Evolution of neutral oxygen during the epoch of reionization and its use in estimating the neutral hydrogen fraction

Abstract

Abstract We use synthetic sightlines drawn through snapshots of the Technicolour Dawn simulations to explore how the statistics of neutral oxygen (${\mathrm{O}\,{\small I}}$) absorbers respond to hydrogen reionization. The ionization state of the circumgalactic medium (CGM) initially roughly tracks that of the intergalactic medium, but beginning at z = 8 the CGM grows systematically more neutral owing to self-shielding. Weak absorbers trace diffuse gas that lies farther from haloes, hence they are ionized first, whereas stronger systems are less sensitive to reionization. The overall ${\mathrm{O}\,{\small I}}$ covering fraction decreases slowly with time owing to competition between ongoing enrichment and gradual encroachment of ionization fronts into increasingly overdense gas. While the declining covering fraction is partially offset by continued formation of new haloes, the ionization of the diffuse gas causes the predicted line-of-sight incidence rate of ${\mathrm{O}\,{\small I}}$ absorbers to decline abruptly at the overlap epoch, in qualitative agreement with observations. In comparison to the recently observed equivalent width (EW) distribution at z ≈ 6, the simulations underproduce systems with $\mathrm{EW} \ge 0.1 \mathring{\rm A}$, although they reproduce weaker systems with $\mathrm{EW} \ge 0.05 \mathring{\rm A}$. By z ≈ 5, the incidence of $\mathrm{EW} \lt 0.1 \mathring{\rm A}$ systems are overproduced, consistent with previous indications that the simulated ionizing background is too weak at z < 6. The summed column densities of $\mathrm{Si}\,{\small II}$ and $\mathrm{Si}\,{\small IV}$ trace the total oxygen column, and hence the ratio of the ${\mathrm{O}\,{\small I}}$ and $\mathrm{Si}\,{\small II}+ \mathrm{Si}\,{\small IV}$ comoving mass densities traces the progress of reionization. This probe may prove particularly useful in the regime where $x_{\mathrm{H}\,{\small I}} \gt 10{{\ \rm per\ cent}}$.