Pith. sign in

REVIEW 3 major objections 6 minor 111 references

Aeos: Transport of Metals from Minihalos following Population III Stellar Feedback

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The first supernovae blow most metals out of minihalos below about 10^7 solar masses, so early chemical enrichment is external rather than self-enrichment.

desk verdict Solid star-by-star simulation paper with a real quantitative threshold that is more model-dependent than the abstract admits; deserves refereeing with requested sensitivity caveats. read the letter →

arxiv 2411.14209 v2 pith:JZO77JWQ submitted 2024-11-21 astro-ph.GA

Jennifer Mead , Kaley Brauer , Greg L. Bryan , Mordecai-Mark Mac Low , Alexander P. Ji , John H. Wise , Andrew Emerick , Eric P. Andersson
show 2 more authors
Anna Frebel Benoit Côté
This is my paper · ORCID
classification astro-ph.GA
keywords PopulationIIIstarschemicalenrichmentminihalossupernovafeedbackmetaltransportintergalacticmediumstar-by-starsimulations-processelements
open problems Dark Matter
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper uses the star-by-star cosmological hydrodynamics simulation Aeos to follow where the metals created by the first stars end up. It claims that the energy from core-collapse supernovae of Population III stars is enough to eject most of a halo's gas and newly produced metals beyond the virial radius whenever the dark matter mass is below about $10^7 \, M_\odot$, regardless of how many supernovae explode. As a result, most minihalos retain almost none of the yields they produce until they grow past that threshold, and halo metallicity does not rise steadily: it spikes with each supernova and then drops as the metals are blown out. If correct, the first supernovae delay the transition to enriched (Population II) star formation, because the metals that eventually form the next stellar generation fall back from the intergalactic medium rather than staying in their birth halo. The paper also resolves ten individual elements and finds that core-collapse supernovae dominate early enrichment, with asymptotic giant branch winds already contributing noticeably to the s-process elements strontium and barium.

What carries the argument

The central quantity is the metal loss fraction $\ell = (M_{\rm met,inj} - M_{\rm met})/M_{\rm met,inj}$, where $M_{\rm met}$ is the gas-phase metal mass currently inside a halo's virial radius and $M_{\rm met,inj}$ is the cumulative metal mass injected by supernovae and AGB winds within that halo or its ancestors. This single diagnostic lets the authors distinguish injection from retention and identify the dark matter mass at which halos switch from losing ~100% of their metals to rebuilding them. The simulation Aeos itself is the enabling object: it treats each star particle as an individual star sampled from an IMF rather than as a whole stellar population, samples Population III masses from a Salpeter-like IMF with characteristic mass 20 $M_\odot$, and deposits $10^{51}$ erg per core-collapse supernova, with yields for ten elements from standard nucleosynthesis tables. The paper uses the loss fraction both for total metals and for individual elements to show that all elements share the same retention threshold until the halo reaches roughly $10^7 \, M_\odot$.

What would settle it

Run the same Aeos initial conditions with pair-instability supernovae (140–260 $M_\odot$, $\sim 10^{53}$ erg) included and check whether halos below $10^7 \, M_\odot$ still lose essentially all their injected metals; a retained fraction in that case would show the central threshold depends on the assumed supernova energies.

Watch

Extended reading notes

Core claim

Using the Aeos simulation, the authors find that supernova feedback from the first stars expels a majority of gas and injected metals beyond the virial radius of halos with $M_{\rm dm} \lesssim 10^7 \, M_\odot$, and that this loss is essentially independent of the number of supernovae. Most minihalos ($M_{\rm dm} \gtrsim 10^5 \, M_\odot$) therefore do not retain significant fractions of their own nucleosynthetic yields until they have grown to $M_{\rm dm} \gtrsim 10^7 \, M_\odot$. The resultant metallicity evolution is non-monotonic: the first supernova produces a rapid spike in metal mass that is immediately followed by a loss of two to three orders of magnitude, and later supernovae do not raise the halo's metallicity until metals begin to reaccrete from the intergalactic medium. Reaccretion is shown to be the dominant process rebuilding halo metals, including exchange between neighboring halos such as Halo 0 and Halo 1. On the timescale of the simulation, core-collapse supernovae dominate the production of all ten tracked elements, but asymptotic giant branch winds contribute significantly to the s-process elements Sr and Ba.

Load-bearing premise

The load-bearing assumption is the adopted Population III stellar prescription: stars from 10 to 100 solar masses explode as ordinary $10^{51}$ erg core-collapse supernovae, stars heavier than 100 solar masses collapse silently, and pair-instability supernovae between 140 and 260 solar masses are left out; if the real first stellar population includes a substantial fraction of pair-instability supernovae at $10^{53}$ erg, the halo mass below which metals are lost would shift upward.

Editorial extensions

If this is right

  • Nearly all minihalos below $10^7 \, M_\odot$ are prevented from self-enriching; the metals of the first stars accumulate in the intergalactic medium and can later be accreted by other halos.
  • Early halo metallicity is non-monotonic, so any model that assumes retained ejecta and a steadily rising metallicity will trigger enriched star formation too early.
  • The loss of metals beyond the virial radius delays the onset of Population II star formation and extends the era in which Population III stars can keep forming.
  • Because gas and metals are lost at somewhat different characteristic masses ($3 \times 10^6 \, M_\odot$ for gas, $10^7 \, M_\odot$ for metals), supernova ejecta do not efficiently mix with the cooler local gas into which they are injected.
  • All ten tracked elements, including s-process elements, show the same loss behavior and the same $10^7 \, M_\odot$ retention threshold at these early times.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A direct test of the energy dependence would be to rerun Aeos with pair-instability supernovae (roughly $10^{53}$ erg) included; the authors themselves expect the retention threshold to move upward, which would make early enrichment even more external and reduce the fraction of halos that ever self-enrich.
  • If the threshold holds, the abundance patterns of the first enriched stars should often carry the signature of externally produced metals mixed in the IGM rather than a single local supernova; comparing Aeos predictions to the observed scatter in ultra-faint dwarf galaxy abundances could constrain how much mixing occurs between ejection and reaccretion.
  • The element-resolved loss fractions imply that at z > 14 s-process elements are not preferentially retained despite being injected by gentler AGB winds; at lower redshifts, zoom-in simulations could test whether the lower injection energy of AGB winds eventually makes s-process elements more centrally concentrated than iron-group elements.
  • An analytic escape condition could be derived from the simulation data: the binding energy of the baryons within the virial radius at the moment of first supernova, compared with the $10^{51}$ erg injection, should predict the $10^7 \, M_\odot$ threshold; if such a criterion works, semi-analytic models of the Pop III–Pop II transition can replace their monotonic-enrichment assumption with a simple
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. This paper uses the Aeos star-by-star cosmological hydrodynamics simulation (1 comoving Mpc box, evolved to z = 14.3) to study how Population III stellar feedback transports metals out of minihalos. The central claim is that energy from core-collapse supernovae (CCSNe) expels a majority of the gas and injected metals beyond the virial radius of halos with dark matter mass Mdm <= 10^7 Msun, almost regardless of the number of supernovae, so that most minihalos do not retain their own nucleosynthetic yields until they grow to Mdm >= 10^7 Msun. The authors infer non-monotonic early metallicity evolution, metal deposition into the IGM followed by reaccretion or external enrichment of neighboring halos, and a delay of the Population III to Population II transition. A second component of the paper decomposes enrichment into 10 elements from CCSNe, Type Ia SNe, and AGB winds, finding CCSN dominance at z ~ 14.3 with a significant AGB contribution to the s-process elements Sr and Ba. The fiducial model adopts a Salpeter Pop III IMF with Mchar = 20 Msun, CCSNe at fixed 10^51 erg for 10-100 Msun stars, direct collapse without feedback for 100-300 Msun stars, no pair-instability supernovae, and zero streaming velocity; the authors describe these choices in Sections 2 and 6.2.

Significance. The metal-loss fraction (Eq. 2) is an emergent bookkeeping quantity rather than a fitted parameter, which is a genuine strength: the result that less than about 8% of injected metals reside within halos until z ~ 15.3, that 92% of halo metals are concentrated in the four most massive halos, and that metals are exchanged between neighboring halos (Halos 0 and 1) are concrete, falsifiable statements. The star-by-star treatment that resolves the Sedov-Taylor phase of most CCSNe at 1 pc resolution, the tracking of 10 elements through distinct nucleosynthetic channels, the candid robustness section (Sec. 6.2), and the quantitative comparison with the Kulkarni et al. (2021) critical mass all work in the paper's favor. If the central threshold holds, the implications for the Pop III to Pop II transition are substantial, because semi-analytic models assuming monotonic enrichment (e.g., Visbal et al. 2020; Hartwig et al. 2022) would need revision. The caveat is that the headline threshold and the non-monotonicity are conditional on the fiducial CCSN-only, vbc = 0 prescription, and the paper should present them as such.

major comments (3)
  1. [Abstract; Secs. 2.3, 4, 6.2] The headline result, expulsion of gas and metals beyond rvir for halos with Mdm <= 10^7 Msun regardless of the number of supernovae, is stated categorically in the abstract and in Summary item 5, but it is established only for the fiducial prescription of Section 2.3: Pop III stars of 10-100 Msun explode as 10^51 erg CCSNe, stars of 100-300 Msun collapse directly with no feedback, and pair-instability SNe (140-260 Msun) are excluded. This model-dependence is load-bearing, not cosmetic. For a 10^7 Msun halo at z ~ 14 (Tvir ~ 3430 K, vc ~ 9.7 km/s, as quoted in Sec. 4), the gas binding energy is of order G Mdm Mg / rvir ~ 10^51 erg, comparable to a single CCSN; a single 10^53 erg PISN therefore exceeds the binding energy by two orders of magnitude and would unbind gas from halos well above the claimed threshold, assuming comparable coupling. This is precisely the point conceded in Section 6.2 item 2, which states that including PISNe could increase the characteristic retention masses and the extent of IGM enrichment. I recommend (i) making the abstract and Section 6.1 explicitly conditional, e.g., in the fiducial CCSN-only, vbc = 0 model, and (ii) adding a short analytic energy-budget estimate, injected energy per event versus gas binding energy as a function of Mdm for 10^51 and 10^53 erg events, so that the reader can see both the origin of the 10^7 Msun threshold and its sensitivity to the adopted SN prescription.
  2. [Abstract; Sec. 6.2 item 3] The abstract's claims of non-monotonic metallicity evolution and a delayed Pop III to Pop II transition are presented as general findings, but they rest on the choice vbc = 0. Section 6.2 item 3 concedes that for halos first forming stars at Mdm > 3 x 10^6 Msun, feedback may no longer expel gas and the halo metallicity could instead increase monotonically with time. Since realistic streaming velocities shift first star formation to larger and later-forming halos, the vbc = 0 assumption biases the simulation toward the paper's qualitative conclusions rather than away from them. The abstract and Summary item 3 should state the non-monotonicity result as conditional on vbc = 0 with a forward reference to the robustness discussion; as written, a reader of the abstract will take the non-monotonicity to be a robust feature of early enrichment rather than a fiducial-model outcome.
  3. [Secs. 4.1.3, 5, 6.2 item 1; Fig. 5] The N-independence of the metal loss fraction is the paper's principal argument for robustness to IMF variation (Section 6.2 item 1), but the evidence supports a narrower claim than the one stated. Figure 5 demonstrates approximately unit loss fraction independent of SN number only over the simulated range of roughly 0-10 CCSNe per halo, and the abstract's regardless-of-number phrasing applies to expulsion beyond rvir; Section 4.1.3 shows for Halo 2 that a clustered series of SNe drives metals beyond 2rvir and states that ejection at farther distances may depend on the number of SNe. The distinction matters because a PISN is not merely an additional event of the same kind: at 10^53 erg it is a roughly 100-fold increase in the energy budget per event, a regime that the N-independence test does not sample. As written, the robustness claim in Section 6.2 item 1 (no change in the characteristic retention mass so long as Pop III CCSNe are prevalent) conflates event count with energy budget. I recommend either scoping the claim to the fiducial 10^51 erg events and the rvir boundary, or adding a test that rescales the injected SN energy to 10^52-10^53 erg to map the threshold's sensitivity.
minor comments (6)
  1. [Abstract] The symbol M* used for the retention threshold in the abstract is ambiguous; the body (Section 4 and Figures 5-10) defines the threshold in terms of dark matter halo mass Mdm, and M* conventionally denotes stellar mass. Please align the abstract notation with the body.
  2. [Sec. 2.2.1] The statement that most of our Pop III stars will have masses between 20 and 300 Msun is a number-count statement whose truth depends on the normalization of the exponential cutoff below Mchar; please clarify whether most refers to the number fraction or the mass fraction of sampled stars.
  3. [Fig. 3 caption] The caption contains the typo Alhough, which should read Although.
  4. [Sec. 5.2; Fig. 12] The Sr and Ba loss-fraction panels are known to be affected by the AGB-injection accounting artifact described in the text; please mark the affected panels in Figure 12 explicitly, or correct the injection-time accounting, so that their anomalous values of the loss fraction are not read as physical.
  5. [Sec. 4.1.2; Fig. 7] The inference that Halo 2 may be entering the metal-retention stage rests on the direction of a single final snapshot; a sentence acknowledging the single-timestep basis of this inference would strengthen the case study.
  6. [Sec. 2.1] The simulation uses 1 pc maximum refinement and no pressure floor, with the Jeans length potentially unresolved; please state explicitly whether the metal-loss statistics are converged with resolution, with a reference to B24, or flag the resolution dependence as a caveat.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the central metal-loss fractions and retention thresholds are emergent simulation bookkeeping, not fitted parameters or self-citation-derived claims.

full rationale

I walked the claimed derivation chain from the abstract and Section 3-4 down to the quantitative definitions. The central quantity, the metal loss fraction, is defined as ℓ = (Mmet,inj - Mmet)/Mmet,inj (Eq. 2), where Mmet,inj is the injected metal mass within the halo or its ancestors and Mmet is the current gas-phase metal mass. Both sides are independently tracked simulation outputs; nothing in the definition presets the value of ℓ, and the claim that most minihalos lose nearly all metals until Mdm reaches about 1e7 Msun is an emergent bookkeeping result rather than an assumed input. No parameter is fitted to this loss fraction and no equation reduces the mass threshold to an assumed yield or IMF parameter. The calibrated inputs that exist, such as the artificial factor-of-2.2 Mg yield boost and the adopted IMF/SN prescription, affect composition and event rates, but they do not by construction force the loss-fraction behavior or the 'regardless of number of SNe' result: halos with different SN counts all exhibit the same loss behavior, which is a simulation outcome. The self-citations to Brauer et al. 2025 and Emerick et al. 2019 describe simulation methods and subgrid prescriptions, but those are normal provenance for the code and are not invoked as an external theorem that forces the paper's conclusions. The paper explicitly flags model dependence in Section 6.2, including the possibility that pair-instability SNe or streaming velocities would shift the retention masses or make metallicity evolution monotonic; these are stated limitations and alternate-model sensitivities, not circular reductions. In short, I found no step where a 'prediction' reduces by construction to its own input, no fitted parameter renamed as a prediction, and no load-bearing self-citation chain. The appropriate circularity finding is no significant circularity, score 0.

Assumptions & free parameters 10 free parameters · 7 assumptions · 0 invented entities

The simulation uses standard subgrid choices and published yield tables; none of the free parameters is fitted to the paper's central metal-loss result, and none was introduced post hoc to produce that result. The assumptions that could most change the conclusions are the IMF and supernova energy prescription and the vbc = 0 setting, both acknowledged in the paper.

free parameters (10)
  • Population III IMF characteristic mass and slope = Mchar = 20 Msun, Salpeter slope alpha = 2.3
    Adopted from Wise et al. (2012); determines the mass distribution of first stars and hence which stars become core-collapse supernovae. The central claim depends on this IMF.
  • Population III mass range = 1 to 300 Msun
    Sets the fraction of stars in the 10 to 100 Msun core-collapse range; stars above 100 Msun are assumed to collapse directly with no feedback.
  • Core-collapse supernova energy = 1e51 erg per supernova
    Fixed thermal energy injection for every CCSN; directly determines whether ejecta escape from minihalos.
  • Star formation efficiency per freefall time = epsilon_ff = 0.02
    Controls how promptly gas turns into stars and therefore the timing and number of supernovae.
  • Star formation density and temperature thresholds = n > 1e4 cm^-3 and T < 500 K
    Used in place of a pressure floor to avoid unresolved Jeans fragmentation; affects where stars form in the simulation.
  • Population III metallicity threshold = Z < 1e-5 Zsun
    Defines the boundary between Pop III and Pop II star formation and is central to the claimed delay of enriched star formation.
  • Population III H2 fraction threshold = fH2 > 0.0005
    Required for Pop III star formation, following Susa et al. (2014); affects which halos form the first stars.
  • Magnesium yield boost factor = 2.2
    Artificially increases Mg yields from all massive stars to match stellar abundance observations; this calibration affects element-specific enrichment conclusions.
  • Stellar wind velocity cap = 100 km/s
    Massive star winds are injected at a fixed 100 km/s for computational efficiency and are fully thermalized; affects AGB and massive wind feedback.
  • Box size and maximum resolution = 1 comoving Mpc box, 1 pc maximum refinement
    Determines which halos are resolved and the volume of IGM available for metal transport; affects statistics such as the 80% of halos with enrichment beyond 2rvir.
assumptions (7)
  • domain assumption The adopted cosmological and baryonic physics, including LCDM structure formation and Grackle non-equilibrium chemistry, faithfully represents the early universe.
    Invoked throughout Section 2; if the chemistry or background model is wrong, the star formation thresholds and feedback behavior would change.
  • domain assumption Gas with metallicity below 1e-5 Zsun forms Population III stars rather than Population II stars.
    Used in Section 2.2.1 to define Pop III star formation; this directly controls the Pop III to Pop II transition timeline.
  • domain assumption Core-collapse supernovae with fixed 1e51 erg thermal energy capture the feedback that drives metal loss, with the kinetic phase resolved or negligible.
    Section 2.3 states that the Sedov-Taylor phase is resolved at 1 pc, so only thermal energy is deposited; this underpins the metal-loss result.
  • domain assumption The adopted yield tables from Heger and Woosley (2010), Limongi and Chieffi (2018), and Cristallo et al. (2015) correctly represent nucleosynthetic outputs.
    Section 2.4 uses these tables for all element channels; the element-specific claims inherit their accuracy.
  • domain assumption The density contour rho200/3 and the virial radius formula in Equation 1 are appropriate halo boundaries for computing metal loss fractions.
    Used in Sections 3 and 4 to define halo and IGM metal fractions; changing the boundary changes the quoted percentages.
  • ad hoc to paper Pair-instability supernovae are rare enough or absent that excluding them does not overturn the central result.
    Section 2.3 excludes PISNe, and Section 6.2 argues they would only strengthen metal loss, but this is a stated assumption of the paper rather than an independently established fact.
  • domain assumption The dark matter streaming velocity vbc = 0 is representative for the conclusions.
    The simulation uses vbc = 0, and Section 6.2 item 3 acknowledges that reasonable nonzero streaming velocities could suppress outflows in larger halos and change the monotonicity of metallicity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Aeos: Transport of Metals from Minihalos following Population III Stellar Feedback." pith.science (2026). https://pith.science/paper/JZO77JWQ

@misc{pith2026241114209,
  author       = {Pith},
  title        = {Pith review of: Aeos: Transport of Metals from Minihalos following Population III Stellar Feedback},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JZO77JWQ}},
  note         = {Machine review of arXiv:2411.14209}
}
abstract

We investigate how stellar feedback from the first stars (Population III) distributes metals through the interstellar and intergalactic medium using the star-by-star cosmological hydrodynamics simulation, Aeos. We find that energy injected from the supernovae of the first stars is enough to expel a majority of gas and injected metals beyond the virial radius of halos with mass $M_* \lesssim 10^7$ M$_\odot$, regardless of the number of supernovae. This prevents self-enrichment and results in a non-monotonic increase in metallicity at early times. Most minihalos ($M \gtrsim 10^5 \, \rm M_\odot$) do not retain significant fractions of the yields produced within their virial radii until they have grown to halo masses of $M \gtrsim 10^7 \, \rm M_\odot$. The loss of metals to regions well beyond the virial radius delays the onset of enriched star formation and extends the period that Population III star formation can persist. We also explore the contributions of different nucleosynthetic channels to 10 individual elements. On the timescale of the simulation (lowest redshift $z=14.3$), enrichment is dominated by core-collapse supernovae for all elements, but with a significant contribution from asymptotic giant branch winds to the s-process elements, which are normally thought to only be important at late times. In this work, we establish important mechanisms for early chemical enrichment which allows us to apply Aeos in later epochs to trace the evolution of enrichment during the complete transition from Population III to Population II stars.

Figures

Figures reproduced from arXiv: 2411.14209 by the authors.

Figure 1
Figure 1. Unweighted projection of metal mass in the AEOS 1 comoving Mpc box at the final redshift of these runs, z = 14.26 (t = 289.2 Myr). The x- and y-axes are in physical units. we use the AEOS20 simulation, which has a minimum Pop III mass of 1 M⊙ and maximum mass of 300 M⊙, with Mchar = 20 M⊙. Thus, most of our Pop III stars will have masses between 20 − 300M⊙. Pop III stars are assigned lifetimes from Schaerer (2002). … view at source ↗
Figure 2
Figure 2. Gas density and metal density slices of the most massive halos in the AEOS box with rvir indicated by the cyan circles. The white contours represent the ρ200/3 density threshold. The x- and y-axes are in physical units. Pop II CCSNe Yields: Yields for Pop II CCSNe were adopted from Limongi & Chieffi (2018) using nine grid points in mass (13M⊙ ≤ M∗ ≤ 120M⊙) and four grid points in metallicity (3.236×10−5 ≤ Z∗ ≤ 0.013… view at source ↗
Figure 4
Figure 4. Orange points represent the halo mass at the redshift of first Pop III star formation in a halo’s history. The blue dashed line is the theoretical critical mass calculated from Kulkarni et al. (2021). 35%, however, this increase is rapid and may reach even higher fractions in a short amount of time following the end of the simulation. Of the metals in all halos, 92% reside in the four most massive halos in the simul… view at source ↗
Figures from the paper (9 more)
Figure 3
Figure 3. Figure 3: Top: Fraction of injected metal mass that remains in halos where the boundary of the halo is either the virial radius (pur￾ple) or the contour where gas is denser than ρ200/3, a density that is typical of collapsed halos (black). Center: The total metallicities (Mmet/M…
Figure 5
Figure 5. Figure 5: Metal loss fraction, ℓ. Top: Metal loss fraction at the last simulation snapshot at z = 14.3 colored by the total number of SNe exploded in each halo. A metal loss fraction of ℓ = 0 indicates that the current metal mass Mmet is equivalent to the total metal mass inject…
Figure 6
Figure 6. Figure 6: shows the time evolution in dark matter, baryon fraction fb, metal mass, and metallicity respectively within 10 5 10 6 10 7 Md m (M ) Halo 0 Halo 1 Halo 2 Halo 3 10 5 10 3 10 1 fb ( Mg Mtot ) 10 6 10 4 10 2 10 0 10 2 Mm et (M ) 100 150 200 250 300 Time (Myr) 10 9 10 7 …
Figure 7
Figure 7. Figure 7: Metal loss fraction of the most massive progenitor line for the four most massive halos in the simulation as a function of Mdm. The grey dashed line indicates the median value of ℓ for the four most massive halos, while the grey shaded region is bounded by the 16th and…
Figure 8
Figure 8. Figure 8: Cumulative metal mass profiles of Halo 0 (left) and Halo 2 (right) in AEOS over time. The gradient purple to yellow represents the highest to lowest redshifts respectively. Lines along the top indicate the virial radius at the redshift for the corresponding profile. ev…
Figure 9
Figure 9. Figure 9: Top panel: Number of SNe in each halo that explode within the virial radius of the halo between snapshots versus Mdm. Bottom panel: Evolution of the metal loss fraction within different multiples of rvir. For Halos 0 and 1, the vertical black line depicts the time (160…
Figure 10
Figure 10. Figure 10: Top: Evolution of fb with dark matter mass of the most massive progenitor of intermediate mass (5×106 < Mdm < 107 M⊙) halos. Halos regain their gas ∼ 106 M⊙. Bottom: ℓ evolution with dark matter mass. The blue dashed line indicates the median value of ℓ for these halo…
Figure 11
Figure 11. Figure 11: Total injected yield (black line) for all metals as well as individual elements from AGB winds (cyan) and Pop III (orange)/Pop II (green) CCSNe occurring within rvir, for each of the four most massive halos over time. Note the different y-scale for elements Sr and Ba.…
Figure 12
Figure 12. Figure 12: Metal loss fraction ℓ within rvir of individual elements as labeled, representing light (O), α (Mg), Fe-peak (Fe), and heavy (Sr) elements for the four most massive halos. Lines and symbols are the same as in [PITH_FULL_IMAGE:figures/full_fig_p015_12.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

111 extracted references · 44 canonical work pages

  1. [1]

    L., & Norman, M

    Abel, T., Bryan, G. L., & Norman, M. L. 2002, Science, 295, 93

    2002

  2. [2]

    I., et al

    Agertz, O., Pontzen, A., Read, J. I., et al. 2020, MNRAS, 491, 1656

    2020

  3. [3]

    P., Agertz, O., Renaud, F., & Teyssier, R

    Andersson, E. P., Agertz, O., Renaud, F., & Teyssier, R. 2023, MNRAS, 521, 2196

    2023

  4. [4]

    P., Rey, M

    Andersson, E. P., Rey, M. P., Pontzen, A., et al. 2025, ApJ, 978, 129

    2025

  5. [5]

    P., Frebel, A., et al

    Brauer, K., Ji, A. P., Frebel, A., et al. 2019, ApJ, 871, 247

    2019

  6. [6]

    2025, ApJ, 980, 41

    Brauer, K., Emerick, A., Mead, J., et al. 2025, ApJ, 980, 41

    2025

  7. [7]

    2012, MNRAS, 427, 127

    Bressan, A., Marigo, P., Girardi, L., et al. 2012, MNRAS, 427, 127

    2012

  8. [8]

    S., & Larson, R

    Bromm, V ., Coppi, P. S., & Larson, R. B. 1999, ApJL, 527, L5 —. 2002, ApJ, 564, 23

    1999

Show all 111 references
  1. [9]

    2004, NewA, 9, 353

    Bromm, V ., & Loeb, A. 2004, NewA, 9, 353

    2004

  2. [10]

    M., Tumlinson, J., Geha, M., et al

    Brown, T. M., Tumlinson, J., Geha, M., et al. 2014, ApJ, 796, 91

    2014

  3. [11]

    2019, The Journal of Open Source Software, 4, 1636

    Brummel-Smith, C., Bryan, G., Butsky, I., et al. 2019, The Journal of Open Source Software, 4, 1636

    2019

  4. [12]

    L., Norman, M

    Bryan, G. L., Norman, M. L., O’Shea, B. W., et al. 2014, ApJS, 211, 19

    2014

  5. [13]

    J., et al

    Chen, K.-J., Tang, C.-Y ., Whalen, D. J., et al. 2024, ApJ, 964, 91

    2024

  6. [14]

    2018, MNRAS, 475, 4378

    Chiaki, G., Susa, H., & Hirano, S. 2018, MNRAS, 475, 4378

    2018

  7. [15]

    Chiaki, G., & Wise, J. H. 2019, MNRAS, 482, 3933

    2019

  8. [16]

    2016, MNRAS, 463, 2781

    Chiaki, G., Yoshida, N., & Hirano, S. 2016, MNRAS, 463, 2781

    2016

  9. [17]

    C., Glover, S

    Clark, P. C., Glover, S. C. O., Smith, R. J., et al. 2011, Science, 331, 1040

    2011

  10. [18]

    T., Fialkov, A., Dixon, K

    Conaboy, L., Iliev, I. T., Fialkov, A., Dixon, K. L., & Sullivan, D. 2023, MNRAS, 525, 5479 Correa Magnus, L., Smith, B. D., Khochfar, S., et al. 2024, MNRAS, 527, 307

    2023

  11. [19]

    A., & van de V oort, F

    Crain, R. A., & van de V oort, F. 2023, ARA&A, 61, 473

    2023

  12. [20]

    2015, ApJS, 219, 40

    Cristallo, S., Straniero, O., Piersanti, L., & Gobrecht, D. 2015, ApJS, 219, 40

    2015

  13. [21]

    2012, Research in Astronomy and Astrophysics, 12, 1197 Davé, R., Rafieferantsoa, M

    Cui, X.-Q., Zhao, Y .-H., Chu, Y .-Q., et al. 2012, Research in Astronomy and Astrophysics, 12, 1197 Davé, R., Rafieferantsoa, M. H., Thompson, R. J., & Hopkins, P. F. 2017, MNRAS, 467, 115 de Bennassuti, M., Salvadori, S., Schneider, R., Valiante, R., &

    2012

  14. [22]

    2017, MNRAS, 465, 926 De Rossi, M

    Omukai, K. 2017, MNRAS, 465, 926 De Rossi, M. E., Bower, R. G., Font, A. S., Schaye, J., & Theuns, T. 2017, MNRAS, 472, 3354 De Silva, G. M., Freeman, K. C., Bland-Hawthorn, J., et al. 2015, MNRAS, 449, 2604

    2017

  15. [23]

    1986, ApJ, 303, 39

    Dekel, A., & Silk, J. 1986, ApJ, 303, 39

    1986

  16. [24]

    L., & Mac Low, M.-M

    Emerick, A., Bryan, G. L., & Mac Low, M.-M. 2019, MNRAS, 482, 1304

    2019

  17. [25]

    L., Mac Low, M.-M., et al

    Emerick, A., Bryan, G. L., Mac Low, M.-M., et al. 2018, ApJ, 869, 94

    2018

  18. [26]

    U., et al

    Ezzeddine, R., Frebel, A., Roederer, I. U., et al. 2019, ApJ, 876, 97

    2019

  19. [27]

    C., Murray, S

    Fragile, P. C., Murray, S. D., & Lin, D. N. C. 2004, ApJ, 617, 1077

    2004

  20. [28]

    Frebel, A., & Norris, J. E. 2015, ARA&A, 53, 631 Gaia Collaboration, Prusti, T., de Bruijne, J. H. J., et al. 2016, A&A, 595, A1 Gaia Collaboration, Vallenari, A., Brown, A. G. A., et al. 2023, A&A, 674, A1

    2015

  21. [29]

    Y ., & Kravtsov, A

    Gnedin, N. Y ., & Kravtsov, A. V . 2006, ApJ, 645, 1054

    2006

  22. [30]

    H., Bromm, V ., Clark, P

    Greif, T. H., Bromm, V ., Clark, P. C., et al. 2012, MNRAS, 424, 399

    2012

  23. [31]

    F., Dooley, G

    Griffen, B. F., Dooley, G. A., Ji, A. P., et al. 2018, MNRAS, 474, 443

    2018

  24. [32]

    A., Pakmor, R., Naab, T., & Springel, V

    Gutcke, T. A., Pakmor, R., Naab, T., & Springel, V . 2021, MNRAS, 501, 5597

    2021

  25. [33]

    2012, ApJ, 746, 125

    Haardt, F., & Madau, P. 2012, ApJ, 746, 125

    2012

  26. [34]

    S., & Glover, S

    Hartwig, T., Bromm, V ., Klessen, R. S., & Glover, S. C. O. 2015, MNRAS, 447, 3892

    2015

  27. [35]

    Hartwig, T., Lipatova, V ., Glover, S. C. O., & Klessen, R. S. 2024, MNRAS, 535, 516

    2024

  28. [36]

    2022, ApJ, 936, 45

    Hartwig, T., Magg, M., Chen, L.-H., et al. 2022, ApJ, 936, 45

    2022

  29. [37]

    R., Bovy, J., Holtzman, J

    Hayden, M. R., Bovy, J., Holtzman, J. A., et al. 2015, ApJ, 808, 132

    2015

  30. [38]

    Hegde, S., & Furlanetto, S. R. 2023, MNRAS, 525, 428

    2023

  31. [39]

    Heger, A., & Woosley, S. E. 2002, ApJ, 567, 532 —. 2010, ApJ, 724, 341 17

    2002

  32. [40]

    M., Wells, A., Norman, M

    Hicks, W. M., Wells, A., Norman, M. L., et al. 2021, ApJ, 909, 70

    2021

  33. [41]

    2019, A&A, 626, A15

    Hill, V ., Skúladóttir, Á., Tolstoy, E., et al. 2019, A&A, 626, A15

    2019

  34. [42]

    R., et al

    Hirai, Y ., Ishimaru, Y ., Saitoh, T. R., et al. 2017, MNRAS, 466, 2474

    2017

  35. [43]

    Hirai, Y ., & Saitoh, T. R. 2017, ApJL, 838, L23

    2017

  36. [44]

    Hirano, S., Hosokawa, T., Yoshida, N., Omukai, K., & Yorke, H. W. 2015, MNRAS, 448, 568

    2015

  37. [45]

    2014, ApJ, 781, 60

    Hirano, S., Hosokawa, T., Yoshida, N., et al. 2014, ApJ, 781, 60

    2014

  38. [46]

    2016, ApJ, 824, 119

    Hosokawa, T., Hirano, S., Kuiper, R., et al. 2016, ApJ, 824, 119

    2016

  39. [47]

    Hosokawa, T., Omukai, K., Yoshida, N., & Yorke, H. W. 2011, Science, 334, 1250

    2011

  40. [48]

    2019, MNRAS, 483, 3363

    Hu, C.-Y . 2019, MNRAS, 483, 3363

    2019

  41. [49]

    Jaura, O., Glover, S. C. O., Wollenberg, K. M. J., et al. 2022, MNRAS, 512, 116

    2022

  42. [50]

    2017, ApJ, 848, 85

    Jeon, M., Besla, G., & Bromm, V . 2017, ApJ, 848, 85

    2017

  43. [51]

    P., Frebel, A., & Bromm, V

    Ji, A. P., Frebel, A., & Bromm, V . 2015, MNRAS, 454, 659

    2015

  44. [52]

    P., Li, T

    Ji, A. P., Li, T. S., Simon, J. D., et al. 2020, ApJ, 889, 27

    2020

  45. [53]

    P., Curtis, S., Storm, N., et al

    Ji, A. P., Curtis, S., Storm, N., et al. 2024, ApJL, 961, L41

    2024

  46. [54]

    I., & Lattanzio, J

    Karakas, A. I., & Lattanzio, J. C. 2014, PASA, 31, e030

    2014

  47. [55]

    2005, ApJ, 630, 675

    Kitayama, T., & Yoshida, N. 2005, ApJ, 630, 675

    2005

  48. [56]

    S., & Glover, S

    Klessen, R. S., & Glover, S. C. O. 2023, ARA&A, 61, 65

    2023

  49. [57]

    2006, ApJ, 653, 1145

    Kobayashi, C., Umeda, H., Nomoto, K., Tominaga, N., & Ohkubo, T. 2006, ApJ, 653, 1145

    2006

  50. [58]

    2001, MNRAS, 322, 231

    Kroupa, P. 2001, MNRAS, 322, 231

    2001

  51. [59]

    R., & Ting, Y .-S

    Krumholz, M. R., & Ting, Y .-S. 2018, MNRAS, 475, 2236

    2018

  52. [60]

    Kulkarni, M., Visbal, E., & Bryan, G. L. 2021, ApJ, 917, 40 Lahén, N., Naab, T., Johansson, P. H., et al. 2020, ApJ, 891, 2 Lahén, N., Naab, T., Kauffmann, G., et al. 2023, MNRAS, 522, 3092

    2021

  53. [61]

    Y ., Naoz, S., et al

    Lake, W., Grudi´c, M. Y ., Naoz, S., et al. 2025, ApJL, 985, L6

    2025

  54. [62]

    A., Schleicher, D

    Latif, M. A., Schleicher, D. R. G., Schmidt, W., & Niemeyer, J. 2013, ApJL, 772, L3

    2013

  55. [63]

    A., Whalen, D., & Khochfar, S

    Latif, M. A., Whalen, D., & Khochfar, S. 2022, ApJ, 925, 28

    2022

  56. [64]

    2018, ApJS, 237, 13

    Limongi, M., & Chieffi, A. 2018, ApJS, 237, 13

    2018

  57. [65]

    F., Faucher-Giguère, C.-A., et al

    Ma, X., Hopkins, P. F., Faucher-Giguère, C.-A., et al. 2016, MNRAS, 456, 2140 Mac Low, M.-M., & Ferrara, A. 1999, ApJ, 513, 142

    2016

  58. [66]

    S., Glover, S

    Magg, M., Klessen, R. S., Glover, S. C. O., & Li, H. 2019, MNRAS, 487, 486

    2019

  59. [67]

    Magg, M., Schauer, A. T. P., Klessen, R. S., et al. 2022, ApJ, 929, 119

    2022

  60. [68]

    R., Schiavon, R

    Majewski, S. R., Schiavon, R. P., Frinchaboy, P. M., et al. 2017, AJ, 154, 94

    2017

  61. [69]

    Marcolini, A., D’Ercole, A., Battaglia, G., & Gibson, B. K. 2008, MNRAS, 386, 2173

    2008

  62. [70]

    C., & Feldmann, R

    Marszewski, A., Sun, G., Faucher-Giguère, C.-A., Hayward, C. C., & Feldmann, R. 2024, ApJL, 967, L41

    2024

  63. [71]

    1986, A&A, 154, 279

    Matteucci, F., & Greggio, L. 1986, A&A, 154, 279

    1986

  64. [72]

    J., & Horta, D

    Mead, J., Ness, M., Andersson, E., Griffith, E. J., & Horta, D. 2024, ApJ, 974, 186

    2024

  65. [73]

    L., Gnedin, O

    Muratov, A. L., Gnedin, O. Y ., Gnedin, N. Y ., & Zemp, M. 2013, ApJ, 772, 106

    2013

  66. [74]

    L., Kereš, D., Faucher-Giguère, C.-A., et al

    Muratov, A. L., Kereš, D., Faucher-Giguère, C.-A., et al. 2017, MNRAS, 468, 4170

    2017

  67. [75]

    Naoz, S., Yoshida, N., & Gnedin, N. Y . 2012, ApJ, 747, 128 —. 2013, ApJ, 763, 27

    2012

  68. [76]

    F., Frenk, C

    Navarro, J. F., Frenk, C. S., & White, S. D. M. 1997, ApJ, 490, 493

    1997

  69. [77]

    B., Stinson, G., et al

    Obreja, A., Brook, C. B., Stinson, G., et al. 2014, MNRAS, 442, 1794 O’Leary, R. M., & McQuinn, M. 2012, ApJ, 760, 4

    2014

  70. [78]

    2018, MNRAS, 476, 3432

    Prantzos, N., Abia, C., Limongi, M., Chieffi, A., & Cristallo, S. 2018, MNRAS, 476, 3432

    2018

  71. [79]

    2012, A&A, 538, A82

    Revaz, Y ., & Jablonka, P. 2012, A&A, 538, A82

    2012

  72. [80]

    2009, A&A, 501, 189

    Revaz, Y ., Jablonka, P., Sawala, T., et al. 2009, A&A, 501, 189

    2009

  73. [81]

    2021, MNRAS, 503, 6026

    Rossi, M., Salvadori, S., & Skúladóttir, Á. 2021, MNRAS, 503, 6026

    2021

  74. [82]

    J., Belczynski, K., Sim, S

    Ruiter, A. J., Belczynski, K., Sim, S. A., et al. 2011, MNRAS, 417, 408

    2011

  75. [83]

    Salpeter, E. E. 1955, ApJ, 121, 161

    1955

  76. [84]

    2010, MNRAS, 402, 1599

    Sawala, T., Scannapieco, C., Maio, U., & White, S. 2010, MNRAS, 402, 1599

    2010

  77. [85]

    2002, A&A, 382, 28

    Schaerer, D. 2002, A&A, 382, 28

    2002

  78. [86]

    Schauer, A. T. P., Glover, S. C. O., Klessen, R. S., & Clark, P. 2021, MNRAS, 507, 1775

    2021

  79. [87]

    Sharda, P., Federrath, C., & Krumholz, M. R. 2020, MNRAS, 497, 336

    2020

  80. [88]

    Sharda, P., & Menon, S. H. 2025, MNRAS, 540, 1745

    2025

  81. [89]

    Skinner, D., & Wise, J. H. 2020, MNRAS, 492, 4386 Skúladóttir, Á., Koutsouridou, I., Vanni, I., et al. 2024, ApJL, 968, L23 Skúladóttir, Á., Salvadori, S., Amarsi, A. M., et al. 2021, ApJL, 915, L30

    2020

  82. [90]

    2015, MNRAS, 452, 2822

    Khochfar, S. 2015, MNRAS, 452, 2822

    2015

  83. [91]

    D., Bryan, G

    Smith, B. D., Bryan, G. L., Glover, S. C. O., et al. 2017, MNRAS, 466, 2217

    2017

  84. [92]

    C., Sijacki, D., & Shen, S

    Smith, M. C., Sijacki, D., & Shen, S. 2018, MNRAS, 478, 302

    2018

  85. [93]

    2011, ApJL, 730, L1

    Stacy, A., Bromm, V ., & Loeb, A. 2011, ApJL, 730, L1

    2011

  86. [94]

    H., & Bromm, V

    Stacy, A., Greif, T. H., & Bromm, V . 2010, MNRAS, 403, 45

    2010

  87. [95]

    2014, ApJ, 792, 32

    Susa, H., Hasegawa, K., & Tominaga, N. 2014, ApJ, 792, 32

    2014

  88. [96]

    J., et al

    Tegmark, M., Silk, J., Rees, M. J., et al. 1997, ApJ, 474, 1

    1997

  89. [97]

    1986, A&A, 158, 17

    Thielemann, F.-K., Nomoto, K., & Yokoi, K. 1986, A&A, 158, 17

    1986

  90. [98]

    X., Woosley, S

    Timmes, F. X., Woosley, S. E., & Weaver, T. A. 1995, ApJS, 98, 617

    1995

  91. [99]

    Tinsley, B. M. 1980, FCPh, 5, 287

    1980

  92. [100]

    2019, MNRAS, 484, 5587

    Torrey, P., V ogelsberger, M., Marinacci, F., et al. 2019, MNRAS, 484, 5587

    2019

  93. [101]

    2010, ApJ, 708, 1398

    Tumlinson, J. 2010, ApJ, 708, 1398

    2010

  94. [102]

    2023, MNRAS, 518, 1128 18

    Valentini, M., Dolag, K., Borgani, S., et al. 2023, MNRAS, 518, 1128 18

    2023

  95. [103]

    L., & Haiman, Z

    Visbal, E., Bryan, G. L., & Haiman, Z. 2020, ApJ, 897, 95

    2020

  96. [104]

    E., Lake, W., Naoz, S., et al

    Williams, C. E., Lake, W., Naoz, S., et al. 2024, ApJL, 960, L16

    2024

  97. [105]

    H., & Abel, T

    Wise, J. H., & Abel, T. 2011, MNRAS, 414, 3458

    2011

  98. [106]

    H., Turk, M

    Wise, J. H., Turk, M. J., Norman, M. L., & Abel, T. 2012, ApJ, 745, 50

    2012

  99. [107]

    Woosley, S. E. 2017, ApJ, 836, 244

    2017

  100. [108]

    2023, Nature, 618, 712

    Xing, Q.-F., Zhao, G., Liu, Z.-W., et al. 2023, Nature, 618, 712

    2023

  101. [109]

    S., et al

    Yong, D., Kobayashi, C., Da Costa, G. S., et al. 2021, Nature, 595, 223

    2021

  102. [110]

    2003, ApJ, 592, 645

    Yoshida, N., Abel, T., Hernquist, L., & Sugiyama, N. 2003, ApJ, 592, 645

    2003

  103. [111]

    2008, Science, 321, 669

    Yoshida, N., Omukai, K., & Hernquist, L. 2008, Science, 321, 669

    2008

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.