-4.3

classification 🌌 astro-ph.SR astro-ph.EP

keywords circumbinary planetsdisc fragmentationgravitational instabilitygas giant planetshydrodynamic simulationsprotoplanetsbinary starsexoplanet formation

0 comments

The pith

Gravitational fragmentation in circumbinary discs produces gas giant planets, more efficiently than around single stars.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Hydrodynamic simulations compare marginally unstable discs around single stars and binaries with two temperature setups. Wider binaries and realistic heating by each star separately cause earlier and more prolific fragmentation than closer binaries or single-star discs. Realistic circumbinary cases yield roughly nine protoplanets per disc, many starting in the planetary-mass range, with formation concentrated beyond a binary-imposed gap near 50 AU and final orbits peaking around 100 AU. The same runs also eject a larger fraction of clumps at speeds of 2-6 km/s, creating free-floating objects. The work concludes that this pathway can account for observed wide-orbit circumbinary gas giants.

Core claim

Simulations of marginally unstable circumbinary discs demonstrate that gravitational fragmentation occurs, earlier and more efficiently for wider binary separations and under realistic individual stellar heating, producing on average 9 protoplanets per disc with initial masses biased toward the gas-giant regime rather than brown dwarfs, forming predominantly beyond a forbidden region of about 50 AU with orbits peaking near 100 AU, while also ejecting more objects than single-star discs.

What carries the argument

Hydrodynamic simulations of marginally unstable discs comparing circumstellar cases, fiducial circumbinary discs with shared temperature profiles, and realistic circumbinary discs heated separately by each star.

If this is right

Discs around wider binaries fragment earlier and more efficiently than those around closer binaries or single stars.
Realistic circumbinary discs form 9 plus or minus 0.9 protoplanets per disc, exceeding the 6.5 plus or minus 0.6 in fiducial circumbinary and 7.5 plus or minus 0.8 in circumstellar cases.
Fragmentation occurs beyond a binary-imposed region of roughly 50 AU, with final orbital radii peaking near 100 AU.
Circumbinary discs eject a higher fraction of protoplanets than circumstellar discs, producing free-floating objects with velocities of 2-6 km/s.
Initial protoplanet masses are lower in realistic circumbinary discs, favoring gas giant planets over brown dwarfs.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The mechanism may help explain the growing sample of observed circumbinary gas giants on orbits wider than 10 AU.
Ejected clumps could contribute to the population of free-floating planets associated with binary systems.
Incorporating magnetic fields or full radiative transfer would provide a direct test of whether the reported fragmentation efficiency persists.
The binary-imposed forbidden region offers a testable signature in the orbital distribution of future detections.

Load-bearing premise

The chosen initial disc conditions, temperature profiles, and purely hydrodynamic treatment without magnetic fields or full radiative transfer accurately represent real circumbinary disc evolution and allow the formed clumps to survive as bound planets.

What would settle it

Detection of circumbinary gas giants with orbital radii or mass distributions that contradict the simulated peaks beyond 50 AU and planetary-mass bias, or absence of fragmentation when the same initial conditions are rerun with magnetohydrodynamics or detailed radiative transfer.

read the original abstract

Over 50 circumbinary exoplanets have been discovered in recent years, with several of them being gas giants on wide orbits ($>10$AU). The aim of this work is to investigate whether these planets can form through circumbinary disc fragmentation due to gravitational instability. We perform hydrodynamic simulations of marginally unstable (i) circumstellar discs, (ii) circumbinary discs with the same temperature profile as the circumstellar discs (fiducial model), and (iii) realistic circumbinary discs heated individually by each star of the binary. We find that discs around binaries with wider separations fragment earlier and more efficiently than those around closer binaries, and earlier than circumstellar discs. Realistic circumbinary discs form a larger number of protoplanets ($9\pm0.9$ protoplanets per disc), than fiducial circumbinary ($6.5\pm0.6$), and circumstellar discs ($7.5\pm0.8$). In realistic circumbinary discs, initial protoplanet masses are lower than those formed in circumstellar discs, and a larger fraction of them lie in the planetary-mass regime, favouring the formation of gas giant planets over brown dwarfs or low-mass stars. Fragmentation occurs predominantly beyond a binary-imposed forbidden region of $\sim50$AU, leading to final orbital radii peaking at $\sim100$AU. We also find that in circumbinary discs dynamical interactions eject a higher fraction of protoplanets than in circumstellar discs, producing free-floating objects, with ejection velocities on the order of $2-6~{\rm km s^{-1}}$. We conclude that gravitational fragmentation of circumbinary discs is a viable and potentially significant formation pathway for circumbinary gas giant planets.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Hydro sims find realistic circumbinary discs fragment into more planetary-mass objects at ~100 AU than single-star discs, with clear differences from fiducial models, though the hydrodynamic-only setup limits how far the viability claim can be pushed.

read the letter

The main result is that circumbinary discs heated separately by each star fragment earlier and more than circumstellar discs, yielding 9 ± 0.9 protoplanets per disc versus 7.5 ± 0.8, with lower initial masses that stay in the gas-giant range and final orbits peaking near 100 AU after a ~50 AU gap. Wider binaries fragment more efficiently, and ejections happen at 2-6 km/s, producing free-floaters at higher rates than in single-star cases. The fiducial circumbinary runs sit in between, showing the effect of the realistic temperature profile. They average over multiple runs, which gives the numbers some weight. This is a straightforward numerical comparison that was not in the earlier literature they cite. The controlled setup across the three disc types is the useful part. It supplies concrete counts and distributions that observers or other simulators can check against. The soft spot is the physics that is left out. Everything is purely hydrodynamic with fixed temperature profiles and no magnetic fields or full radiative transfer. Real discs have fields that can add pressure or drive turbulence, and cooling rates that set the fragmentation line depend on proper radiation transport rather than the imposed heating. The runs show initial clumps and ejections, but do not track whether those clumps stay bound and grow over long timescales without being torn apart. That directly affects whether the higher planetary-mass fraction survives as a real formation channel. This is for people who model planet formation in binary systems and want quantitative outputs from GI simulations to compare with the growing sample of wide circumbinary giants. It is worth sending to peer review because the experiments are set up with clear controls and produce specific, testable numbers even if the physics needs more scrutiny.

Referee Report

2 major / 2 minor

Summary. The paper investigates the formation of circumbinary planets via gravitational instability in discs using hydrodynamic simulations. It compares three cases: circumstellar discs, fiducial circumbinary discs with identical temperature profiles, and realistic circumbinary discs with individual stellar heating. Key results include earlier and more efficient fragmentation in wider binaries and realistic models, with realistic circumbinary discs producing 9±0.9 protoplanets per disc, lower initial masses favoring gas giants, fragmentation outside ~50 AU leading to orbits peaking at ~100 AU, and higher ejection rates producing free-floating objects. The authors conclude that this is a viable pathway for circumbinary gas giant planets.

Significance. If the central findings are robust, the work would be significant as it provides the first quantitative comparison of fragmentation efficiency between circumstellar and circumbinary environments, with specific predictions for protoplanet numbers, masses, and orbital distributions that can be tested against observations of the over 50 known circumbinary exoplanets. The use of multiple runs to report averages with uncertainties is a strength, as is the identification of a binary-imposed forbidden region for fragmentation.

major comments (2)

[§2 (Numerical Methods)] The simulations employ a purely hydrodynamic treatment with imposed temperature profiles and no magnetic fields or self-consistent radiative transfer. This is a load-bearing assumption for the fragmentation results, as magnetic fields can provide additional pressure support or drive turbulence that suppresses gravitational instability, and the cooling rates critical for fragmentation depend on full radiative transfer rather than the simplified heating used here.
[§4 (Results and Discussion)] While the initial formation of protoplanets is reported, including their masses and the higher fraction in the planetary-mass regime for realistic circumbinary discs, there is no long-term evolution shown to confirm that these clumps remain bound, avoid disruption, and accrete sufficient mass to become gas giants. The noted ejections indicate dynamical interactions, but survival as planets is assumed rather than demonstrated.

minor comments (2)

[Abstract] The abstract reports averaged results with uncertainties (e.g., 9±0.9), but the main text should explicitly state the number of independent simulations performed and any tests for numerical convergence or resolution dependence to support the statistical claims.
[Throughout] Some notation for protoplanet masses and orbital radii could be clarified with explicit definitions or references to standard conventions in the field.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive review and positive assessment of the significance of our work. We address each major comment below, indicating where revisions have been made to the manuscript.

read point-by-point responses

Referee: The simulations employ a purely hydrodynamic treatment with imposed temperature profiles and no magnetic fields or self-consistent radiative transfer. This is a load-bearing assumption for the fragmentation results, as magnetic fields can provide additional pressure support or drive turbulence that suppresses gravitational instability, and the cooling rates critical for fragmentation depend on full radiative transfer rather than the simplified heating used here.

Authors: We acknowledge that the purely hydrodynamic treatment with imposed temperature profiles is a simplification that omits magnetic fields and self-consistent radiative transfer. This choice follows standard practice in gravitational instability studies to isolate the effects of binary heating and enable direct comparisons across our three disc models. We will add a new subsection to the discussion explicitly addressing these limitations, including the potential stabilizing role of magnetic fields and the approximate nature of the cooling, with supporting references to relevant MHD and radiative transfer studies. revision: yes
Referee: While the initial formation of protoplanets is reported, including their masses and the higher fraction in the planetary-mass regime for realistic circumbinary discs, there is no long-term evolution shown to confirm that these clumps remain bound, avoid disruption, and accrete sufficient mass to become gas giants. The noted ejections indicate dynamical interactions, but survival as planets is assumed rather than demonstrated.

Authors: Our simulations are focused on the fragmentation epoch and the immediate post-formation properties of the protoplanets, including their initial masses, numbers, and orbital evolution up to the point where the reported orbital distribution (peaking near 100 AU) and ejections are measured. We agree that demonstrating long-term survival, avoidance of disruption, and growth to full gas-giant masses would require substantially longer runs or a hybrid hydrodynamical plus N-body approach, which lies outside the present scope. We will revise the results and discussion sections to clarify the simulation timescales, state that the reported outcomes represent the fragmentation phase, and discuss the implications of the higher ejection rates for free-floating objects while noting the need for future work on long-term evolution. revision: partial

Circularity Check

0 steps flagged

No circularity: purely numerical hydrodynamic experiments with no analytical derivation or self-referential reduction.

full rationale

The paper reports outcomes from a set of 3D SPH simulations under three classes of initial disc conditions (circumstellar, fiducial circumbinary with imposed temperature profile, and realistic circumbinary with per-star heating). All reported quantities—fragmentation timing, protoplanet counts (9±0.9, 6.5±0.6, 7.5±0.8), mass distributions, orbital radii peaking near 100 AU, and ejection statistics—are direct numerical outputs rather than quantities derived from equations that loop back to fitted parameters or prior self-citations. No self-definitional relations, fitted-input predictions, or ansatz smuggling appear in the presented chain; the viability conclusion follows from the simulation results under the stated hydrodynamic, barotropic, and cooling assumptions. This is the expected non-circular finding for a parameter-study simulation paper.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

The central claim rests on the validity of the hydrodynamic setup and the interpretation that simulated fragments correspond to observed planets; no new physical entities are introduced.

free parameters (2)

initial disc temperature profile
Chosen separately for fiducial and realistic heating cases to enable comparison
binary separation
Varied across runs to test wider versus closer binaries

axioms (2)

domain assumption Discs are set to be marginally unstable to gravitational instability
This is the explicit condition chosen to trigger fragmentation in the simulations
standard math Standard hydrodynamic equations plus self-gravity govern the evolution
Core assumption of all such disc-fragmentation simulations

pith-pipeline@v0.9.0 · 5618 in / 1348 out tokens · 117148 ms · 2026-05-08T01:44:47.331505+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

2 extracted references · 2 canonical work pages

[1]

, Lubow S

Artymowicz P. , Lubow S. H., 1994, ApJ , 421, 651 Betti S. K. et al., 2022, ApJ , 935, L18 Bodenheimer P. , Pollack J. B., 1986, Icarus , 67, 391 Boley A. C. , 2009, ApJ , 695, L53 Boley A. C. , Hayfield T., Mayer L., Durisen R. H., 2010, Icarus , 207, 509 ´Calovi ´c A. , Nayakshin S., Casewell S., Miret-Roig N., 2026, MNRAS , 545, staf2097 Chen C. , Armi...

work page 1994
[2]

Published by Oxford University Press on behalf of R oy al Astronomical Society. This is an Open Access article distributed under the t erms of the Creativ e Commons A t tribution License ( https://creativecommons.org/licenses/by/4.0/ ), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly c...

work page 2026

[1] [1]

, Lubow S

Artymowicz P. , Lubow S. H., 1994, ApJ , 421, 651 Betti S. K. et al., 2022, ApJ , 935, L18 Bodenheimer P. , Pollack J. B., 1986, Icarus , 67, 391 Boley A. C. , 2009, ApJ , 695, L53 Boley A. C. , Hayfield T., Mayer L., Durisen R. H., 2010, Icarus , 207, 509 ´Calovi ´c A. , Nayakshin S., Casewell S., Miret-Roig N., 2026, MNRAS , 545, staf2097 Chen C. , Armi...

work page 1994

[2] [2]

Published by Oxford University Press on behalf of R oy al Astronomical Society. This is an Open Access article distributed under the t erms of the Creativ e Commons A t tribution License ( https://creativecommons.org/licenses/by/4.0/ ), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly c...

work page 2026