{"id":"76be0c0a-0535-4a19-8d19-de40a0c7381f","arxiv_id":"2509.02165","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":5,"one_line_summary":"Primordial black holes from the QCD phase transition could accrete dark matter and gas to form globular clusters before galaxies assemble.","lead":"This paper proposes that globular clusters formed before galaxies, seeded by 1000-solar-mass primordial black holes created during the QCD phase transition. The authors run dark-matter-only simulations of black hole accretion and argue the resulting dark clusters pull in gas and ignite star formation by redshift 100.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The gas-to-star transition is assumed, not simulated; without it the PBH dark cluster does not yield a GC by z~100.","rationale":"The reader's weakest_assumption identifies exactly the same load-bearing gap: the unmodeled conversion of a dark-matter PBH cluster into a luminous globular cluster via baryonic infall, cooling, and star formation. The paper explicitly disclaims further simulation in Section 3, and the counter-evidence section does not mitigate this omission—it only places upper limits on nuclear PBHs based on kinematics and X-rays. The concern is not a matter of disagreement with consensus; it is an internal logical gap between the simulated quantity (dark matter accretion) and the claimed product (a GC). An honest non-finding is not appropriate here because the paper's own words flag the missing step, and the central claim depends on it. I do not see a more fundamental, independent weakness that would change the verdict. The numerical simplifications (simple leapfrog-like integration, no convergence tests) are concerning but secondary; even if the accretion rates are accurate, the gas physics could still fail. The concrete hydrodynamics test proposed would settle the issue. Since the reader already rated the paper CONDITIONAL with high correctness risk—which I agree with—no verdict adjustment is needed beyond UNCHANGED.","tokens_in":8298,"tokens_out":6831,"duration_ms":80565,"concrete_test":"Run a zoom-in 3D radiation-hydrodynamics simulation starting from the dark-matter PBH distribution of Run 82 at z~100, with primordial gas chemistry (H2 cooling), the CMB background, and star formation/feedback, resolving the Jeans mass. Verify whether a self-bound stellar cluster of ~1e5 M_sun forms within a few free-fall times. If not, the central claim fails. An analytic first check: compute the cooling-to-free-fall time ratio in the potential well, including supersonic gas–dark-matter relative velocities.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires baryons to fall into the PBH potential well, cool, and fragment into a globular cluster by z~100. Section 3 explicitly states 'We do not follow this with further simulation, but outline what we expect to happen.' This is the weakest link: the N-body runs are dark-matter-only, and the baryonic mass estimate is based on the most favorable run (Run 82: final PBH mass 2.5e4 M_sun), while other runs show minimal growth (e.g., Run 76, 78, 85). Moreover, Section 2 acknowledges that gas is 'significantly hotter than the dark matter' and has a 'significant net relative velocity on small scales,' yet the star-formation outline assumes gas quietly accumulates and reaches T<116K and densities ~1e-6 g/cm^3. No cooling mechanism (e.g., H2 in pristine gas) is demonstrated to overcome these initial conditions, nor is the Jeans analysis connected to a realistic fragmentation criterion. If the gas does not collapse and form stars as assumed, the proposed mechanism does not produce luminous GCs, even if the PBH accretion runs are correct.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes that primordial black holes formed at the QCD phase transition, with a log-flat IMF up to ~1000 Msun, serve as gravitational seeds for globular clusters. N-body simulations of PBH accretion in an expanding universe show that a dominant 'nucleus' can grow in some configurations (e.g., Run 82 from 5000 to 25000 Msun). The authors assume the accompanying baryons (15% of the final PBH mass) fall into the dark matter potential well, cool, and fragment into a luminous globular cluster by z~100, and they argue that old GCs would then harbor nuclear PBHs. The paper surveys X-ray, kinematic, and CMB constraints and concludes that current observations do not rule out the scenario.","tokens_in":8618,"tokens_out":7922,"duration_ms":76263,"significance":"If the gas-to-star step could be established, this would be an important and falsifiable formation channel for globular clusters and would connect PBHs to an observable population. The paper's strengths are its relatively simple, reproducible N-body code (available at github/jrmould/darkmatter), its explicit engagement with kinematic and X-ray counter-evidence, and its specific predictions about nuclear PBHs and a possible gravitational wave background. The main weakness is that the decisive baryonic collapse and fragmentation step is not modeled; the paper itself says it 'outline[s] what we expect to happen.' The quantitative baryon budget from the fiducial run is also far below a typical GC mass. The scenario is conditional, not established.","major_comments":[{"comment":"The central claim that PBH dark clusters become luminous GCs rests on an unmodeled baryonic transition. Section 3 states 'We do not follow this with further simulation, but outline what we expect to happen,' and Section 2 notes that gas is 'significantly hotter than the dark matter' and has 'significant net relative velocity on small scales.' No cooling mechanism (e.g., H2) is demonstrated to bring pristine gas to T<116 K, and the Jeans-length estimate is not tied to a fragmentation criterion. The simulations by themselves establish only dark clusters; the GC claim requires a hydrodynamical treatment or at least a quantitative cooling/condensation argument.","section":"Section 3"},{"comment":"The baryon mass budget is orders of magnitude too small. The paper takes Run 82's final PBH mass of 25000 Msun and assigns 'accompanying baryons Omega_b/Omega_m = 15% of that mass,' i.e. ~3750 Msun. A globular cluster contains ~1e5-1e6 Msun in stars. The claim that 'GC sized gas clouds need little encouragement to collapse' does not follow unless the dark cluster subsequently accretes ~1e5 Msun of baryons from the surrounding medium; that accretion is not included in the runs or quantified.","section":"Section 3 / Run 82"},{"comment":"The fiducial growth run is not representative. Table 1 shows substantial growth only in Runs 80, 81, 82, and 84; Runs 75, 76, 78, 79, 80a, 83, and 85 grow by factors below ~1.5 or not at all. The text adopts Run 82 for the baryon scaling and figures emphasize its growth. There are no error bars, no multiple realizations of the same parameters, and no convergence or resolution tests. The reader cannot assess whether the 'nucleus remains' conclusion is robust or specific to initial conditions.","section":"Table 1 / Section 2.1"},{"comment":"The kinematic test in Section 4.2 is treated too optimistically. For the archetypal metal-poor cluster M92, Kamann et al. (2014) find a 1-sigma upper limit of 980 Msun and a 3-sigma limit of 2700 Msun on a central IMBH. The model's predicted final masses range from ~10^3 to 2.5e4 Msun; thus the 3-sigma M92 limit rules out the high-growth runs, and the 1-sigma limit is at the lower end of the predicted range. The paper's response (NGC 6362 is not metal-poor) does not address M92, which is directly relevant to 'oldest globular clusters.' A quantitative likelihood statement is needed.","section":"Section 4.2"}],"minor_comments":[{"comment":"Reference list formatting: 'V olonteri', 'Senchnya', and 'Carr & Kuhnel 2021 arxiv 21100282' contain typos or incomplete fields; the text cites Harris (1986) while the reference entry is listed as Harris 1996.","section":"References"},{"comment":"The caption says 'The scale is in pixels, which are effectively AU' but the pixel-to-AU conversion is not defined; please add units or a scale bar.","section":"Figure 2"},{"comment":"The text says '100000 particle n-body runs' but Table 1 lists Runs 80 and 82 as having 150000 particles; clarify the default particle number.","section":"Section 2.1"},{"comment":"The abstract and conclusions emphasize 1000 Msun nuclei, but several simulations end at 25000 Msun. Specify whether the initial or final mass is the prediction for GC nuclei.","section":"Abstract/Conclusions"},{"comment":"The accretion formula dm/dt = pi sigma rho v is not numbered; number the equations for easier reference in the text.","section":"Section 2.1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript would be publishable if reframed as a study of PBH dark-cluster formation, but the GC claim needs either hydrodynamical modeling or an explicit, quantified accretion/cooling argument. The reliance on two related papers by the same group (Mould 2025; Mould & Batten 2025) for the IMF and PBH mass-temperature relation makes external validation difficult; I would ask the editor to ensure those works are available and independently assessed. I recommend major revision rather than rejection because the observational tests proposed are concrete and the simulations are reproducible."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this paper is a scenario sketch, not a completed model. The new piece is the specific chain: a ~1000 M_sun PBH formed at the QCD transition acts as a nucleus, gravitationally accretes smaller PBHs into a dark cluster at z~1300, and then gas supposedly falls into that potential, cools, and fragments into a globular cluster by z~100. That specific proposal, with new N-body runs and an explicit prediction of nuclear PBHs in old GCs, is genuinely not in the cited literature. The authors deserve credit for running the simulations, for being explicit that the runs are dark-matter-only, and for writing a counter-evidence section that takes the strongest observational constraints (X-ray upper limits, kinematic limits like M92 and NGC 6362) seriously. They don't oversell: the abstract says “may,” and the conclusions call the model “currently viable” with a decisive test within reach.\n\nThe soft spot is the one the stress-test note names, and it's real. Section 3 says, in effect, we don't simulate the gas, we just outline what we expect. That is the load-bearing step. If the baryons don't actually cool and collapse into stars, you don't get a luminous GC, no matter how well the PBH accretion runs work. The paper's own Section 2 notes that the gas is hotter than the dark matter and has a significant relative velocity on small scales, which makes quiet infall and cooling less automatic than the outline implies. No cooling mechanism (H2 or otherwise) is shown to bring the gas from those conditions down to the T<116 K and densities used for the Jeans argument. On top of that, the baryonic mass estimate uses Run 82, one of the most favorable runs (5000 -> 25000 M_sun), while several other runs show almost no growth (e.g., Runs 75, 78, 79). There are no error bars, no control runs, and the code is available but unversioned. These are not fatal for a scenario paper, but they mean the quantitative part is shakier than the narrative suggests.\n\nThat said, the authors are transparent about the limits. They aren't hiding the unmodeled transition; they flag it explicitly. The work is clear thinking on its own terms, the citations are appropriate, and the dependence on their earlier results (Mould 2025, Mould & Batten 2025) is a dependency, not a circularity. The paper is for people working on GC formation and PBH phenomenology who want a testable alternative to the usual protogalactic formation channels. It deserves a serious referee, with the main request being a real baryonic simulation or at least a quantified cooling and fragmentation argument before the central claim is taken as supported. I'd accept it for review and let the referee push on that step.","headline":"A speculative but honest scenario paper that links QCD-transition PBHs to globular cluster formation; the N-body growth runs are real and the counter-evidence section is candid, but the gas-to-star transition is explicitly not simulated, which is exactly the link that needs to hold.","tokens_in":9113,"tokens_out":1695,"would_cite":true,"duration_ms":20273,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["97.60.Lf","98.20.Gm"],"model":"deepseek-v4-flash","headline":"~1000-solar-mass black holes from the QCD phase transition can act as pregalactic nuclei, accreting smaller black holes into a dark cluster that draws in gas and triggers globular cluster formation by z~100, with the nucleus persisting.","keywords":["globular clusters","primordial black holes","QCD phase transition","pregalactic star formation","Bondi-Hoyle accretion","intermediate-mass black holes","dark clusters"],"falsifier":"Measure the stellar velocity dispersion in the cores of a large sample of the oldest metal-poor globular clusters: a 1000-solar-mass nucleus at 7.6 kpc would raise the dispersion to ~5.4 km/s within 2 arcseconds of the centre, and the paper notes M92's current 1σ upper limit of 980 solar masses sits right at that edge. If a survey of blue clusters at that sensitivity finds no central dark mass in the 10^3–10^4 solar-mass range, the mechanism is ruled out for those clusters; a clean complement is a deep stacked X-ray search showing no central sources where Bondi accretion predicts detectable lu","tokens_in":8133,"feed_emoji":"🕳️","tokens_out":13223,"duration_ms":117693,"temperature":0.7,"pith_summary":"This paper proposes a new answer to an old puzzle: how globular clusters — the oldest star clusters in the Universe — formed before galaxies did. The authors argue that primordial black holes created at the QCD phase transition, with masses spread up to about 1000 solar masses, would act as pregalactic nuclei: the dominant black hole accretes its smaller companions into a dense 'dark cluster' that emits no light but creates a deep potential well. Gas falling into this well would reach densities around 10^-6 grams per cubic centimeter and become Jeans-unstable, cooling and fragmenting into a luminous globular cluster by redshift ~100. Once formed, the nuclear black hole stays, continuing to feed on a fraction of the mass lost by evolved stars. The paper stops short of simulating the gas: the link between the simulated dark cluster and a real star cluster is asserted from density and Jeans-length estimates, and the authors specify kinematic and X-ray observations that could detect — or rule out — such nuclei.","feed_headline":"1000-solar-mass black holes may seed globular clusters","feed_subtitle":"QCD-era black hole nuclei could grow by accretion and trigger star formation by z=100","key_machinery":"The load-bearing object is the QCD-transition primordial black hole: a black hole of order 1000 solar masses formed when the Universe cooled through the QCD phase transition at T~220 MeV, the top of a mass distribution that spreads equal mass per decade. A dominant-mass nucleus accretes smaller PBHs at the Bondi-Hoyle rate dm/dt = πσρv, with the sound speed taken from prior work, in N-body simulations of 10^5 particles started near z=1300. What carries the argument toward a globular cluster is the 'dark cluster' — the tightly concentrated potential well (several thousand solar masses within ~10 AU) formed by the accreted PBHs. Gas drawn into this well reaches densities ~10^-6 g/cm^3, where t","core_discovery":"The paper's claim is that QCD-transition primordial black holes up to ~1000 solar masses can seed globular clusters. In gravity-only N-body runs of 100,000 particles, a dominant-mass nucleus grows by accreting smaller PBHs through Bondi-Hoyle capture — e.g., from 5000 to 25,000 solar masses by late times — producing a strongly concentrated dark cluster whose potential well deepens by an order of magnitude. The authors infer that baryons falling into this well reach densities ~10^-6 g/cm^3 with Jeans lengths of order 0.1 AU below 116 K, so star formation ignites around z~100 and a globular cluster forms around the persistent intermediate-mass black hole. Old metal-poor globular clusters are t","pith_inferences":["If the mechanism is right, the mass of the nuclear black hole should correlate with cluster age and metallicity — the oldest, most metal-poor clusters formed earliest when PBH densities were highest, so they should show the most massive nuclei; this correlation is not derived in the paper but is directly testable.","The model could turn globular clusters into indirect probes of the primordial black hole abundance: the observed incidence of nuclear dark masses in metal-poor clusters would calibrate the QCD-transition PBH fraction far below the ~1% dark-matter limit quoted in the paper, and even null results would set a competitive upper bound.","The paper's required endpoint — gas cooling to ~100 K and fragmenting by z~100 inside a dark-matter-dominated well — is precisely the regime of first-star formation; a natural extension would be a cosmological hydrodynamical simulation seeded with one of these dark clusters, turning the asserted step into a computed one."],"forward_implications":["Metal-poor (blue) globular clusters, the oldest stellar systems, should harbour nuclear black holes of order 10^3–10^4 solar masses that persist today and keep accreting a small fraction of the mass shed by evolved stars.","High-resolution kinematics of cluster cores can now test this: a 1000-solar-mass nucleus would raise the velocity dispersion to ~5.4 km/s within 2 arcseconds of the centre of a cluster at 7.6 kpc, a signal within reach of current instruments on clusters like M92.","If accretion disks form around the nuclei, the clusters should emit weakly in X-rays or the far ultraviolet — or be hidden by dust and gas; present Chandra data for 75 of 81 Milky Way clusters already constrain such emission.","Supernovae in the first stellar generation would enrich pristine gas, and gas loss from the shallow potential would yield the low metal abundances and bimodal colour distribution observed in globular cluster systems.","Stellar-mass black holes captured and merged around these nuclei during and after cluster formation could contribute to the cosmic gravitational-wave background."],"supporting_citations":[{"why":"First to consider pregalactic accretion onto black holes; supplies the accretion concept this model extends.","marker":"Carr (1981)"},{"why":"Supplies the accretion rate formula dm/dt = πσρv that sets how fast the nucleus grows.","marker":"Bondi-Hoyle (1944)"},{"why":"Provides the equation for the characteristic mass of PBHs formed at the QCD transition.","marker":"Carr & Kuhnel (2021)"},{"why":"Provides the mass distribution function that sets the assumed PBH initial mass function.","marker":"Byrnes et al. (2018)"},{"why":"Supplies the PBH mass–temperature relation and evolutionary tracks used for Figure 1 and the 1000-solar-mass ceiling.","marker":"Mould (2025)"},{"why":"Supplies the sound speed entering the Bondi-Hoyle accretion rate.","marker":"Thomas, Kopp & Skordis (2016)"},{"why":"Supports the claim that GC-sized gas clouds collapse readily around z~100 once seeded.","marker":"Peebles (1969)"},{"why":"Supplies the physical case that the QCD phase transition produces PBHs in the 1–1000 solar-mass range.","marker":"Alonso-Monsalve & Kaiser (2023)"},{"why":"The Chandra search for central black holes in 81 Milky Way GCs that currently constrains — but does not exclude — the predicted nuclei.","marker":"Su et al. (2022)"},{"why":"The compilation of black hole detections and upper limits in GCs used to check the predicted 10^3–10^4 solar-mass nuclei.","marker":"Lützgendorf et al. (2013)"}],"fun_headline_variants":["QCD black holes ignite globular clusters at z≈100","Primordial black holes seed star birth in globulars","1000-solar-mass PBHs may spark globular cluster birth","Heavy PBHs from QCD era could seed ancient globulars","Black hole seeds from the QCD transition make globulars"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"Everything rests on what happens to the gas: the paper assumes baryons falling into the simulated dark cluster cool and fragment into a luminous globular cluster by redshift 100, but it does not simulate that step — if the gas fails to collapse and make stars, the model fails even if the black-hole accretion runs are correct.","fun_headline_variants_meta":{"raw":{"variants":["QCD black holes ignite globular clusters at z≈100","Primordial black holes seed star birth in globulars","1000-solar-mass PBHs may spark globular cluster birth","Heavy PBHs from QCD era could seed ancient globulars","Black hole seeds from the QCD transition make globulars"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000709,"raw_usage":{"total_tokens":2959,"prompt_tokens":600,"completion_tokens":2359,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":344,"completion_tokens_details":{"reasoning_tokens":2271}},"tokens_in":344,"tokens_out":2359,"duration_ms":17132,"temperature":1.0,"reasoning_tokens":2271,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T11:50:23.193063+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the stellar velocity dispersion in the cores of a large sample of the oldest metal-poor globular clusters: a 1000-solar-mass nucleus at 7.6 kpc would raise the dispersion to ~5.4 km/s within 2 arcseconds of the centre, and the paper notes M92's current 1σ upper limit of 980 solar masses sits right at that edge. If a survey of blue clusters at that sensitivity finds no central dark mass in the 10^3–10^4 solar-mass range, the mechanism is ruled out for those clusters; a clean complement is a deep stacked X-ray search showing no central sources where Bondi accretion predicts detectable lu","supporting_citations":[{"cited_title":"See Mould (2025) for the code for Figure","cited_arxiv_id":null,"evidence_quote":"Supplies the PBH mass–temperature relation and evolutionary tracks used for Figure 1 and the 1000-solar-mass ceiling."}],"review_version":1}