Hunting the first Cosmic Giants: formation and detectability of Direct Collapse Black Holes around high-redshift quasars

Alessandro Lupi; Alessandro Trinca; Marta Volonteri; Raffaella Schneider; Roberto Decarli; Rosa Valiante; Zolt\'an Haiman

REVIEW 3 major objections 5 minor 1 cited by

Hunting the first Cosmic Giants: formation and detectability of Direct Collapse Black Holes around high-redshift quasars

T0 review · 3 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read This paper predicts that direct-collapse black holes form as early as redshift 22 inside quasar-building overdensities and that tens of them survive unmerged to z≈7, where JWST can catch a few in bright accretion episodes.

desk verdict Useful semi-analytic forecast with honest sensitivity tests, but the headline 'tens of DCBHs' is conditional on a volume-averaged metal pollution model; the JWST predictions are worth engaging but need an error budget and public data. read the letter →

arxiv 2601.14370 v2 pith:Z6ZNZ4IX submitted 2026-01-20 astro-ph.GA

Alessandro Trinca , Alessandro Lupi , Zoltán Haiman , Marta Volonteri , Rosa Valiante , Raffaella Schneider , Roberto Decarli This is my paper · ORCID

classification astro-ph.GA

keywords directcollapseblackholeshigh-redshiftquasarsLyman-WernerradiationholeseedsJWSTobservationssemi-analyticgalaxyformationcosmicdawnquasarenvironments

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

The paper tries to establish that the dense cosmic regions that grow into the bright quasars seen at redshift ~7 are natural factories of direct-collapse black holes — black holes seeded at ~10^5 solar masses by the monolithic collapse of pristine, radiation-sterilized gas. Its central move is to calculate the ultraviolet radiation that dissociates hydrogen molecules (the Lyman-Werner field) locally, halo by halo, rather than as a single uniform background. Doing so, it finds heavy seeds can form as early as z≈22, with 54 forming in the reference model of one quasar's assembly region (24 even under a stricter radiation threshold). About 40 per cent of these seeds survive outside the main galaxy down to z≈7.5, mostly still near their birth mass, where a few should be visible to deep JWST imaging during bright accretion episodes. This turns the direct-collapse scenario into a concrete, testable prediction for quasar fields.

What carries the argument

The key mechanism is the resolved local Lyman-Werner flux: rather than a uniform background, each halo receives the summed UV radiation from all neighbouring star-forming galaxies and AGN, using the three-dimensional halo positions from the cosmological N-body merger tree. This local flux often exceeds the critical threshold in pristine atomic-cooling halos — particularly in 'synchronized pairs', where a close neighbour supplies dissociating radiation — allowing direct collapse at z≈22 when the global background is still far too weak. The second essential ingredient is that seeds form in low-mass satellites that do not merge with the central halo; they therefore survive as almost un-grown fo

What would settle it

A targeted JWST survey of at least five z≈7 quasar fields to F444W≈31 that finds no AGN-dominated companions at projected separations of 5–70 kpc from the quasar would falsify the predicted average of ~1.8 such systems per field.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that direct-collapse black hole seeds form abundantly and early inside the overdensity that builds a z≈7 quasar: 54 seeds with the reference critical Lyman-Werner flux (300 J_21), 24 with a stricter 1000 J_21 threshold, the first appearing at z≈22. Metal enrichment of the intergalactic medium shuts off further direct collapse by z≈14. Of the 54 seeds, 22 survive in satellite halos 5–70 kpc from the central quasar at z≈7.5, most having grown little from the 10^5 M_sun seed mass. The paper then argues these surviving seeds are observable: they alternate between quasar-like and dormant states on few-Myr timescales, and a typical quasar field at z 7

Load-bearing premise

The model spreads metals from each galaxy uniformly across the surrounding intergalactic medium, treating enrichment as a single volume-averaged metallicity; if supernova winds instead quickly pollute every nearby pristine halo out to ~7 kpc, the predicted seed population falls from tens to only about 4–13, depending on the adopted critical flux.

Editorial extensions

If this is right

A z≈7 quasar field should contain ~20 unmerged heavy-seed remnants at separations of 5–70 kpc, most with black hole masses near 10^5 M_sun.
Deep JWST imaging in the F444W band (rest-frame optical) to magnitude ~29–31 should catch an average of ~3–5 active companions per quasar when counting AGN emission alone, with about 1–2 clearly AGN-dominated; a multi-quasar survey can therefore test the model statistically.
Because heavy seeds form early (z≳14) and stop once the intergalactic medium is metal-enriched, the surviving population at z≈7 is a fossil record of the earliest direct-collapse episodes.
A non-detection in dedicated deep surveys would not be empty: the paper argues it would constrain seed masses to be below ~10^5 M_sun or favour alternative seeding channels such as rapid growth of lighter seeds.
The number of seeds is not highly sensitive to the assumed critical Lyman-Werner flux (54 vs 24 between J_crit=300 and 1000 J_21), making the orders-of-magnitude prediction stable.

Reading between the lines

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

The per-halo local LW flux recipe is a general improvement: any merger-tree-based formation model with halo positions could apply it, and doing so is likely to push heavy-seed formation to earlier epochs than homogeneous-background treatments suggest.
The same mechanism should operate in the assembly regions of any massive galaxy at high redshift, so the prediction of fossil heavy seeds around quasars plausibly extends to the environments of the most luminous galaxies at z≈5–6, where JWST has already mapped satellite populations.
Because surviving seeds remain near their birth mass and switch between luminous and dormant states on few-Myr timescales, a two-epoch variability search in NIRCam imaging could detect many more of them than a single-epoch survey would; this is a testable extension the paper itself does not quantify.
If a few such companions are confirmed as broad-line AGNs, their Eddington ratios and spectral shapes could distinguish direct-collapse seeds (~10^5 M_sun, metal-poor surroundings) from light-seed descendants that grew rapidly — giving a direct observational handle on the seeding channel.

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.

Desk Editor's Note

read the letter

Read this as a competent SAM forecast with a concrete JWST angle, not as a settled discovery. The genuinely new piece is coupling CAT to the L19 N-body tree with a spatially resolved LW field, which yields the first quantitative population: 54 DCBH seeds, first formation at z~22, 22 survivors at z~7.5, and a detectability estimate of ~1.8 AGN-dominated systems per quasar field at m_F444W < 31. The comparison to the L19 hydro run, with BH mass tracking within 0.5 dex, is a useful sanity check. The sensitivity tests on Jcrit and metal pollution are honest, and the observational strategy discussion is concrete and current. Credit is due for that.

The soft spot is exactly the one flagged in the stress-test note, and it holds up on reading. The headline numbers come from a run with volume-averaged IGM metallicity. The paper's own 7 pkpc enriched-bubble test reduces the Jcrit=1000 seed count from 24 to 4. Since the authors themselves say Jcrit=1000 is likely more appropriate for realistic SEDs, the combined stringent case is 4 seeds, not tens. The abstract's phrase 'tens ... under more stringent conditions' is misleading unless 'more stringent' means Jcrit alone without the localized enrichment test. The 22-survivor population used for the JWST forecasts inherits this fragility; the detectability numbers would drop if the 4-seed case is closer to reality. This is not an internal inconsistency—the paper is upfront about the risk—but it is load-bearing, and the abstract overstates the robustness.

Two more moderate concerns. First, this is a single halo realization. There is no scatter estimate from varying merger histories, so the 54/24/22 numbers have no error bars. Second, data are only 'shared on reasonable request'; no code or catalogs are released. For a paper whose main claim is a population-level prediction, that is a reproducibility gap.

Who should read this: observers planning JWST searches for companions around z~7 quasars will find the strategy and magnitudes useful, and theorists working on DCBH seeding should know it exists. It deserves a serious referee, because the framework is sound and the sensitivity tests are informative even if the headline needs recalibration. My recommendation: send it to review, but require (a) a weaker or properly qualified abstract, (b) a discussion of single-realization variance, and (c) public data/code. With those changes it would be a solid contribution.

Referee Report

3 major / 5 minor

Summary. The paper couples the CAT semi-analytic model of galaxy/BH evolution with high-resolution N-body merger trees of a ~10^12 Msun halo at z~6 in order to predict when and where direct-collapse black hole (DCBH) seeds form in the environment of a bright high-redshift quasar, and whether their descendants are observable with JWST. The central claims are that, when the local Lyman-Werner (LW) radiation field is resolved, DCBHs form as early as z~22; that a reference model with Jcrit=300 J21 yields 54 seeds, with 24 seeds even for the more realistic Jcrit=1000 J21; that 22 of 54 seeds survive outside the main halo at z~7.5; and that deep JWST imaging/spectroscopy can detect a few active, AGN-dominated heavy-seed descendants per quasar field. The paper includes a careful comparison with the L19 hydrodynamical run and an explicit sensitivity analysis to Jcrit and to localized metal enrichment.

Significance. If the predictions hold, the paper provides a concrete, falsifiable demographic of heavy-seed remnants around z~7 quasars, connecting the DCBH formation scenario to JWST observations. The main strengths are the use of a high-resolution merger tree resolving minihalos, the self-consistent local LW field treatment, the quantitative calibration check against L19 (BH mass agreement within 0.5 dex), and the honest reporting of sensitivity tests for Jcrit and external metal pollution. The detectability forecasts, with predicted Ndet values for different F444W depths, are also usefully concrete. The result is significant for the high-redshift SMBH seeding debate, provided the headline numbers are positioned as conditional on model choices rather than as robust point predictions.

major comments (3)

[§3.3.2, Fig. 4] The fiducial 54-seed count and the 22 survivors at z≈7.5 used in §4 are produced with a single volume-averaged IGM metallicity floor. The paper's own conservative test—7 pkpc enriched bubbles around every star-forming galaxy—reduces the Jcrit=1000 seed count from 24 to 4. Since §3.3.1 states that Jcrit=1000 is 'likely more appropriate for realistic galaxy spectra at these early epochs,' the combined model with realistic Jcrit and localized metal pollution yields about 4 seeds, not 'tens.' This directly weakens the abstract's claim 'Even under more stringent conditions ... tens of DCBHs' and, because the detectability analysis in §4.2 is based on the 22-survivor reference population, the JWST predictions inherit this fragility. The authors should present the combined Jcrit=1000 + 7 pkpc case as the conservative baseline (or at least as a headline variant) and re-derive the expected satell
[§3.3.1] The fiducial critical LW flux is Jcrit=300 J21, yet the same section explains that this value is appropriate only for soft, metal-enriched post-starburst SEDs that are unlikely to dominate at z≳10. The authors justify retaining it for calibration consistency and because DCBHs may form even with incomplete H2 suppression. This is a legitimate model choice, but it means that the '>50 seeds' and '22 survivors' headline numbers are not the physically preferred model. The paper should make Jcrit=1000 the default presentation, with Jcrit=300 shown as a lower-threshold sensitivity case. As written, the abstract and conclusions emphasize the Jcrit=300 numbers while burying the more realistic threshold in a sensitivity subsection.
[§4.2 and Fig. 7] The detectability predictions are based on a single merger-tree realization and on the reference 54-seed / 22-survivor model. The paper acknowledges the single-realization limitation in §5, but the quantitative Ndet values are still quoted as if they were average expectations for a typical quasar. A minimal propagation of the §3.3.2 sensitivity would reduce the seed population by more than an order of magnitude in the conservative case (4 vs 54), and the expected number of AGN-dominated detectable satellites would likely drop below ~0.2 at mF444W<31. The authors should either include a detectability estimate for the conservative combined case or explicitly state that all JWST forecasts apply only to the volume-averaged-enrichment, Jcrit=300 model.

minor comments (5)

[Abstract] The phrase 'Even under more stringent conditions for heavy seed formation, tens of DCBHs are predicted' is inconsistent with the combined stringent case (Jcrit=1000 plus 7 pkpc enriched bubbles) that yields 4 seeds. Please rephrase to specify which conditions produce 'tens' and which produce only a handful.
[Fig. 7] The figure panels (a) and (b) are labeled only in the caption text; the in-figure labels would be clearer if they explicitly said 'AGN only' and 'AGN-dominated total' on each panel.
[§5] The sentence 'they accrete at the Eddington limit for approximately ~60% of the time spent in an active accretion phase, defined as M/MEdd > 0.1' uses M/MEdd where the accretion rate ratio is meant; please use \dot{M}/\dot{M}_Edd.
[Appendix A] The transition from local to global LW flux once the average background exceeds Jcrit is a reasonable approximation, but it should be explicitly stated that this removes spatial fluctuations at late times; otherwise the reader may wonder why the local treatment is not used throughout.
[§2.2.1] The description of the homogeneous IGM metallicity floor is clear, but the two opposing biases of volume averaging (over-enrichment of isolated halos vs under-enrichment of synchronized pairs) would benefit from being summarized in one sentence at the point of the model description, not only in §3.3.2.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the DCBH counts, survival fractions, and JWST detectability estimates are forward outputs of a semi-analytic model, not fits to the predicted quantities, and key assumptions are varied or checked against an independent hydrodynamical simulation.

full rationale

The paper's central claims—54 DCBH seeds, their formation epochs down to z~22, the 22 survivors at z~7.5, and the predicted JWST-detection rates—are genuine outputs of the CAT SAM applied to N-body merger trees. The model parameters, including the SFR efficiencies and the Jcrit threshold, are inputs chosen from prior calibration and literature, not fitted to the DCBH population being predicted. The Lyman-Werner flux is computed forward from the spatial distribution of sources (Eq. A5), and the seed count is then counted, not imposed. The paper explicitly varies the most consequential input: 'we re-ran our simulation with the resolved local LW flux, adopting a higher threshold value of JLW,crit=1000' (§3.3.1), and also tests localized external metal enrichment with 2 and 7 pkpc bubbles (§3.3.2, Fig. 4). These variants change the numbers but do not alter the qualitative conclusion, and they demonstrate that the headline numbers are not fixed by construction. The reliance on the authors' own CAT calibration is real but not circular: CAT was calibrated to reproduce high-redshift UV luminosity functions, SFR densities, and BH mass functions (Trinca et al. 2022), not to produce a specific number of direct-collapse seeds. Moreover, the model is benchmarked against the independent hydrodynamical simulation L19: 'the main properties and evolution of the simulated galaxy remain broadly consistent with previous results from hydrodynamical simulations, without requiring any fine-tuning of the model free parameter outside of the fiducial ranges previously adopted' (§3.1). The weakest point, the volume-averaged IGM metallicity, is acknowledged and tested by the authors themselves ('Neglecting the role of these localized early enrichment events could therefore lead to overestimating the number of potential DCBH formation sites'), but this is a sensitivity limitation, not a circular reduction. No step in the derivation chain uses the target result as an input or renames a fitted quantity as a prediction.

Assumptions & free parameters 8 free parameters · 6 assumptions · 0 invented entities

The central population counts (54/24/22) are controlled by a small set of sub-grid parameters inherited from the CAT calibration and by the assumed DCBH criteria; the paper varies Jcrit and approximates metal pollution as a uniform floor. No new physical entity is introduced; all sensitivity parameters are listed.

free parameters (8)

Star formation efficiency, PopIII (epsilon_SF,PopIII) = 0.15
Set to align with hydro simulations of pristine clusters (Chon et al. 2022); affects LW production and metal enrichment, hence DCBH sites. §2.2.1.
Star formation efficiency, PopII (epsilon_SF,PopII) = 0.05
Calibrated to reproduce high-z galaxy UV luminosity functions and SFR density (Trinca et al. 2022). §2.2.1.
Critical LW flux Jcrit = 300 J21 (fiducial); 1000 J21 (sensitivity)
Fiducial adopted even though text says 1000 is more appropriate for realistic SEDs; chosen to avoid recalibrating CAT. Directly sets number of DCBH hosts. §2.2.2, §3.3.1.
DCBH seed mass Mseed = 10^5 Msun
Postulated heavy seed mass; recent simulations suggest 10^4 Msun, which would reduce detectability. §2.2.2, §5.
Bondi accretion boost alpha = 90
Calibrated to reproduce high-z quasar population; controls growth of seeds and hence final masses. §2.2.2.
Gas core radius R_core = 0.012 Rvir
Assumed density profile normalization; affects BH accretion rates. §2.2.2.
Dust-to-metal ratio = 0.3
Assumed for AGN attenuation in synthetic spectra, affects predicted magnitudes. §4.1.
Metallicity threshold Zcrit = 10^-3.8 Zsun
Boundary between PopIII/PopII and pristine vs enriched halos; sets which halos can host DCBH. §2.2.1.

assumptions (6)

domain assumption LCDM cosmology with standard parameters and the L19 constrained initial conditions reproduce the z=6 ~3e12 Msun halo.
The whole calculation inherits the N-body zoom-in simulation; any error in the merger tree propagates to all final counts. §2.1.
ad hoc to paper A single volume-averaged IGM metallicity floor captures external metal pollution of pristine halos.
Paper itself notes this could over- or underestimate enrichment in synchronized pairs; central counts strongly depend on it. §3.3.2.
domain assumption Atomic cooling halos with Tvir>1e4 K and pristine gas, if irradiated above Jcrit, collapse monolithically to a 1e5 Msun BH without fragmentation.
Standard DCBH criterion, adopted as a binary condition; the paper discusses but does not relax it. §2.2.2.
domain assumption Accretion is Eddington-limited; no super-Eddington phases.
Growth is capped at Eddington; the paper notes super-critical accretion could increase detectable counts. Eq. (5), §5.
domain assumption BH mergers are instantaneous in major mergers; minor-merger BHs become wandering and are not followed.
Affects survival and final BH inventory around z~7. §2.2.2.
domain assumption The halo mass resolution (m_DM=9.19e4 Msun, ~10 particles per minihalo) is sufficient to resolve all DCBH host halos.
Missing halos below resolution would change counts. §2.1.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Hunting the first Cosmic Giants: formation and detectability of Direct Collapse Black Holes around high-redshift quasars." pith.science (2026). https://pith.science/paper/Z6ZNZ4IX

@misc{pith2026260114370,
  author       = {Pith},
  title        = {Pith review of: Hunting the first Cosmic Giants: formation and detectability of Direct Collapse Black Holes around high-redshift quasars},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Z6ZNZ4IX}},
  note         = {Machine review of arXiv:2601.14370}
}

abstract

The rapid emergence of supermassive black holes (SMBHs) in the early Universe poses a challenge to current models of black hole growth. One promising formation pathway is the direct collapse black hole (DCBH) scenario, in which gas in pristine, low-metallicity halos forms supermassive (or quasi-) stars leading to massive black holes seeds under specific environmental conditions. In this work, we investigate the potential host environments of DCBHs by coupling a semi-analytic model tracing BH formation and galaxy co-evolution with high-resolution N-body dark matter merger trees. This allows us to trace the population of DCBHs formed during the hierarchical assembly of a $\sim 10^{12} ~\rm M_\odot$ dark matter halo hosting a bright $10^9 ~\rm M_\odot$ quasar at redshift $z \approx 7$. We find that, when accounting for local fluctuations in the UV radiation field within this early cosmic structure, massive BH seeds can form via direct collapse as early as $z \approx 22$. Even under more stringent conditions for heavy seed formation, tens of DCBHs are predicted to emerge within the simulated overdensity down to $z \sim 14$, at which point metal enrichment of the intergalactic medium inhibits further episodes of direct collapse. A significant fraction of the massive black hole population formed at $z > 14$ is expected to survive in satellite galaxies that do not merge with the central halo down to $z \approx 7$. We show that the existence of such a population of ungrown heavy BH seeds can be probed through deep JWST observations targeting regions surrounding bright high-redshift quasars, and we discuss tailored observational strategies to detect and identify these elusive systems.

Figures

Figures reproduced from arXiv: 2601.14370 by the authors.

**Figure 1.** Evolution of the properties of the main simulated galaxy. From top to bottom we show the BH mass, stellar mass, SFR, and BH accretion rate. cat predictions (orange lines) are compared with the results of the hydrodynamical simulation by L19 (gray lines). For reference, the stellar mass panel also shows the DM halo mass growth and the total stellar mass assembled in the simulated overdensity with dotted and dashed li… view at source ↗

**Figure 2.** [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

**Figure 3.** Redshift distribution of newly formed heavy BH seed for different model variants. Panel (a) shows the results for a homogeneous LW field (blue), while the results including the resolved local flux seen by each galaxy are shown in panels (c) and (e) for a critical threshold of, respectively, JLW,crit = 300 J21 (orange) and 1000 𝐽21 (red). Panel (b) shows the comparison between the corresponding cumulative seed distri… view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Number of heavy seeds formed in the simulation as a function of redshift (upper panel) and related cumulative distribution (lower panel). Orange and yellow histograms show, respectively, the fraction of DCBH host halos having - at the time of seed formation - the close…

**Figure 5.** Figure 5: Projection of the simulated overdensity at 𝑧 ≈ 7.5, with distance from the centre of the structure marked in physical kpc and arcseconds. Star forming galaxies are marked with yellow points, while halos hosting heavy seed descendants or SMBHs (log(MBH/M⊙ ) > 6.5) are h…

**Figure 6.** Figure 6: Spectrum of the quasar satellite galaxy ID48871, highlighted as a green square in [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]

**Figure 7.** Figure 7: Average number of detectable galaxies hosting heavy BH seeds descendants (5 < log(MBH ) < 6.5) in the redshift range 7 < 𝑧 < 8, assuming different magnitude threshold limits 𝑚F444W < 29, 30, 31. The galaxy detectability has been estimated: a) considering only the - int…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

How Massive Can a Population III Starburst Be? Simulating the First Galaxies with High Lyman-Werner Background
astro-ph.GA 2026-03 conditional novelty 6.0 of 10

Strong Lyman-Werner radiation can delay star formation in 10^8-solar-mass halos, enabling Population III starbursts of up to ~10^6 solar masses before internal metal enrichment ends them.

Reference graph

Works this paper leans on

2 extracted references · cited by 1 Pith paper

[1]

R., Iliev I

Ahn K., Shapiro P. R., Iliev I. T., Mellema G., Pen U.-L., 2009, ApJ, 695, 1430 Akins H. B., et al., 2025, ApJ, 991, 37 Amaro-Seoane P., et al., 2023, Living Reviews in Relativity, 26, 2 Begelman M. C., Volonteri M., Rees M. J., 2006, Monthly Notices of the Royal Astronomical Society, 370, 289 Behroozi P. S., Wechsler R. H., Wu H.-Y., 2013, ApJ, 762, 109 ...

arXiv 2009
[5]

MNRAS000, 1–17 (2026)

2026

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [1]

R., Iliev I

Ahn K., Shapiro P. R., Iliev I. T., Mellema G., Pen U.-L., 2009, ApJ, 695, 1430 Akins H. B., et al., 2025, ApJ, 991, 37 Amaro-Seoane P., et al., 2023, Living Reviews in Relativity, 26, 2 Begelman M. C., Volonteri M., Rees M. J., 2006, Monthly Notices of the Royal Astronomical Society, 370, 289 Behroozi P. S., Wechsler R. H., Wu H.-Y., 2013, ApJ, 762, 109 ...

arXiv 2009

[2] [5]

MNRAS000, 1–17 (2026)

2026