Reconciling chemical-enrichment models with the observed high-redshift intergalactic-medium abundance pattern ($[{\rm C/H}]$, $[{\rm Si/C}]$, $[{\rm O/C}]$) and the transport of metals into the lowest-density regions

Calura & Matteucci (2006) computed the cosmic evolution of metal densities with detailed chemical-evolution models for spheroidal and dwarf irregular galaxies, and found that while their predictions grossly reproduce the oxygen abundance of the intergalactic medium (IGM) at high redshift, they overpredict the IGM carbon abundance $[{\rm C/H}]$ and underpredict the observed $[{\rm Si/C}]$ and $[{\rm O/C}]$ ratios at $z\gtrsim 2$. They also showed, via order-of-magnitude stall-radius arguments, that dispersing metals to $\sim 1$ Mpc within $\sim 1$ Gyr requires galactic-wind velocities of order 1500 km/s, leaving open whether realistic hydrodynamic transport and mixing can enrich the lowest-density IGM. The tension has not been resolved; it has been sharpened: observations now firmly establish $[{\rm C/H}]\approx -3.5$ at $z\sim 3$--$4.5$, strongly inhomogeneous enrichment, and $\alpha$-enhanced $[{\rm Si/C}]$, while modern hydrodynamical simulations still cannot simultaneously reproduce the observed C IV and Si IV properties and underpredict C IV and Si IV at $z\approx 3.5$. The open problem is to determine quantitatively which combination of stellar carbon yields, initial-mass-function assumptions, ultraviolet-background ionization corrections, dust depletion, and spatially inhomogeneous or biased enrichment reconciles models with the observed IGM abundance pattern, and to determine the transport physics that governs whether metals reach the lowest-density gas.

Background

The intergalactic medium (IGM) is the diffuse baryonic gas between galaxies, which at redshifts $2\lesssim z\lesssim 6$ (earlier cosmic epochs; $1+z$ is the stretch factor of the cosmic scale) is observed mainly through the Lyman-$\alpha$ forest: a thicket of neutral-hydrogen (H I) absorption lines imprinted on spectra of quasars — brilliantly luminous, accreting supermassive black holes in galactic nuclei — by overdense regions of a photoionized cosmic web. Heavier elements ('metals') synthesized in stars are ejected from galaxies by supernova-driven galactic winds and pollute this gas; they are detected through associated absorption lines of ions such as C IV (triply ionized carbon), Si IV (triply ionized silicon), and O VI (quintuply ionized oxygen). Abundances are written in the standard bracket notation, e.g. $[{\rm C/H}] = \log_{10}[(n_{\rm C}/n_{\rm H})/(n_{\rm C}/n_{\rm H})_\odot]$, where $n_X$ is the number density of element $X$ and $\odot$ denotes the solar value; abundance ratios such as $[{\rm Si/C}]$ and $[{\rm O/C}]$ encode which stellar populations produced the metals: $\alpha$-elements (O, Mg, Si) are made primarily by core-collapse supernovae — the explosions of massive stars — on short time-scales, while a large fraction of carbon comes from low- and intermediate-mass stars and from massive stars with rotation and binary mass loss, with yields that remain uncertain, especially at low metallicity.

Converting observed ionic column densities (the amount of a given ion intercepted per unit area along the line of sight) into elemental abundances requires ionization corrections, which depend on the shape of the metagalactic ultraviolet background (UVB) — the extragalactic ionizing radiation field — itself uncertain and time-varying, notably during helium reionization (the epoch, roughly $z\sim 3$, when quasar radiation doubly ionized intergalactic helium); abundance-ratio inferences are therefore degenerate with the assumed ionizing spectrum. A further ingredient is the stellar initial mass function (IMF), the distribution of stellar birth masses. Calura & Matteucci (2006) built chemo-photometric chemical-evolution models (models following gas inflow, star formation, stellar nucleosynthesis with finite stellar lifetimes, and galactic winds for each galaxy type) for spheroids (elliptical and bulge-dominated galaxies) and dwarf irregular galaxies (small, gas-rich, slowly evolving systems), under a pure luminosity-evolution assumption (fixed galaxy comoving number density, no spiral-disc winds, no Population III — first-generation — stars), and integrated the metals ejected into the IGM over cosmic time. They found: (i) the bulk of IGM enrichment comes from spheroids, with dwarf irregulars negligible; (ii) predicted $[{\rm O/H}]$ broadly matches observations, but predicted $[{\rm C/H}]$ exceeds the observed high-redshift carbon measurements; (iii) predicted $[{\rm Si/C}]$ and $[{\rm O/C}]$ at $z\gtrsim 2$ fall below observational inferences; and (iv) by their stall-radius estimates (the distance a spherical, wind-blown metal bubble travels before stalling in the ambient IGM, assuming uniform IGM pressure and constant wind speed), enriching regions $\sim 1$ Mpc away within $\sim 1$ Gyr would require wind velocities $v_W\sim 1500$ km/s. They noted that a top-heavy (high-mass-dominated) early IMF could raise the predicted $[{\rm Si/C}]$ and $[{\rm O/C}]$ but would overproduce the total local metal budget, closing that escape route, and they proposed several candidate explanations for the abundance tension — observational systematics (UVB-shape-dependent ionization corrections; C IV losing sensitivity as a metallicity tracer in gas near the cosmic mean density because carbon is ionized to higher states there, while O VI remains sensitive), incorrect or incomplete carbon yields, carbon locked into dust grains (depletion), spatial inhomogeneities and biased sightline sampling, or IMF variations — but could not decide among them, and called for numerical hydrodynamic simulations adopting their chemical-evolution prescriptions.

Later work sharpened every ingredient without resolving the tension. Simcoe (2011) measured ionization-corrected $[{\rm C/H}] = -3.55$ at $4.0 < z < 4.5$, showing the inferred abundance evolution is robust to, but its strength depends on, the assumed UVB spectrum. Fumagalli (2013) reviewed the observations: a median IGM abundance $[{\rm C/H}]\sim -3.5$ at $z\sim 3$ and a five-dex (a dex is a factor of ten in logarithmic units) metallicity spread among denser Lyman-limit absorbers — clouds dense enough to absorb all photons shortward of the 912 Å hydrogen ionization edge — confirming strong spatial inhomogeneity. D'Odorico et al. (2016), from an ultra-high signal-to-noise spectrum taken with UVES (the ultraviolet–visual echelle spectrograph on the Very Large Telescope), constrained the volume filling factor of IGM gas enriched to $\log Z/Z_\odot\geq -3$ to only $\sim 10$–$13$ per cent, favouring pollution by low-mass galaxies and/or early enrichment. On the theoretical side, Tescari et al. (2010) implemented exactly the requested coupling of a detailed chemical-evolution enrichment scheme to cosmological hydrodynamic simulations with various wind prescriptions, obtaining only partial agreement with observed C IV statistics. Turner et al. (2016) showed that the EAGLE simulations (a large cosmological hydrodynamical simulation project) at $z\approx 3.5$ underpredict the observed C IV and Si IV optical depths (by up to $\sim 1$ dex in dense gas) while matching O VI, and that UVB softening, resolution, unresolved turbulence, and increased abundances each fail to close the gap. D'Odorico et al. (2022) found that simulations that reproduce the observed Si IV properties underpredict C IV, and vice versa. Shull et al. (2014) measured $[{\rm Si/C}]\approx 3$ times solar in the low-redshift IGM. On transport, Cen & Chisari (2011) found superwind influence limited to $\gtrsim 0.5$ Mpc and overdensities $> 10$, Pratt et al. (2018) found O VI extending at most $\sim 0.6$ Mpc from low-redshift galaxies, with metals not leaving galaxy groups, and Sharma et al. (2026) measured the baryon content of cosmic-void interiors for the first time (via dispersion measures of fast radio bursts — millisecond extragalactic radio flashes whose frequency-dependent arrival times measure the electron column along the path), leaving the metal content of the emptiest regions still uncharacterized. Carbon yields remain unsettled: Ma et al. (2025) showed the carbon yields of binary-stripped massive stars are dominated by metallicity dependence and cannot be extrapolated from solar metallicity. Tie et al. (2022) noted that background-IGM metallicity measurements at $z\sim 4$ are sparse and mutually inconsistent, and developed a clustering-based technique to constrain both metallicity and enrichment topology. The stage is therefore: firm observations of a low, inhomogeneous, $\alpha$-enhanced IGM abundance pattern; enrichment models and hydrodynamic simulations that still fail to reproduce the carbon and silicon absorption simultaneously; and an undetermined decomposition of the discrepancy among yields, IMF, ionization physics, dust, and transport.

Problem Statement

Determine the physical origin of the discrepancy, identified by Calura & Matteucci (2006), between chemical-evolution predictions and the observed abundances and abundance ratios of the high-redshift intergalactic medium, and establish the mechanism by which metals are transported into the lowest-density IGM regions. Specifically, the unresolved question is which combination of (a) incorrect or incomplete stellar nucleosynthetic yields, particularly for carbon, (b) initial-mass-function assumptions (noting that a strongly top-heavy early IMF is excluded by the local cosmic metal budget), (c) ultraviolet-background shape and ionization-correction systematics in converting C IV, Si IV, and O VI absorption into elemental abundances, (d) depletion of carbon onto dust grains, and (e) spatially inhomogeneous enrichment with biased sightline sampling, reconciles the predicted IGM $[{\rm C/H}]$ (overpredicted relative to observed values of order $-3.5$ at $z\gtrsim 2$) and the predicted $[{\rm Si/C}]$ and $[{\rm O/C}]$ (underpredicted relative to observational inferences at $z\gtrsim 2$, and to the $\alpha$-enhanced, roughly three-times-solar Si/C found in the low-redshift IGM) with observations, while remaining consistent with the observed redshift evolution of these abundances, the measured volume-filling-factor and detection-rate constraints on the enriched fraction of IGM gas, and the local metal inventory. As part of the same objective, establish whether realistic hydrodynamic transport, interaction with large-scale structure, and mixing — beyond the order-of-magnitude spherical stall-radius estimates — can deliver metals from individual galaxies to the lowest-density regions of the IGM on the observationally required time-scales, as probed for example by the limited ($\sim 0.5$–$0.6$ Mpc) present-day reach of metals around galaxies and by the still-unassayed metal content of cosmic-void interiors. A complete answer decomposes the discrepancy quantitatively among the candidate causes and identifies the enrichment history and transport physics consistent with all of these constraints simultaneously.

References

1. Calura, F., Matteucci, F. (2006). Cosmic evolution of metal densities: the enrichment of the intergalactic medium. Monthly Notices of the Royal Astronomical Society 369, 465–478. DOI: 10.1111/j.1365-2966.2006.10329.x, https://arxiv.org/abs/astro-ph/0602311
2. Tescari, E., Viel, M., D'Odorico, V., Cristiani, S., Calura, F., Borgani, S., Tornatore, L. (2010). Cosmic evolution of the C IV in high-resolution hydrodynamic simulations. Monthly Notices of the Royal Astronomical Society 411, 826–848. DOI: 10.1111/j.1365-2966.2010.17761.x, https://arxiv.org/abs/1007.1628
3. Simcoe, R. A. (2011). The Carbon Content of Intergalactic Gas at z=4.25 and its Evolution Toward z=2.4. The Astrophysical Journal 738, 159. DOI: 10.1088/0004-637x/738/2/159, https://arxiv.org/abs/1106.2810
4. Cen, R., Chisari, N. (2011). Star Formation Feedback and Metal-Enrichment History of the Intergalactic Medium. The Astrophysical Journal 731, 11. DOI: 10.1088/0004-637x/731/1/11, https://arxiv.org/abs/1005.1451
5. Fumagalli, M. (2013). Metal abundances in the high-redshift intergalactic medium. arXiv:1312.5321, https://arxiv.org/abs/1312.5321
6. Shull, J. M., Danforth, C. W., Tilton, E. M. (2014). Tracing the Cosmic Metal Evolution in the Low-Redshift Intergalactic Medium. The Astrophysical Journal 796, 49. DOI: 10.1088/0004-637x/796/1/49, https://arxiv.org/abs/1409.6720
7. D'Odorico, V., Cristiani, S., Pomante, E., Carswell, R. F., Viel, M., Barai, P., Becker, G. D., Calura, F., Cupani, G., Fontanot, F., Haehnelt, M. G., Kim, T.-S., Miralda-Escudé, J., Rorai, A., Tescari, E., Vanzella, E. (2016). Metals in the z~3 intergalactic medium: results from an ultra-high signal-to-noise ratio UVES quasar spectrum. Monthly Notices of the Royal Astronomical Society 463, 2690–2707. DOI: 10.1093/mnras/stw2161, https://arxiv.org/abs/1608.06116
8. Turner, M. L., Schaye, J., Crain, R. A., Theuns, T., Wendt, M. (2016). Observations of metals in the z≈3.5 intergalactic medium and comparison to the EAGLE simulations. Monthly Notices of the Royal Astronomical Society 462, 2440–2464. DOI: 10.1093/mnras/stw1816, https://arxiv.org/abs/1605.08700
9. Pratt, C. T., Stocke, J. T., Keeney, B. A., Danforth, C. W. (2018). The Spread of Metals into the Low-redshift Intergalactic Medium. The Astrophysical Journal 855, 18. DOI: 10.3847/1538-4357/aaaaac, https://arxiv.org/abs/1706.05103
10. Tie, S. S., Hennawi, J. F., Kakiichi, K., Bosman, S. E. I. (2022). Constraining IGM enrichment and metallicity with the C IV forest correlation function. Monthly Notices of the Royal Astronomical Society 515, 3656–3673. DOI: 10.1093/mnras/stac2021, https://arxiv.org/abs/2201.10571
11. D'Odorico, V., Finlator, K., Cristiani, S., Cupani, G., Perrotta, S., Calura, F., Cènturion, M., Becker, G., Berg, T. A. M., Lopez, S., Ellison, S., Pomante, E. (2022). The evolution of the Si IV content in the Universe from the epoch of reionization to cosmic noon. Monthly Notices of the Royal Astronomical Society 512, 2389–2401. DOI: 10.1093/mnras/stac545, https://arxiv.org/abs/2202.12206
12. Ma, J.-Z., Farmer, R., de Mink, S. E., Laplace, E. (2025). Carbon from massive binary-stripped stars: Effect of metallicity. Astronomy & Astrophysics 704, A179. DOI: 10.1051/0004-6361/202555350, https://arxiv.org/abs/2505.02918
13. Rowlands, S., Davies, R. L., Ryan-Weber, E., Keating, L. C., Sebastian, A. M., Becker, G. D., Bischetti, M., Bosman, S. E. I., Chen, H., Davies, F. B., D'Odorico, V., Gaikwad, P., Gallerani, S., Haehnelt, M. G., Kulkarni, G., Meyer, R. A., Welsh, L., Zhu, Y. (2026). E-XQR-30: Evidence for an Increase in the Ionization State of Metal Absorbers from z6 to z2. Monthly Notices of the Royal Astronomical Society 546, staf2254. DOI: 10.1093/mnras/staf2254, https://arxiv.org/abs/2601.02634
14. Sharma, K., Ravi, V., Anbajagane, D., Coulton, W. R., Krause, E., Schuster, N., Pisani, A., McCarty, S., Connor, L., Ferraro, S., Hamaus, N., Pranjal R. S. (2026). Baryons in the Darkest Sites of the Universe. arXiv:2605.01994, https://arxiv.org/abs/2605.01994