{"id":"5b375071-4dbc-4a18-9095-0667ec565a8f","arxiv_id":"2607.05509","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.5,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"GC light-element abundance patterns can be inherited at birth from chemically structured GMCs formed by O-rich inflow colliding with N-rich gas after mini-quenching, without extended in-cluster self-enrichment.","lead":"Cosmological simulations show giant molecular clouds can inherit globular-cluster-like light-element spreads and N–O anticorrelations from galactic gas flows after starbursts. This offers a birth-inheritance channel that sidesteps the mass-budget problem of classic self-enrichment models.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.5","headline":"The Na–O anticorrelation is asserted by chemical analogy rather than demonstrated; without it the inheritance claim covers only a subset of the defining GC light-element signatures.","rationale":"The Reader correctly isolates the Na gap and the unresolved long-term evolution as the weakest assumptions. The Na issue is the more immediate load-bearing concern for the chemical-inheritance claim itself: the paper already shows that dense clusters form from the selected GMCs (Fig. 4), so the dynamical-evolution caveat is secondary to whether the birth gas even carries the full set of light-element anticorrelations. The proposed passive-Na post-processing is a concrete, low-cost check that uses the same yield tables already in the model and would settle whether the analogy holds inside the very clouds the authors have identified. Because the paper is transparent about the limitation and the rest of the mechanism is well-supported by the simulations, the verdict remains CONDITIONAL; the test simply makes the condition sharper.","tokens_in":17561,"tokens_out":585,"duration_ms":5682,"concrete_test":"Post-process the existing THESAN-ZOOM snapshots by tagging passive Na yields from the same AGB tables already used for N (Karakas 2010 and successors) and recompute the Pearson r([Na/Fe],[O/Fe]) and Δ[Na/Fe] inside every chemically selected GC-like GMC. If the majority of clouds that satisfy the published N–O and Fe criteria fail r([Na/Fe],[O/Fe]) < −0.3 (or show Δ[Na/Fe] ≲ 0.3 dex), the inheritance claim for the full light-element pattern is weakened.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is that GC abundance patterns can be inherited at birth from chemically structured GMCs. The paper’s own selection and figures establish N–O anticorrelation and low Fe dispersion inside those GMCs (Fig. 3, Methods thresholds). The Na–O anticorrelation, however, is not tracked: sodium is absent from the nine-element network. The authors argue by analogy that “sodium is produced alongside nitrogen in AGB stars through the NeNa cycle” and therefore “we expect” the same processes to produce Na–O (discussion paragraph after Fig. 4). That expectation is load-bearing: if the O-rich re-accreted gas and the N-rich galactic gas do not also carry the complementary Na contrast at the moment of GMC assembly, the clouds would reproduce only part of the observed GC chemical fingerprint. Because the inheritance scenario is offered as an alternative to self-enrichment models that do reproduce Na–O, the untested Na step is the single most critical unclosed link between the simulated GMCs and the full set of GC signatures.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The manuscript uses the THESAN-ZOOM cosmological radiation-hydrodynamic suite (standard nine-element enrichment) to identify a chemically selected population of giant molecular clouds that exhibit large log(N/O) spreads, [N/Fe]–[O/Fe] anticorrelations, and low iron dispersion. These GC-like GMCs form preferentially when oxygen-rich gas ejected by a prior starburst re-accretes and collides with nitrogen-rich galactic gas left by AGB enrichment during a temporary quench; they are more massive than ordinary GMCs, appear at early times and low metallicity consistent with MW GCs, and host dense bound star clusters with surface densities comparable to observed GCs. The authors conclude that key GC light-element patterns can be inherited at birth from baryon-cycle-structured ISM rather than requiring extended in-cluster self-enrichment, thereby avoiding the mass-budget problem and exotic polluters.","tokens_in":17880,"tokens_out":1124,"duration_ms":8484,"significance":"If the inheritance channel holds, it reframes GC multiple populations as a fossil of high-redshift gas flows and stochastic star-formation histories rather than an internal cluster process. The work is grounded in self-consistent cosmological simulations with a standard yield set, supplies a concrete formation pathway (re-accretion after mini-quenching), links the chemically selected clouds to dense cluster formation, and offers falsifiable predictions (formation timescales ≲3–5 Myr; GC-like chemistry in young high-z clusters and in galaxies quenched for ≳50 Myr). These are genuine strengths that make the paper a valuable contribution even if sodium remains untracked.","major_comments":[{"comment":"The central claim is that GC abundance patterns can be inherited at birth. The paper demonstrates N–O anticorrelation and low Fe dispersion inside the selected GMCs (Fig. 3; Methods chemical cuts Δlog(N/O)>0.5, r([N/Fe],[O/Fe])<−0.3, σ[Fe/H]<0.1). Sodium is absent from the nine-element network, so the Na–O anticorrelation is asserted only by chemical analogy (“sodium is produced alongside nitrogen in AGB stars through the NeNa cycle… we expect”). Because Na–O is a defining GC signature and self-enrichment models already reproduce it, the inheritance scenario remains incomplete until the complementary Na contrast between the re-accreted O-rich and residual N-rich gas is shown (or a quantitative yield-based argument is supplied). This is the single load-bearing gap.","section":"Discussion after Fig. 4; Methods (chemical network)"},{"comment":"Long-term dynamical evolution of the dense clusters is unresolved (stellar softening lengths 138–554 cpc; Methods). Figure 4 shows that chemically selected birth gas can form bound systems with Σ_* ≳700 M_⊙ pc^{-2} and M_* ≳10^5 M_⊙, but survival to z=0, mass loss, and the final GC mass function are not demonstrated. The paper correctly flags this limitation; a clearer statement of which present-day GC properties are predicted versus which remain open is needed so that the inheritance claim is not over-read as already producing the full z=0 GC population.","section":"Methods (Star cluster identification); Discussion"}],"minor_comments":[{"comment":"The free parameters of the GC-like selection (Δlog(N/O)>0.5, Pearson r<−0.3, σ[Fe/H]<0.1) and of CloudPhinder (α_max=20) should be stated more prominently in the main text or a table, with a short sensitivity check, so readers can judge robustness without digging into Methods.","section":"Methods: GMC identification and GC-like GMC selection"},{"comment":"Extended Data Figures 5–7 are useful but the main text could briefly note that the N–O anticorrelation appears across all three resolution levels, strengthening the claim that the pattern is not a resolution artefact.","section":"Results / Extended Data"},{"comment":"The surface-density threshold Σ_*≈700 M_⊙ pc^{-2} used to split chemical behaviour in Fig. 4 is described as “defined to split”; a short justification relative to observed GC densities or prior simulation thresholds would help.","section":"Fig. 4 caption and surrounding text"},{"comment":"A few typographical issues remain (e.g., missing spaces in “Globularclusters”, “star-to-starvariations” in the abstract/preprint header). A careful proof-read of the compiled PDF is warranted.","section":"Abstract / title block"}],"recommendation":"major_revision","confidential_remarks":"The Na gap is real but fixable within the paper’s scope by a quantitative yield argument or by flagging Na–O as a prediction rather than an established result; I would not reject on that basis alone. The work is a genuine alternative pathway and fits a high-impact astrophysics journal once the inheritance claim is carefully scoped."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The new piece is clear: in THESAN-ZOOM they find a population of massive GMCs that already carry large log(N/O) spreads, N–O anticorrelations, and tight [Fe/H] at formation, assembled when O-rich re-accreted ejecta slam into AGB-enriched N-rich gas after a mini-quench. Those clouds sit at the sites of dense bound-cluster formation. That is a genuine alternative to extended in-cluster self-enrichment and it sidesteps the mass-budget problem without exotic polluters.\n\nWhat they do well is keep the chemistry standard (nine-element network, published yields) and still recover the N–O pattern plus the mass and metallicity trends that match observed GC fractions. The random-forest ranking (mass, redshift, recent SFR, net inflow) and the surface-density cut that separates anticorrelated dense clusters from looser systems are clean. Figures 2–4 and the extended examples make the mechanism easy to see. They also flag the limits themselves: no sodium, no long-term dynamical evolution, selection thresholds tuned to Type-I GC chemistry.\n\nThe soft spots are real but proportionate. Sodium is not tracked, so the Na–O anticorrelation is an expectation from the NeNa cycle rather than a demonstrated result; that is the single biggest unclosed link if the claim is “full GC light-element fingerprint.” Long-term survival and z=0 demographics are also below the resolution floor. Neither sinks the inheritance channel that is actually shown. The free parameters (spread cuts, α_max, density threshold) are selection choices, not free knobs that invent the physics.\n\nThis is for people working on multiple populations, high-z N-rich galaxies, and baryon-cycle enrichment. It is not yet a complete replacement for self-enrichment models, but it is a serious, falsifiable alternative grounded in the same simulations that already explain N-rich galaxies. I would send it to referees; the gaps are the kind that invite follow-up rather than desk rejection. Worth reading and citing if you work on GC formation or JWST chemical outliers.","headline":"Solid simulation result: standard-yield cosmological runs produce massive GMCs with GC-like N–O anticorrelations and low Fe scatter via post-starburst inflow collisions; Na is missing and z=0 survival is unresolved, but the inheritance channel is real and worth engaging.","tokens_in":18554,"tokens_out":546,"would_cite":true,"duration_ms":5373,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Globular cluster light-element patterns can be inherited at birth from chemically structured giant molecular clouds shaped by galactic gas flows, without extended in-cluster self-enrichment.","keywords":["globular clusters","giant molecular clouds","chemical abundances","nitrogen-oxygen anticorrelation","baryon cycle","high-redshift galaxies","star cluster formation","self-enrichment"],"falsifier":"Obtain ages and light-element abundances for very young (less than about 1 Myr) high-redshift GC progenitors: inheritance predicts anomalous chemistry already present and star formation ending within roughly 3–5 Myr before iron spreads appear; extended multi-generation formation with delayed pollution would favor classical self-enrichment.","tokens_in":18454,"feed_emoji":"⭐","tokens_out":932,"duration_ms":15891,"temperature":0.7,"pith_summary":"Globular clusters show large star-to-star spreads and anticorrelations in light elements such as nitrogen and oxygen, almost always at nearly constant iron. The usual explanation is self-enrichment inside the cluster over many millions of years, but those models face a severe mass-budget problem or need exotic stars. This paper uses cosmological radiation-hydrodynamic simulations with ordinary chemical yields and finds giant molecular clouds that already carry the same abundance signatures. The clouds form when oxygen-rich gas ejected in an earlier starburst falls back and collides with nitrogen-rich gas left after the galaxy briefly quenches, then become sites of dense star-cluster formation. If the picture is right, the famous GC chemistry is a birth imprint of the high-redshift baryon cycle rather than a product of long internal pollution, and the clusters themselves become fossils of early galaxy gas flows.","feed_headline":"Globular clusters inherit their chemistry at birth","feed_subtitle":"Simulations show N–O spreads arise when oxygen-rich inflows collide with nitrogen-rich gas after starbursts.","key_machinery":"Inflow-driven GMC formation after a burst-lull cycle: oxygen-rich core-collapse ejecta leave the shallow potential, AGB stars continue to enrich the remaining gas with nitrogen while the galaxy is temporarily quenched, and the returning oxygen-rich inflow collides with that nitrogen-rich gas, compressing massive clouds that mix both abundances and imprint N–O anticorrelations at fixed iron.","core_discovery":"Cosmological radiation-hydrodynamic simulations with a standard chemical enrichment model produce a population of giant molecular clouds whose internal abundances already match several defining globular-cluster signatures: large light-element spreads and nitrogen–oxygen anticorrelations at nearly constant iron. These clouds form at the restart of star formation after a starburst, where previously ejected oxygen-rich gas collides with nitrogen-rich galactic gas, and they are the sites of dense star-cluster formation. The chemical patterns of globular clusters can therefore be inherited at birth from chemically structured interstellar gas shaped by the baryon cycle.","pith_inferences":["Abundance-spread size should scale with parent GMC mass and with the surface density of the resulting bound cluster, even among still-forming high-redshift systems.","Because sodium is produced with nitrogen in AGB stars, the same collision channel should also produce the Na–O anticorrelation once sodium is tracked.","Mildly nitrogen-rich galaxies may lack GC-like clouds if their burst-lull cycles are too weak, giving a chemical diagnostic of burst strength.","Resolved long-term evolution to z=0 is still required to check whether the dense clusters survive with the correct mass function and without forced preferential stripping of one abundance group."],"forward_implications":["GC abundance patterns need not require extended in-cluster star formation or exotic polluters.","Highly nitrogen-rich high-redshift galaxies are likely sites of ongoing GC formation via the same baryon-cycle process.","GCs form on short timescales (about 3–5 Myr) and avoid the mass-budget problem because both abundance groups form together from already mixed gas.","GCs become a fossil record of chemical enrichment and gas flows in high-redshift galaxies.","Some nitrogen-rich field stars may form outside clusters from diffuse nitrogen-rich gas rather than only from dissolved GCs."],"fun_headline_variants":["Giant molecular clouds seed GC light-element abundance patterns","Post-starburst gas collisions create GC-like N–O anticorrelations","GCs inherit chemical signatures from birth clouds after starbursts","Simulations show GCs form in clouds with ready-made abundance spreads","Baryon cycle imprints GC chemistry on high-redshift molecular clouds"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The argument assumes that chemically selected birth clouds with the right light-element spreads will, after unresolved long-term dynamical evolution and without tracked sodium, still yield the full observed properties of present-day globular clusters.","fun_headline_variants_meta":{"raw":{"variants":["Giant molecular clouds seed GC light-element abundance patterns","Post-starburst gas collisions create GC-like N–O anticorrelations","GCs inherit chemical signatures from birth clouds after starbursts","Simulations show GCs form in clouds with ready-made abundance spreads","Baryon cycle imprints GC chemistry on high-redshift molecular clouds"]},"model":"grok-4.5","effort":"low","cost_usd":0.007722,"raw_usage":{"total_tokens":1853,"prompt_tokens":801,"num_sources_used":0,"completion_tokens":79,"cost_in_usd_ticks":77220000,"prompt_tokens_details":{"text_tokens":801,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":973,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":801,"tokens_out":79,"duration_ms":12446,"temperature":1.0,"reasoning_tokens":973,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-11T06:37:12.015598+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Obtain ages and light-element abundances for very young (less than about 1 Myr) high-redshift GC progenitors: inheritance predicts anomalous chemistry already present and star formation ending within roughly 3–5 Myr before iron spreads appear; extended multi-generation formation with delayed pollution would favor classical self-enrichment.","supporting_citations":[],"review_version":1}