{"id":"1006c41d-db90-4706-b7ad-f07b0455ba6c","arxiv_id":"2507.21216","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A cold Jupiter greatly increases the chance that inner super-Earth planets are ejected, and most surviving systems are stable two-planet systems.","lead":"This paper uses computer simulations to show that a distant Jupiter-like planet can fling inner super-Earths out of their star systems, creating free-floating planets. The ejected planets move slowly relative to their stars, which could help astronomers identify them in microlensing surveys.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 1000 au ejection criterion may count bound, returning planets as ejected; at 1000 au the local escape speed is only ~1.3 km/s, so the 38% FFP fraction and low-velocity distribution need an energy-based check.","rationale":"The reader's weakest assumption (perfect mergers) is real and acknowledged in Section 5, but the 1000 au ejection criterion is more fundamental because it defines what counts as an ejected/free-floating planet. The paper's own velocity histogram (Figure 7) makes the test urgent: at 1000 au the local escape speed is only 1.33 km/s for a solar-mass star, so the quoted comparison with the 1 au escape speed (40 km/s) does not establish that the removed objects are unbound. A formally bound object with apocenter beyond 1000 au would be removed by the code but is not a free-floating planet; whether it later returns depends on the subsequent evolution, which is truncated. Thus the headline 38% 'ejected' super-Earth fraction and the low-velocity FFP prediction could both be inflated by the cutoff. The perfect-merger approximation shifts the collision/ejection branching but is unlikely to erase the qualitative enhancement from the cold Jupiter; the ejection criterion could. I therefore recommend keeping the reader's CONDITIONAL verdict, pending the energy-based reclassification. This is a partial agreement: the reader noted the ejection definition in the rationale but chose perfect merger as the weakest assumption.","tokens_in":16946,"tokens_out":10494,"duration_ms":121540,"concrete_test":"Extract from each super-Earth+cold-Jupiter run the state (position, velocity, host mass) at the moment a planet is removed at 1000 au. Compute specific orbital energy epsilon = v^2/2 - GM/r. Classify as genuinely ejected only if epsilon > 0 (equivalently v > 1.33 km/s at r=1000 au); for epsilon < 0, continue integration or re-inject the body to see whether it returns and interacts. Recompute f_ej,SE and the velocity histogram using only epsilon > 0 events. If the unbound fraction is within Poisson noise of 38% and the v_inf distribution is still mostly <6 km/s, the concern does not change the conclusion; if it drops materially, the quantitative claim must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 2.2 defines ejection as the moment a planet's distance from the system barycenter exceeds 1000 au, and removes it. For a 1 Msun host, the two-body escape speed at 1000 au is sqrt(2GM/1000 au) ~ 1.33 km/s, not the 40 km/s quoted in Section 3.2. Figure 7 shows most ejected super-Earths have relative velocities in 0-6 km/s, so any object in that histogram with v < 1.33 km/s is formally bound and will return on a very eccentric orbit (semimajor axis > 1000 au) unless later scattered again. The simulation does not check, or report, how many 'ejections' satisfy a positive-energy criterion. If a substantial fraction are bound but temporarily distant, the 38% ejection fraction overstates true free-floating super-Earth production, and the 'mostly below 6 km/s' velocity distribution is partly an artifact of the removal radius. This is independent of the acknowledged perfect-merger limitation, and it bears directly on the paper's central claim that cold Jupiters create free-floating super-Earths.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses N-body simulations (REBOUND with IAS15 and Mercurius, plus a GR correction) to study instability, collisions, and ejections in systems of five super-Earths and in systems of five super-Earths plus one cold Jupiter. It reports that the first close encounter time in units of the innermost period is nearly independent of the innermost semimajor axis, whereas the half-loss time increases with it; that a cold Jupiter ejects about 38% of super-Earths at low relative velocities; and that surviving two-planet systems are mostly stable and contain inward-migrated, eccentric super-Earths. The paper connects these results to free-floating planet observations and to the Safronov number.","tokens_in":17165,"tokens_out":6126,"duration_ms":71796,"significance":"If the quantitative ejection statistics are correct, the paper makes a useful contribution: it identifies a plausible dynamical channel for free-floating super-Earths, gives a predictor for their low ejection velocities, and quantifies the role of cold Jupiters. The modeling uses standard, well-tested integrators with GR corrections, and the main statistical results are derived directly from simulations rather than from circular fitting. The paper also honestly states the perfect-merger limitation and gives a conservative stability estimate. However, the headline 38% ejection fraction and the velocity histogram are directly tied to the distance-based ejection criterion, and the lack of an energy check means the quantitative central claim is not yet established.","major_comments":[{"comment":"The ejection criterion is purely distance-based: a planet is removed when its barycentric distance exceeds 1000 au (§2.2). For a 1 M☉ host, the two-body escape speed at 1000 au is about 1.33 km/s, not the 40 km/s quoted in §3.2 as the escape speed at 1 au. Figure 7 reports most ejected super-Earths at relative velocities of 0–6 km/s, so any object in that histogram with relative speed below about 1.33 km/s is formally bound and will return on an orbit with semimajor axis greater than 1000 au unless it is subsequently scattered again. Because the manuscript does not check the sign of the total energy at removal, the reported 38% ejection fraction and the “mostly below 6 km/s” velocity distribution may significantly overstate the true production of free-floating super-Earths. I request an energy-based ejection criterion (or a tracking test for returning planets), a report of how many removals are actually unbound, and recomputed fej,SE and velocity statistics.","section":"§2.2, §3.2, Fig. 7"},{"comment":"The perfect-merger assumption is acknowledged in §5 but not tested. In the super-Earth–cold-Jupiter systems, a super-Earth can receive velocity kicks near the surface escape velocity of Jupiter during encounters with the cold Jupiter, so subsequent super-Earth–super-Earth collisions may occur at relative velocities above their mutual escape speeds; hit-and-run or partial accretion outcomes are then plausible. Such outcomes would change the number of surviving planets, the fraction of super-Earths removed by collision versus ejection, and the final two-planet fraction. Since the 38% ejection fraction and the 94% two-planet fraction are quantitative central claims, a sensitivity test with a simple alternative collision prescription, or at least a quantitative estimate of the affected encounters, is needed before these numbers can be taken as robust.","section":"§5, §2.2"}],"minor_comments":[{"comment":"The abstract says the first close encounter time “is identical” regardless of the innermost semimajor axis, but the text and Figure 3 show no clear dependence, not exact equality; I suggest “nearly independent of a1,0.”","section":"Abstract, §3.1, Fig. 3"},{"comment":"The table reports fej,SE = 38% for all three K values and gives eccentricity percentiles without uncertainties; with 180 systems per sample, Poisson errors on the ejection counts are nontrivial, and the exact equality across K should be discussed with error bars.","section":"Table 3"},{"comment":"The histogram should state the number of ejected planets included and the bin width; without these, the shape of the low-velocity peak cannot be assessed.","section":"Fig. 7"},{"comment":"The observational sample is selected by requiring eccentricity measurements, which may bias the comparison; this selection effect should be acknowledged, and the small sample size (16 single-super-Earth systems and 8 multi-super-Earth systems) should be stated as a limitation.","section":"§4.3"},{"comment":"The parameter ℓ ≈ 3 in Equation (5) is introduced as an approximate value, but no fit or derivation is shown; please clarify whether it is an empirical constant or an adjusted parameter.","section":"§4.1, Eq. (5)"}],"recommendation":"major_revision","confidential_remarks":"The 1000 au ejection-criterion issue is the decisive one. It is not a disagreement with consensus but an internal-consistency problem: the paper’s own quoted escape speeds show that the comparison in §3.2 is made at the wrong radius. If the requested energy check substantially reduces the ejection fraction, the central conclusions will need to be reframed; if it does not, the quantitative claims can stand after a clear statement of the criterion. The observational comparison in §4.3 is thin but not central to the paper’s main dynamical result."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper's real news is that a cold Jupiter in a compact system of super-Earths ejects a large fraction of those super-Earths—38% in the fiducial run—and that the ejected planets leave with low relative velocities, mostly below 6 km/s. That is a concrete, previously unquantified channel for making free-floating super-Earths, and it is worth taking seriously. The simulations are competent: REBOUND with IAS15 and Mercurius, GR included, 180 systems per bin, and they compare against earlier work cleanly. The half-loss time scaling t1/2 ~ a1,0^2.92 for systems above ~0.35 au is a nice quantitative result.\n\nThe soft spot is the ejection criterion. They define ejection as crossing 1000 au and removing the planet. The stress-test is right: the local escape speed at 1000 au from a solar-mass star is about 1.3 km/s, not 40 km/s, so any planet removed with relative speed below ~1.3 km/s is actually bound and will return on a very wide orbit. The paper doesn't report how many of the 38% satisfy a positive-energy condition. Because Figure 7 shows a large share of ejected super-Earths with velocities between 0 and 6 km/s, some of those are likely bound. That means 38% may be an overestimate, and the low-velocity tail in Figure 7 is partly an artifact of the removal radius. This is not fatal to the main story—most planets scattered by a giant planet probably do become unbound—but it needs to be checked before quoting the number. The perfect-merger assumption is acknowledged, and the discussion of hit-and-run is reasonable; that's a second-order issue compared to the energy check.\n\nAlso, the paper should stop using the 40 km/s escape speed at 1 au to explain the 0–6 km/s velocities; the relevant escape speed is at the removal radius. A short rerun counting only positive-energy orbits would settle this.\n\nI'd send this to a competent referee. The qualitative conclusion is important enough, and the simulation work is solid. Just make the ejection criterion rigorous.","headline":"A plausible new channel for free-floating super-Earths, but the 38% ejection fraction needs an energy-based check at the 1000 au removal radius.","tokens_in":17735,"tokens_out":3303,"would_cite":true,"duration_ms":38618,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["70F15","85A04"],"pacs":["95.10.Ce"],"model":"deepseek-v4-flash","headline":"In planetary systems with super-Earths plus a cold Jupiter, close encounters with the Jupiter eject about 38% of the super-Earths, mostly at speeds below 6 km/s, making dynamical instability a promising source of free-floating super-Earths.","keywords":["free-floating planets","super-Earths","cold Jupiters","planetary dynamics","orbital instability","N-body simulations","Safronov number","microlensing"],"falsifier":"Rerun the super-Earth+cold-Jupiter suite with a collision model that allows hit-and-run or partial mass loss; if the ejected fraction drops well below 38% or the surviving two-planet systems no longer favor inward, eccentric super-Earths, the central numerical claim fails. Observational check: measure the space velocities of microlensing free-floating super-Earth candidates; if most move much faster than the local stellar population, the predicted low ejection speeds are wrong.","tokens_in":16713,"feed_emoji":"🪐","tokens_out":11869,"duration_ms":118785,"temperature":0.7,"pith_summary":"This paper asks whether the dynamical instability of multi-planet systems can produce the free-floating super-Earths suggested by recent microlensing surveys. Using N-body simulations of five super-Earths, with and without an outer cold Jupiter, it argues that the cold Jupiter is the decisive agent: the ejection fraction jumps from under 2% to about 38% when the Jupiter is present, and nearly all ejected super-Earths leave with relative speeds below 6 km/s. The simulations also show that first close encounters and first planet losses follow different time scalings, and that the Safronov number, rather than planet mass alone, controls whether encounters end in collision or ejection. If these results hold, they give concrete predictions for the kinematics of free-floating super-Earths and for the orbital architecture of the systems left behind.","feed_headline":"Cold Jupiters eject 38% of sibling super-Earths","feed_subtitle":"Simulations show a distant giant turns 38 percent of inner super-Earths into low-velocity free-floating planets.","key_machinery":"The argument runs on dimensionless quantities. The orbital spacing $K$ divides the separation between neighboring planets by their mutual Hill radius and sets how quickly interactions begin. Times are measured in units of the innermost planet's orbital period $P_1$, which makes the differing scalings of first-encounter and first-loss times visible. The Safronov number $\\Theta = (M_p/M_*)(a_p/r_p)$ — the ratio of a planet's surface escape speed squared to its orbital speed squared — is the switch that decides whether a close encounter ends in collision ($\\Theta<1$) or ejection ($\\Theta>1$); the cold Jupiter's large Safronov number is why it scatters super-Earths out of the system. The machinery also includes the simulation's event rules, especially the definition of ejection at 1000 au from the system center of mass.","core_discovery":"The paper's central discovery is that a cold Jupiter can act as an ejection engine for inner super-Earths. In the reference sample with a $0.5$ au innermost super-Earth, orbital spacing $K=5$, and one Jupiter-mass planet at the outer edge, 38% of all super-Earths end up escaping the system, with 91% of those ejections happening during the early chaotic phase; in the identical super-Earth-only sample the ejection fraction is below 2%. The ejected planets have low velocities relative to their host stars, most below 6 km/s, so their Galactic motions should match the stellar population rather than form a fast-moving halo. Systems that survive typically contain one super-Earth and the cold Jupiter; the surviving super-Earth has migrated inward and gained eccentricity (median 0.21), and more than 86% of these two-planet systems are judged long-term stable by the empirical criterion used. A supporting discovery is that the time to first planet loss, measured in units of the innermost orbital period, grows roughly as $a_1^{2.1}$ with the inner semimajor axis, while the first-close-encounter time does not, so encounter times alone do not describe instability.","pith_inferences":["If the 38% ejection rate applies to real super-Earth+cold-Jupiter systems, then the observed coexistence of super-Earths with cold Jupiters implies a substantial steady production of free-floating super-Earths; the paper notes this possibility but does not quantify a galactic rate.","A testable extension: because the Safronov number grows with distance from the star and shrinks with stellar mass, super-Earths around low-mass stars with a cold Jupiter should be ejected even more readily; a scaled simulation suite could check this.","The low relative ejection speeds imply that the free-floating planets from this channel should be kinematically indistinguishable from their birth population; any future detection of fast-moving free-floating super-Earths would point to other channels such as stellar flybys or binary disruption.","The predicted inward migration and eccentricity growth of surviving super-Earths give a demographic signature: observed systems with a cold Jupiter and a tight, eccentric inner super-Earth may be the aftermath of this instability, which transit and radial-velocity surveys can search for."],"forward_implications":["Super-Earth+cold-Jupiter systems become a quantitatively significant source of free-floating super-Earths: about 38% of the super-Earths in such systems are ejected during instability.","The ejected planets' low relative speeds mean microlensing surveys should detect them with timescales and kinematics similar to those of the ambient stellar population, not as a high-velocity population.","Surviving systems are typically one super-Earth plus the cold Jupiter; their inner super-Earths should be on inward-migrated, eccentric orbits, and over 86% of the survivors should be dynamically stable on long timescales.","In super-Earth-only systems, the first-close-encounter time in units of $P_1$ is independent of orbital radius, while the first-planet-loss time grows as $a_1^{2.1}$, so instability should be described by planet-loss times rather than encounter times.","Ejection fraction is organized by the Safronov number: systems with the same Safronov number but different masses eject planets at similar rates, which explains why earlier low-Safronov-number simulations saw no ejections."],"supporting_citations":[{"why":"Supplies the observed coexistence statistics (about 30% of super-Earths have cold Jupiters) that motivate the simulated architecture.","marker":"W. Zhu & Y. Wu 2018"},{"why":"Provides the simulation methodology of in-situ scattering, the two-phase integration scheme, and the random initial orbital parameter choices used here.","marker":"K. R. Anderson et al. 2020"},{"why":"Defines the mutual-Hill-radius orbital spacing K and the first-close-encounter time scaling used as the baseline for the instability analysis.","marker":"J. Chambers et al. 1996"},{"why":"Earlier four-planet simulations whose first-loss-time results are compared and whose lack of ejections is reinterpreted through the Safronov number.","marker":"D. R. Rice et al. 2018"},{"why":"Source of the Safronov number used to organize collision-versus-ejection outcomes across different planet masses.","marker":"V. S. Safronov 1972"},{"why":"Gives the empirical two-planet stability criterion used to estimate that over 86% of surviving two-planet systems are long-term stable.","marker":"C. Petrovich 2015"},{"why":"Gives the observed free-floating-planet abundance and super-Earth mass range that motivate the ejection channel.","marker":"T. Sumi et al. 2023"},{"why":"Supplies the N-body integration package used to evolve the simulated planetary systems.","marker":"H. Rein & S. F. Liu 2012"}],"fun_headline_variants":["Cold Jupiters launch 38% of inner super-Earths into free space","Distant giant planets eject 38% of their inner counterparts","Simulations: one cold Jupiter can fling away a third of inner worlds","Cold Jupiters act as slingshots, ejecting 38% of inner super-Earths","How cold Jupiters turn inner planets into free-floating worlds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The simulations assume that every planet-planet collision results in a perfect merger of the two bodies, conserving total mass and momentum; if collisions instead bounce, fragment, or lose mass, the 38% ejection fraction and the final orbit statistics could change.","fun_headline_variants_meta":{"raw":{"variants":["Cold Jupiters launch 38% of inner super-Earths into free space","Distant giant planets eject 38% of their inner counterparts","Simulations: one cold Jupiter can fling away a third of inner worlds","Cold Jupiters act as slingshots, ejecting 38% of inner super-Earths","How cold Jupiters turn inner planets into free-floating worlds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000814,"raw_usage":{"total_tokens":3632,"prompt_tokens":1070,"completion_tokens":2562,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":686,"completion_tokens_details":{"reasoning_tokens":2458}},"tokens_in":686,"tokens_out":2562,"duration_ms":21848,"temperature":1.0,"reasoning_tokens":2458,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T13:00:00.524223+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Rerun the super-Earth+cold-Jupiter suite with a collision model that allows hit-and-run or partial mass loss; if the ejected fraction drops well below 38% or the surviving two-planet systems no longer favor inward, eccentric super-Earths, the central numerical claim fails. Observational check: measure the space velocities of microlensing free-floating super-Earth candidates; if most move much faster than the local stellar population, the predicted low ejection speeds are wrong.","supporting_citations":[],"review_version":1}