[Paper Review] Measuring the HI content of individual galaxies out to the epoch of reionization with [CII]

K. E. Heintz, D. Watson+3 more|arXiv (Cornell University)|Aug 30, 2021

Galaxies: Formation, Evolution, PhenomenaPhysics and Astronomy174 references50 citations

TL;DR

This paper presents a direct calibration of the [C II]-to-H I conversion factor using γ-ray burst afterglows, enabling the first robust measurement of neutral atomic hydrogen (H I) mass in individual high-redshift galaxies up to z ≈6. The method reveals that H I mass exceeds stellar mass at z ≈1 and increases with redshift, with H I fraction reaching ~60% at z ∼6, and shows a universal anti-correlation with gas-phase metallicity across cosmic time.

ABSTRACT

The HI gas content is a key ingredient in galaxy evolution, the study of which has been limited to moderate cosmological distances for individual galaxies due to the weakness of the hyperfine HI 21-cm transition. Here we present a new approach that allows us to infer the HI gas mass $M_{ m HI}$ of individual galaxies up to $z\approx 6$, based on a direct measurement of the [CII]-to-HI conversion factor in star-forming galaxies at $z\gtrsim 2$ using $\gamma$-ray burst afterglows. By compiling recent [CII]-158 $\mu$m emission line measurements we quantify the evolution of the HI content in galaxies through cosmic time. We find that the HI mass starts to exceed the stellar mass $M_\star$ at $z\gtrsim 1$, and increases as a function of redshift. The HI fraction of the total baryonic mass increases from around $20\%$ at $z = 0$ to about $60\%$ at $z\sim 6$. We further uncover a universal relation between the HI gas fraction $M_{ m HI}/M_\star$ and the gas-phase metallicity, which seems to hold from $z\approx 6$ to $z=0$. The majority of galaxies at $z>2$ are observed to have HI depletion times, $t_{ m dep,HI} = M_{ m HI}/{ m SFR}$, less than $\approx 2$ Gyr, substantially shorter than for $z\sim 0$ galaxies. Finally, we use the [CII]-to-HI conversion factor to determine the cosmic mass density of HI in galaxies, $ ho_{ m HI}$, at three distinct epochs: $z\approx 0$, $z\approx 2$, and $z\sim 4-6$. These measurements are consistent with previous estimates based on 21-cm HI observations in the local Universe and with damped Lyman-$\alpha$ absorbers (DLAs) at $z\gtrsim 2$, suggesting an overall decrease by a factor of $\approx 5$ in $ ho_{ m HI}(z)$ from the end of the reionization epoch to the present.

Motivation & Objective

• To overcome the limitation of 21-cm line detection, which is only feasible up to z ≈0.4 for individual galaxies, by developing an alternative tracer of H I gas content.
• To calibrate the [C II]-158 µm line luminosity as a direct tracer of H I mass in high-redshift galaxies, avoiding assumptions about gas physical conditions.
• To quantify the evolution of H I gas content in galaxies from z ≈6 to the present, including its dependence on metallicity and star formation rate.
• To measure the comoving H I mass density ρHI(z) at z ≈0, z ≈2, and z ∼4–6 using the new calibration, and compare with existing estimates.
• To assess the potential of [C II] as a tracer for probing H I content beyond the reach of 21-cm observations, especially during the epoch of reionization.

Proposed method

• Utilize high signal-to-noise X-shooter/VLT spectroscopy of 15 γ-ray burst afterglows at z ≳2 to directly measure H I column densities via Lyman-α absorption and [C II]* column densities via λ1335.7 absorption.
• Derive the empirical [C II]-to-H I conversion factor β[CII] ≡ MHI / L[CII] from the observed abundance ratios in the line-of-sight through the ISM of GRB host galaxies.
• Apply the derived β[CII] calibration to existing [C II]-emitting galaxy samples at z ≈2–6 to infer individual H I masses.
• Use the [C II] luminosity function at z ≈0, z ≈2, and z ∼4–6 to compute the comoving H I mass density ρHI(z) via ρHI(z) = β[CII] × L[CII](z).
• Compare the inferred ρHI(z) with previous estimates from 21-cm surveys and damped Lyman-alpha absorbers (DLAs).
• Investigate the relation between H I gas fraction MHI/M⋆, gas-phase metallicity 12+log(O/H), and H I depletion timescale tdep,HI = MHI/SFR across cosmic time.

Experimental results

Research questions

• RQ1 Can the [C II]-158 µm line be reliably calibrated as a tracer of H I mass in high-redshift galaxies, independent of assumptions about gas conditions?
• RQ2 How does the H I gas mass fraction MHI/M⋆ evolve from z ≈6 to z = 0, and at what redshift does H I mass exceed stellar mass?
• RQ3 Is there a universal relation between H I gas fraction and gas-phase metallicity across cosmic time, from z ≈6 to z = 0?
• RQ4 What is the H I depletion timescale tdep,HI = MHI/SFR at z ≳2, and how does it compare to local galaxies?
• RQ5 How does the comoving H I mass density ρHI(z) evolve from z ∼4–6 to the present, and is it consistent with DLA and 21-cm observations?

Key findings

• The H I mass exceeds the stellar mass M⋆ at z ≈1, and the H I fraction MHI/M⋆ increases from ~20% at z = 0 to ~60% at z ∼6.
• A universal anti-correlation is found between H I gas fraction MHI/M⋆ and gas-phase metallicity 12+log(O/H), valid from z ≈6 to z = 0.
• The majority of galaxies at z ≳2 have H I depletion timescales tdep,HI < 2 Gyr, significantly shorter than the ≈5–10 Gyr seen in local galaxies (z ≈0).
• The comoving H I mass density ρHI(z) at z ∼4–6 is measured as 1.6+0.5−0.4 × 10⁸ M⊙ Mpc⁻³, a factor of ≈5 higher than at z ≈0, consistent with DLA and 21-cm estimates.
• The ρHI(z) evolution is best described by a power-law function ρHI(z) = 3 × 10⁷(1 + z)⁰.⁸⁵ M⊙ Mpc⁻³, predicting a H I mass density at z = 0 that is ~2× lower than previous estimates.
• At z ∼4–6, H I mass is consistent with the total dynamical mass Mdyn on average, indicating H I dominates the ISM mass budget, whereas molecular gas contributes more significantly at z ≈2.