Probing Ultralate Reionization: Direct Measurements of the Mean Free Path over 5 < z < 6

Abstract

The mean free path of ionizing photons, λ mfp, is a critical parameter for modeling the intergalactic medium (IGM) both during and after reionization. We present direct measurements of λ mfp from QSO spectra over the redshift range 5 &lt; z &lt; 6, including the first measurements at z ≃ 5.3 and 5.6. Our sample includes data from the XQR-30 VLT large program, as well as new Keck/ESI observations of QSOs near z ∼ 5.5, for which we also acquire new [C ii] 158 μm redshifts with ALMA. By measuring the Lyman continuum transmission profile in stacked QSO spectra, we find λ mfp = 9.33 − 1.80 + 2.06 , 5.40 − 1.40 + 1.47 , 3.31 − 1.34 + 2.74 , and 0.81 − 0.48 + 0.73 pMpc at z = 5.08, 5.31, 5.65, and 5.93, respectively. Our results demonstrate that λ mfp increases steadily and rapidly with time over 5 &lt; z &lt; 6. Notably, we find that λ mfp deviates significantly from predictions based on a fully ionized and relaxed IGM as late as z = 5.3. By comparing our results to model predictions and indirect λ mfp constraints based on IGM Lyα opacity, we find that the evolution of λ mfp is consistent with scenarios wherein the IGM is still undergoing reionization and/or retains large fluctuations in the ionizing UV background well below redshift 6.

Constraints on the Evolution of the Ionizing Background and Ionizing Photon Mean Free Path at the End of Reionization

“…The single power-law evolution with redshift that Worseck et al (2014) find to be a good fit at z = 2-5 is disfavored at z  5.4, with our constraints indicating a more rapid decline to higher redshift. Intriguingly, our constraints are in good agreement with the z = 6-5 trend found by Becker et al (2021), Zhu et al (2023, effectively bridging the gaps between the more direct quasar-stacking measurements. In the right panel of Figure 8, we compare our constraints to theoretical predictions for the evolution of λ mfp .…”

“…In the left panel of Figure 8, we compare our constraints to the measurements from Worseck et al (2014), Becker et al (2021), andZhu et al (2023), as well as the lower limit from Bosman (2021). The single power-law evolution with redshift that Worseck et al (2014) find to be a good fit at z = 2-5 is disfavored at z  5.4, with our constraints indicating a more rapid decline to higher redshift.…”

“…At the low-redshift end of our data set, z ∼ 5, our ability to constrain λ mfp is limited by the insensitivity of the Lyα forest to long mean free paths, as the resulting fluctuations in the radiation field become weak compared to the intrinsic fluctuations already imprinted by the IGM density field. However, at these redshifts, the direct stacking of quasar spectra appears to be a far more sensitive probe with fewer underlying assumptions and systematic uncertainties (Worseck et al 2014;Becker et al 2021;Zhu et al 2023). At the highredshift end, z ∼ 6, we are instead limited by the relative sparsity of sightlines and the poor spatial resolution of our ionizing background simulations.…”

Lyman Continuum Emission from Active Galactic Nuclei at 2.3 ≲ z ≲ 3.7 in the UVCANDELS Fields

“…The end of the reionization epoch has been constrained to z ∼ 6 by multiple studies of Gunn-Peterson troughs seen in active galactic nucleus (AGN) spectra (e.g., Becker et al 2001;Djorgovski et al 2001;Fan et al 2006;Goto et al 2011), though more recent evidence has been presented favoring a longer epoch ending around z ∼ 5.3-5.5 (e.g., Qin et al 2021;Bosman et al 2022;Zhu et al 2023). The leading candidates for the main producers of hydrogen-ionizing radiation (λ 912 Å, i.e., Lyman continuum (LyC)) are thought to be low-mass, starforming galaxies which existed in abundance at z  6 (e.g., Bouwens et al 2015;Parsa et al 2018;Saldana-Lopez et al 2023).…”