{"id":"bf226cce-90c3-4792-81c5-0a91e269bd06","arxiv_id":"2501.12986","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Internal Lyman-Werner feedback from a first star inside an atomic-cooling halo reduces accretion onto neighboring protostars and cannot produce a supermassive star, even when molecular hydrogen cooling is artificially removed.","lead":"Using hydrodynamical simulations, the authors show that ultraviolet radiation from the first star forming in a primordial atomic-cooling halo suppresses, rather than boosts, the growth of a second star in the same halo. This closes off a proposed mechanism for forming supermassive black hole seeds and sharpens the conditions required for this channel.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Generality of the negative result hinges on a single, already-collapsed halo; the no-cooling outcome may reflect Halo C's shallow inner potential rather than a universal ACH property.","rationale":"The reader's weakest_assumption identifies the single-halo generality problem, and I agree that it is the most load-bearing issue. The internal simulation logic is sound for Halo C: adding LW reduces accretion, and even complete H2 removal does not produce atomic-cooling collapse. But the paper's abstract and conclusions make a broader statement about ACHs, and the bridge from one halo to the population is a qualitative 'conservative' argument in Sec 6. The concern is not simply that another halo might differ; it is that the physical mechanism responsible for the negative result—the absence of infall energy to heat already-collapsed, pressure-supported gas in a shallow inner potential—is tied to the halo's late, H2-cooled assembly. Standard DCBH scenarios rely on preventing H2 cooling before collapse so that the gas remains near the virial temperature during infall. In Halo C, that window has already closed. The no-cooling run therefore tests a regime (post-collapse H2 removal) that may not correspond to the internal-LW scenario in typical ACHs, where the first star's radiation could act earlier, during the collapse of the second clump. This is a correctness risk, not a mere disagreement with consensus: the paper's own no-cooling run demonstrates that the outcome is controlled by the halo's dynamical state. The concrete test—repeating the runs on a second halo or an idealized cuspy, pre-collapse halo—would directly determine whether the negative conclusion is a general property or an artifact of Halo C. Since the reader already made the generality issue the basis for a conditional verdict, I do not recommend changing the verdict, but I emphasize that the condition is essential and currently unmet.","tokens_in":22281,"tokens_out":10819,"duration_ms":119055,"concrete_test":"Repeat the no-cooling and short-delay simulations on a second atomic-cooling halo from Kulkarni et al. (2019), e.g. Halo A or B, which forms at higher redshift and is not yet fully collapsed, using the same ENZO sink and LW prescriptions. Additionally, run an idealized spherical collapse in an NFW potential with concentration c~10 and virial temperature ~10^4 K, starting at the virial radius with H2 cooling disabled ab initio. If in either test the central gas temperature reaches ~8000 K or a sink particle sustains accretion >0.01 M_sun/yr for >10^5 yr, the paper's conclusion that internal LW feedback cannot form SMSs in ACHs does not generalize to those conditions.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central negative claim is that internal LW feedback inside atomic-cooling halos (ACHs) is unlikely to facilitate SMS formation. The strongest evidence is the no-cooling run (Sec 5.2): even with a uniform 10^10 J21 LW flux that fully dissociates H2, the gas in Halo C only reaches ~1200 K, well below the atomic-cooling threshold, and the sink stalls at 23 M_sun. However, this result is obtained in a single halo whose gas has already collapsed via H2 cooling and settled into a pressure-supported core in a shallow inner dark-matter potential (Figs. 13-15). The virial temperature at small radii is only ~1300 K, whereas gas near the virial radius (~100 pc) has T~8000 K but a dynamical time of ~50 Myr. This is precisely why the no-cooling run fails: there is no remaining infall energy to heat the core. In a more typical higher-redshift ACH, which is less evolved and has a cuspier NFW-like density profile, internal LW may dissociate H2 before or during the initial collapse; the infalling gas can then shock to ~10^4 K and maintain the high accretion rates needed for an SMS. The paper's Sec 6 argument that Halo C is 'conservative' because it is massive and late-forming conflates total halo mass with inner potential-well depth and does not address the evolutionary-state issue. Thus the generality of the headline conclusion rests on an untested assumption about halo structure and assembly timing.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper uses ENZO adaptive-mesh-refinement simulations of Halo C, an atomic-cooling halo at z~6.56 taken from Kulkarni et al. (2019), to test whether Lyman-Werner (LW) radiation from a first Population III star can dissociate H2 around a nearby protostellar clump and thereby promote the formation of a second, supermassive star. Four runs are compared: a background-only run, two runs with an additional internal LW flux of 1e4 J21 added after either a short (20 kyr) or long (250 kyr) delay, and an extreme no-cooling run with a uniform 1e10 J21 flux that fully dissociates H2. The simulations show that the additional LW flux reduces the accretion rates and final sink masses in both clumps relative to the background-only run (e.g., sink B1 reaches ~2000 Msun in the background-only run but only ~250 Msun and ~1000 Msun in the short- and long-delay runs). The no-cooling run raises the gas temperature to only ~1200 K, well below the atomic-cooling threshold, and the sink stalls at ~23 Msun. The paper concludes that internal LW feedback inside atomic-cooling halos is unlikely to facilitate the formation of supermassive stars or massive BH seeds.","tokens_in":22530,"tokens_out":9887,"duration_ms":110150,"significance":"If the conclusion holds, this is a useful negative result: it challenges an appealing mechanism for producing direct-collapse black hole seeds through sequential Population III star formation within a single atomic-cooling halo, complementing earlier work on external LW backgrounds. The paper's strengths are the clean controlled comparison of four runs, the explicit no-cooling diagnostic, the identification of a plausible physical explanation (dense-core H2 survives while outer gas heats and expands, and the shallow inner potential cannot raise the temperature to ~1e4 K), and the quantitative discussion of cooling vs. dynamical times. The simulations are internally consistent, and the negative trend is robust across two clumps and two flux turn-on delays. The main weakness is that the population-level conclusion is extrapolated from a single, already-collapsed halo, as discussed below.","major_comments":[{"comment":"The generalization from Halo C to the broader ACH population is not supported by the simulations. The text states that Halo C is conservative because it is massive and late-forming ('We therefore regard our result as conservative, and expect our conclusions to hold for higher-redshift ACHs'), but the failure mode identified in the no-cooling run (Sec 5.2, Figs 12-13, 15) is set by the shallow inner potential at r < 1 pc, where the gas has already collapsed and is pressure-supported at T~1200 K. This is an evolutionary-state property, not a total-halo-mass property. A higher-redshift ACH with a cuspier inner density profile, or one in which H2 is dissociated before/during the initial collapse, could have deeper inner potential and residual infall energy to shock the gas to ~1e4 K. To support the headline conclusion, the authors should either test additional halos or idealized density-profile variations, or explicitly restrict the conclusion to already-collapsed massive halos like Halo C.","section":"Section 6"},{"comment":"The no-cooling experiment is not a clean test of whether internal LW feedback can prevent H2 cooling in a typical ACH, because the extreme 1e10 J21 flux is switched on after Halo C's gas has already undergone H2-cooling collapse and assembled a dense, quasi-static core. The authors themselves note that gas at ~100 pc has T~8000 K but a dynamical time of ~50 Myr; the simulation shows quasi-hydrostatic settling at ~1200 K, demonstrating that this particular core cannot be re-heated by residual infall within the simulated time. This does not exclude the possibility that in a less-evolved halo, dissociating H2 before the initial collapse would allow infalling gas to shock to ~1e4 K and maintain high accretion. A simulation with the extreme flux applied before/at the onset of H2 cooling, or an idealized halo with a deeper inner potential, is needed to separate these two interpretations.","section":"Section 5.2"}],"minor_comments":[{"comment":"The hydrogen number density normalization in the unnumbered equation is printed as n0 = 1.04e-6 cm^-3; this appears to be a typo for 1.04e6 cm^-3, since the electron-density fit and the protostellar-core densities in Figs 5 and 9 imply values near 1e6 cm^-3 at r < 0.1 pc. As written, the analytic dissociation-timescale estimate of ~150 yr is computed for a medium many orders of magnitude less dense than the simulated core and should be corrected.","section":"Section 3"},{"comment":"The first paragraph of Section 3 states that the second protostar is forming ~20 kpc away; this should be ~20 pc, consistent with the rest of the paper and with the quoted 1e4 J21 flux normalization.","section":"Section 3"},{"comment":"The text refers to 'snapshot #50' when discussing the central XH2 values in the short- and long-delay runs, but Figure 5 shows columns for snapshots #17, #35, and #60 only; please clarify whether the values refer to snapshot #60 or to an additional plotted/unplotted snapshot.","section":"Section 4.2.3"},{"comment":"The no-cooling run is listed in Table 1 with t_delay = 0, but Section 5.2 says the 1e10 J21 flux is added 'beginning at dynamic collapse (z = 6.5618)', which is slightly later than the simulation restart at z = 6.5648; the time normalization should be made consistent.","section":"Table 1 / Section 5.2"}],"recommendation":"major_revision","confidential_remarks":"The paper is candid about its single-halo setup and states its limitations in Sec 6, but the abstract and title claim a general conclusion. If the authors are unwilling to add another halo simulation, the editor may consider asking them to reframe the paper as a case study of Halo C; alternatively, an idealized test of the potential-well dependence (e.g., varying the inner density profile or redshift) would substantially strengthen the population-level claim. The simulation methodology is sound and the negative trend is clear in the runs presented; the main risk is over-generalization from one already-collapsed halo."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Jim,\n\nQuick take on arXiv:2501.12986 (Sullivan et al.). The headline is a clean negative result: in Halo C from Kulkarni et al. 2019, adding a large internal Lyman-Werner flux from a first protostar reduces the accretion rate and final mass of a neighboring protostar, rather than boosting it into the supermassive-star regime. The four runs (background-only, short-delay, long-delay, and an extreme no-cooling run with J_LW = 1e10 J21) are consistent: lower accretion, lower stellar masses, and the no-cooling run stalls at 23 solar masses because the core only heats to ~1200 K, well below the atomic cooling threshold. The physical explanation is coherent: H2 survives in the dense core, while outer lower-density gas is heated and expands, suppressing infall.\n\nWhat's new: this is the first high-resolution simulation test of whether a first star's internal LW radiation in the same atomic-cooling halo can promote a second supermassive star. Dunn et al. 2018 included internal LW but could not resolve the cores. The paper also does nice work with analytical estimates of the H2 dissociation front versus the hydrogen ionization front, and the cooling-time/dynamical-time comparison is a useful diagnostic. The no-cooling run is a clever control: even with H2 fully destroyed, the shallow inner potential of this halo prevents the gas from reaching ~1e4 K.\n\nThe main soft spot is generality. The conclusion is phrased broadly (\"internal LW feedback inside atomic-cooling halos is unlikely to facilitate SMS formation\"), but every run uses a single halo that has already collapsed and settled into a pressure-supported core. The no-cooling failure is partly an evolutionary-state effect: there is no remaining infall energy to shock-heat the gas. The Sec 6 argument that Halo C is conservative because it is massive and late-forming is asserted, not demonstrated. A higher-redshift ACH with a cuspier inner profile and gas still infalling could behave differently, especially if the LW flux dissociates H2 before the initial collapse. The authors acknowledge some of this, but the main conclusion goes a bit further than the evidence. A second halo or an idealized variation with a deeper potential well would have made the claim much stronger.\n\nMinor: there is no public code, data, or parameter file deposit, which limits reproducibility. The uniform LW treatment is idealized but clearly stated.\n\nVerdict: this deserves a serious referee. The result is solid for the halo they studied, and the suppression mechanism is physically plausible. I would cite it if I worked on direct-collapse black hole seeding, and I would bring it to our reading group. The revision should either soften the population-level claim or test another halo.\n\n-- [Your name]","headline":"Clean, well-executed negative result: internal Lyman-Werner feedback suppresses accretion in the one halo they simulate, but the single-halo generalization to all atomic-cooling halos is the real soft spot.","tokens_in":23136,"tokens_out":2984,"would_cite":true,"duration_ms":34948,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Adding a nearby star's Lyman-Werner radiation suppresses, rather than enables, supermassive-star formation in atomic-cooling halos.","keywords":["Population III stars","supermassive star formation","direct-collapse black hole seeds","Lyman-Werner feedback","atomic-cooling halos","molecular hydrogen dissociation","protostellar accretion","cosmological zoom-in simulations"],"falsifier":"Run the same experiment on a grid of halos with deeper potential wells and clump separations from $2$ to $40$ pc, and check whether any configuration sustains accretion above $0.01 M_\\odot$ yr$^{-1}$ for longer than $10^5$ yr after H$_2$ is fully dissociated; the paper's estimate implies the threshold lies near $6$ pc separation or a first star more massive than about $150 M_\\odot$, where the H$_2$ survival density exceeds $10^6$ cm$^{-3}$. A positive result would show the negative conclusion is specific to Halo C, not general.","tokens_in":121,"feed_emoji":"🌌","tokens_out":13721,"duration_ms":247199,"temperature":0.7,"pith_summary":"This paper asks whether the Lyman-Werner (LW) ultraviolet light emitted by the first star born inside a protogalactic gas halo can strip molecular hydrogen from a neighboring clump of collapsing gas, letting that second clump collapse into a supermassive star, a proposed seed for supermassive black holes. The authors test the idea with hydrodynamical simulations of a massive halo containing two adjacent collapsing clumps about $20$ pc apart, adding an internal LW flux of $10^4 J_{21}$ ($J_{21}=10^{-21}$ erg s$^{-1}$ cm$^{-2}$ Hz$^{-1}$ sr$^{-1}$ is the standard LW intensity unit) either $20{,}000$ or $250{,}000$ years after the first protostar forms. In every case the extra radiation lowers the protostars' accretion rates and final masses rather than raising them: the second clump's star ends at $250$--$1000 M_\\odot$ versus about $2000 M_\\odot$ without the extra flux, and even a run with molecular hydrogen fully destroyed stalls at about $1200$ K, far below the about $10^4$ K atomic-cooling threshold needed for the supermassive route. The paper concludes that this internal LW feedback inside atomic-cooling halos is unlikely to form supermassive stars or massive black hole seeds.","feed_headline":"Nearby star's UV light stalls protostar growth, simulations show","feed_subtitle":"Internal Lyman-Werner radiation cuts accretion and leaves young stars at hundreds of solar masses, not supermassive.","key_machinery":"The machinery is a set of four ENZO adaptive-mesh-refinement simulations of a single $3.2\\times 10^8 M_\\odot$ atomic-cooling halo, Halo C, in which two protostellar clumps (A and B, about $20$ pc apart) collapse nearly simultaneously; protostars are represented by sink particles, and the runs differ only in the added internal Lyman-Werner flux and its turn-on time (none; $10^4 J_{21}$ at $20{,}000$ yr; $10^4 J_{21}$ at $250{,}000$ yr; and $10^{10} J_{21}$ with all H$_2$ cooling removed). Two analytic tools carry the interpretation: a front-propagation balance (equation 2) that compares H$_2$ dissociation by LW photons with H$_2$ re-formation through the H$^-$ bottleneck reaction, giving an H$_2$ `survival density' above which molecules survive, and a virial-temperature profile that shows the halo's shallow inner potential well can only support gas at about $1200$ K rather than the about $10^4$ K needed for atomic cooling. Together they show why the added LW flux heats the low-density envelope but not the dense core, and why even full H$_2$ dissociation cannot drive the collapse to supermassive-star conditions within the halo's dynamical time.","core_discovery":"Using a suite of ENZO adaptive-mesh-refinement simulations of a $3.2\\times 10^8 M_\\odot$ atomic-cooling halo with two nearby collapsing cores (clumps A and B), the authors show that adding an internal Lyman-Werner flux after the first protostar forms reduces the gas density and H$_2$ abundance in the accretion regions, raises the average gas temperature to several hundred kelvin but not above about $1000$ K, and cuts the accretion rate below the critical $0.01$--$0.04 M_\\odot$ yr$^{-1}$ needed to avoid main-sequence contraction. The short-delay and long-delay runs produce second-clump protostars of about $250$ and $1000 M_\\odot$, compared with about $2000 M_\\odot$ in the background-only run; the first clump falls from about $7000$ to $4600$--$6000 M_\\odot$. In an extreme `no-cooling' run with a $10^{10} J_{21}$ flux that fully dissociates H$_2$, the gas heats to about $1200$ K, reaches quasi-hydrostatic equilibrium near the local virial temperature, and accretion shuts off, leaving a $23 M_\\odot$ star. Because the already-collapsed core has no reservoir of infall energy left to convert into heat, the dynamical time for about $10^4$ K gas at about $100$ pc is about $50$ Myr, far longer than the protostar's Kelvin-Helmholtz contraction time of $10^4$--$10^5$ yr. The paper therefore concludes that adjacent protostellar cores do not help form a supermassive star and that this route is unlikely to produce massive black hole seeds.","pith_inferences":["The negative result is demonstrated for one halo; scaling the clump separation down to about $6$ pc or raising the first star's mass would push the H$_2$ survival density above about $10^6$ cm$^{-3}$, and the paper does not simulate such configurations, so the conclusion is strongest for roughly $20$ pc separations and roughly $150 M_\\odot$ first stars.","The no-cooling run isolates a timing bottleneck that any in-situ warm-atomic-gas mechanism must beat: gas at the atomic-cooling threshold about $100$ pc away has a $50$--Myr dynamical time, while the protostar contracts to the main sequence in $10^4$--$10^5$ yr; this suggests the limiting factor is the halo's potential well, not the LW flux alone.","If typical higher-redshift atomic-cooling halos have cuspier inner density profiles or more tightly packed cores, the internal-LW route could be reactivated; a systematic parameter sweep over halo mass, clump separation, and first-star mass would test the paper's conservative-generalization claim.","The paper's estimate that a roughly $150 M_\\odot$ star at about $20$ pc gives $10^4 J_{21}$ implies that only a modest boost in LW luminosity or proximity is needed to cross the H$_2$ survival threshold, so the mechanism is not ruled out everywhere, only in the simulated configuration."],"forward_implications":["Internal LW feedback from a first protostar suppresses rather than promotes growth of a nearby second protostar, lowering final stellar masses by factors of 2--8 in the simulated halo.","Even complete H$_2$ dissociation does not enable a supermassive star in this halo: the gas settles at about $1200$ K and accretion stops, because the dynamical time at the atomic-cooling radius exceeds the protostar's contraction time.","The accretion rates in all runs stay below the $0.01$--$0.04 M_\\odot$ yr$^{-1}$ threshold, so the resulting protostars join the main sequence as ordinary-mass Population III stars rather than growing to $10^5$--$10^6 M_\\odot$.","Consequently, sequential star formation within individual atomic-cooling halos is unlikely to be a major source of the massive black hole seeds that later grow into supermassive black holes.","A scenario in which a normal Population III star forms inside a region of warm gas that collapses tens of Myr later remains possible, and the paper identifies it as a follow-up study."],"supporting_citations":[{"why":"It supplies Halo C, its zoom-in initial conditions, the sink-particle method, and the ionizing and LW background used as the base state for all four runs.","marker":"Kulkarni et al. 2019"},{"why":"It provides the H$^-$ formation and H$_2$ dissociation rate coefficients used in the analytic front-propagation and H$_2$-survival-density estimates.","marker":"Shang et al. 2010"},{"why":"It defines the critical LW flux required to suppress molecular cooling and establishes the warm atomically-cooled collapse route to supermassive stars.","marker":"Omukai 2001"},{"why":"It supplies the critical accretion rate near $0.01$--$0.04 M_\\odot$ yr$^{-1}$ above which a protostar can avoid joining the main sequence and grow supermassive.","marker":"Hosokawa et al. 2012"},{"why":"It refines the supermassive-star accretion threshold and is used to judge whether the simulated protostars could still grow massive.","marker":"Hosokawa et al. 2013"},{"why":"It provides updated critical-flux estimates for atomic-cooling halos that motivate the choice of $10^4 J_{21}$ as the internal LW flux.","marker":"Wolcott-Green & Haiman 2019"},{"why":"It is the review that frames supermassive-star and direct-collapse black hole formation and the role of LW radiation in suppressing molecular cooling.","marker":"Inayoshi et al. 2020"},{"why":"It supplies the hydrogen self-shielding prescription applied in the simulations.","marker":"Rahmati et al. 2013"}],"fun_headline_variants":["Internal Lyman-Werner feedback does not enable supermassive star formation","Nearby star's UV radiation cuts accretion, no supermassive star forms","In atomic-cooling halo, internal LW feedback prevents SMS seeds","Simulations: internal UV feedback reduces accretion, stops SMS","Second protostar's growth stunted by first star's Lyman-Werner flux"],"cache_read_input_tokens":25088,"weakest_assumption_plain":"The negative result rests on a single very massive, late-forming halo (which the authors argue is the easy case for making a supermassive star); if more typical halos have steeper-density cores, more closely spaced clumps, or heavier first stars, the internal ultraviolet flux and the depth of the gas's potential well would both be larger, and the outcome could change.","fun_headline_variants_meta":{"raw":{"variants":["Internal Lyman-Werner feedback does not enable supermassive star formation","Nearby star's UV radiation cuts accretion, no supermassive star forms","In atomic-cooling halo, internal LW feedback prevents SMS seeds","Simulations: internal UV feedback reduces accretion, stops SMS","Second protostar's growth stunted by first star's Lyman-Werner flux"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000306,"raw_usage":{"total_tokens":1900,"prompt_tokens":1235,"completion_tokens":665,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":851,"completion_tokens_details":{"reasoning_tokens":570}},"tokens_in":851,"tokens_out":665,"duration_ms":7122,"temperature":1.0,"reasoning_tokens":570,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T16:34:01.637920+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same experiment on a grid of halos with deeper potential wells and clump separations from $2$ to $40$ pc, and check whether any configuration sustains accretion above $0.01 M_\\odot$ yr$^{-1}$ for longer than $10^5$ yr after H$_2$ is fully dissociated; the paper's estimate implies the threshold lies near $6$ pc separation or a first star more massive than about $150 M_\\odot$, where the H$_2$ survival density exceeds $10^6$ cm$^{-3}$. A positive result would show the negative conclusion is specific to Halo C, not general.","supporting_citations":[],"review_version":1}