Epoch of Reionization (EoR) Explained

• Epoch of Reionization is the period when the first luminous sources transformed an almost neutral intergalactic medium into a highly ionized state, ending the cosmic dark ages.
• The state of reionization is tracked through observables such as CMB optical depth, Gunn–Peterson troughs, and redshifted 21-cm emission, providing key constraints on the timing and morphology of the process.
• Advanced simulations and multiwavelength observations are used to model ionizing photon production, radiative feedback, and the evolving bubble topology in the early universe.

The Epoch of Reionization (EoR) designates the critical transitional phase in cosmic history during which the intergalactic medium (IGM) was transformed from an almost completely neutral state, following cosmological recombination, to a highly ionized state, primarily under the influence of the first luminous sources. This process marked the conclusion of the Universe’s so-called “dark ages” and set the stage for galaxy evolution and structure formation observable in the late Universe. The EoR is typically delineated to extend from redshifts $z\sim15$–20, corresponding to the birth of the first stars (“cosmic dawn”), to $z\sim5$–6, by which time observations indicate the IGM was nearly fully ionized [(Zaroubi, 2012); (Lidz, 2015); (Choudhury, 2022)].

1. Physical Processes and Timeline of the EoR

The EoR was initiated by the emergence of the first astrophysical sources—Population III stars, Population II galaxies, and potentially accreting black holes—whose ultraviolet (UV) and X-ray radiation generated expanding H II regions in the neutral hydrogen-dominated IGM (Basu, 24 Oct 2025, Alvarez et al., 2019). The global progress of reionization is quantified by the volume-averaged ionized fraction $$x_e(z)\equiv \langle n_{\rm HII}\rangle/\langle n_{\rm H}\rangle$$, which evolves according to the competitive balance between ionizations and recombinations:

$$\frac{dx_e}{dt} = \frac{\dot n_{\rm ion}}{\bar n_{\rm H}} - \frac{x_e}{t_{\rm rec}}$$

with $\dot n_{\rm ion}$ the comoving rate of ionizing photon production, $\bar n_{\rm H}$ the mean hydrogen number density, and $t_{\rm rec}$ the mean recombination time [(Lidz, 2015); (Zaroubi, 2012)].

The minimum mass of star-forming halos ($M_{\rm min}$) and the efficiency of ionizing photon escape ($f_{\rm esc}$) are pivotal parameters. The “inside-out” topology—wherein over-dense, galaxy-rich regions ionize first, and voids ionize last—dominates early and mid-stages, with a transition to “outside-in” closure as recombinations in residual high-density clumps control the tail end (at $x_{\rm HI,v}\sim10^{-4}$) (Giri et al., 2024).

The EoR elapsed over an extended redshift interval. Current modeling constrained by CMB and quasar observations yields $z_{75}=9.2^{+1.2}_{-1.2}$, $z_{50}=8.1^{+1.1}_{-1.0}$, and $z_{25}=7.3^{+1.1}_{-1.0}$ as the redshifts where the global neutral fraction was 75%, 50%, and 25%, respectively (Greig et al., 2016).

2. Theoretical Framework for Modeling Reionization

Most EoR models rely on “photon-counting” equations, coupled to the statistics of structure formation, source physics, and small-scale radiative transfer (Lidz, 2015, Basu, 24 Oct 2025). The governing equation is

$$\frac{d Q_{\rm HII}}{dt} = \frac{\dot n_{\rm ion}}{\langle n_{\rm H}\rangle} - \frac{Q_{\rm HII}}{t_{\rm rec}}$$

where $Q_{\rm HII}$ is the volume-filling factor of H II regions, and $$t_{\rm rec} = [\alpha_B C \langle n_{\rm H}\rangle (1+z)^3]^{-1}$$, with $\alpha_B$ the recombination coefficient (case B) and $C$ the clumping factor of the ionized IGM (Lidz, 2015, Melia et al., 2015).

The ionizing emissivity is parameterized as

$$\dot n_{\rm ion}(z) = f_{\rm esc}\,f_\ast\,N_\gamma\,\frac{df_{\rm coll}}{dt}\,n_{\rm H}$$

where $f_\ast$ is the star formation efficiency, $N_\gamma$ the ionizing photons per stellar baryon, and $f_{\rm coll}$ the collapse fraction of matter into halos above $M_{\rm min}$ (Lidz, 2015).

Radiative transfer and feedback from the evolving ionizing background are incorporated using semi-numerical excursion-set algorithms or full 3D radiative-hydrodynamics simulations (e.g., RAMSES-RT in SPICE (Basu, 24 Oct 2025), pyC$^2$Ray (Giri et al., 2024)). Models require self-consistent calibration to high-$z$ observed UV luminosity functions (UVLFs), quasar Lyman-$\alpha$ forest absorption, and Ly$\alpha$ emitter statistics.

3. Observational Diagnostics and Constraints

A multi-probe strategy is essential to dissect the reionization process (Alvarez et al., 2019):

• CMB Optical Depth ($\tau_e$): Thomson scattering off free electrons suppresses primary temperature anisotropies and imprints a polarization bump on large angular scales (Planck: $\tau_e=0.054 \pm 0.007$) (Reichardt, 2015, Greig et al., 2016, Qin et al., 2020).
• Gunn–Peterson Effect: The saturation of Lyman-$\alpha$ absorption in high-$z$ quasar spectra (the “Gunn–Peterson trough”) sets stringent upper limits on $x_{\rm HI}$ at $z\sim 6$–7 ($x_{\rm HI}\lesssim 0.1$ at $z=6.3$ in GRB 210905A (Fausey et al., 2024)).
• Redshifted 21-cm Line: The differential brightness temperature

$$\delta T_b(\mathbf{x},z) \approx 27\,{\rm mK}\,x_{\rm HI}(\mathbf{x},z)\,(1+\delta_b)\left(\frac{1+z}{10}\right)^{1/2}\left[1-\frac{T_{\rm CMB}(z)}{T_s(\mathbf{x},z)}\right]$$

probes the spatiotemporal topology of reionization through its power spectrum, higher-order statistics, and imaging [(Koopmans et al., 2015); (Zaroubi, 2010)].

• Lyman-$\alpha$ Emitting Galaxies: The rapid decline in Ly$\alpha$ visibility at $z\gtrsim 7$ signifies a jump in the IGM neutral fraction, with modeling indicating $x_{\rm HI}(z=7)\sim 0.5$ (Dijkstra, 2014).
• Fast Radio Burst (FRB) Dispersion Measures: The mean IGM DM as a function of $z$ is sensitive to both the timing and topology (inside-out vs outside-in) of reionization; samples of $\gtrsim10^4$ high-$z$ FRBs can constrain the reionization duration to $\Delta z\sim2$ (Pagano et al., 2021).

A synthesis of Planck $\tau_e$, quasar dark fraction, galaxy UVLFs, Ly$\alpha$ emitter statistics, and kSZ power consistently yields a midpoint $z_{\rm re}\sim7.6\pm0.8$ and duration $\Delta z_{25–75}\sim1.7$ (Greig et al., 2016, Qin et al., 2020).

4. Topological and Statistical Properties of the 21-cm Signal

The 21-cm field is inherently non-Gaussian during the EoR due to the overlapping, evolving H II morphology. The power spectrum $P_{21}(k,z)$ peaks at scales corresponding to typical bubble sizes (e.g., $k_{\rm knee}\sim0.4$–0.6\,Mpc$^{-1}$ when mean island radii are $\sim$20–40\,cMpc) (Giri et al., 2024). The error-covariance of $P_{21}$ is strongly non-Gaussian: the trispectrum term dominates the diagonal and off-diagonal elements at $k \gtrsim 0.1$\,Mpc$^{-1}$ in the late EoR, producing enhanced variance and bin correlations (boost factors up to $\sim 200\times$ at $k\sim1$\,Mpc$^{-1}$, $x_{\rm HI}=0.15$) (Mondal et al., 2016).

To capture LoS evolution (light-cone effect), the evolving power spectrum formalism decomposes the full MAPS $\mathcal{C}_\ell(\nu_1,\nu_2)$ into multipoles $P_{e,q}(k,z)$, with the monopole at large $k$ tracing the global neutral fraction and higher multipoles encoding RSD and anisotropies (Pramanick et al., 27 Mar 2025).

5. Astrophysical Drivers: Sources, Sinks, and Feedback

Early galaxies (including Pop III and Pop II stars in halos down to $T_{\rm vir}^{\min}\sim10^5\,{\rm K}$), dominate the ionizing photon budget. The “reionization Drake parameter” $\zeta=f_{\rm esc}\,f_\ast\,N_\gamma$ incorporates escape fraction, star formation efficiency, and stellar yields (Lidz, 2015, Basu, 24 Oct 2025). High-redshift measurements and simulations suggest $f_{\rm esc}$ of $\sim 0.05$–0.2 and steep UVLF faint-end slopes, though parameter degeneracies remain strong until resolved by 21-cm or next-generation galaxy surveys (Greig et al., 2016).

Feedback from supernovae and radiative processes regulates star formation, modifies the UVLF, and impacts reionization topology and duration (Basu, 24 Oct 2025). Sinks of photons—primarily Lyman-limit systems and self-shielded clumps—set the photon mean free path ($R_{\rm mfp}\sim 10\,$cMpc at $z\gtrsim7$, increasing at late times), limiting bubble growth and controlling the late-time ionizing background (Giri et al., 2024, Lidz, 2015).

Thermodynamics of the IGM is set by photoheating (IGM temperature $T_0\sim$1–2$\times 10^4\,{\rm K}$ post-reionization), Compton cooling, and ongoing photoionizations. Variations in source spectral hardness and X-ray pre-heating (from HMXBs, miniquasars) can drive significant temperature and spin-temperature fluctuations, imprinting signatures in the 21-cm absorption and emission history (Basu, 24 Oct 2025).

6. Current Observational Status and Future Prospects

Direct IGM 21-cm detections are pending. Statistical upper limits from LOFAR, MWA, PAPER, HERA, and GMRT approach $\Delta^2_{21}(k)\leq(100\,{\rm mK})^2$ at $z=7$–9 but remain above the fiducial EoR signal [(Koopmans et al., 2015); (Zaroubi, 2010)]. Forthcoming SKA-Low, HERA, and pathfinder experiments are forecast to deliver high-S/N power spectra, direct imaging of $\sim$10–100\,cMpc-scale ionized regions at S/N$\gtrsim 5$ per $k$ bin, and cross-correlation measurements with galaxy and LIM surveys (Pramanick et al., 27 Mar 2025).

Observational “boundary conditions” at the EoR onset are within reach: measuring the turnover in the Pearson cross-correlation between 21-cm brightness and star-formation tracers at mean ionized fraction $\bar x_{\rm HII}\sim1\%$–10\% provides a robust anchor on reionization’s start (Libanore et al., 10 Sep 2025). The closing phase is accessible via detection of knee features in $P_{21}$ and mapping large residual “neutral islands” through Lyman-$\alpha$ dark troughs, as simulated in (Giri et al., 2024).

Multiwavelength synergy—combining high-$z$ galaxies, 21-cm tomography, LIM, CMB polarization/kSZ, and FRB DM mapping—will enable the separation of timing, topology, and astrophysical drivers, breaking extant parameter degeneracies and assembling a comprehensive narrative of the EoR (Alvarez et al., 2019, Pagano et al., 2021).

---

Table 1: Representative Redshifts for Key EoR Milestones

Ionization State	Redshift (1σ)	Reference
75% neutral ($x_{\rm HI}=0.75$)	$z_{75}=9.2^{+1.2}_{-1.2}$	(Greig et al., 2016)
50% neutral ($x_{\rm HI}=0.50$)	$z_{50}=8.1^{+1.1}_{-1.0}$	(Greig et al., 2016)
25% neutral ($x_{\rm HI}=0.25$)	$z_{25}=7.3^{+1.1}_{-1.0}$	(Greig et al., 2016)
Complete ($x_{\rm HI}<10^{-3}$)	$z\approx6$	(Fausey et al., 2024, Giri et al., 2024)

---

7. Open Questions and Next Steps

Despite significant progress, open issues persist regarding the faint end of the ionizing source population, the contribution of hard-spectrum X-ray sources, the nature and evolution of photon sinks, the impact of feedback mechanisms, and the topology of the late EoR transition (“inside-out” to “outside-in”) (Basu, 24 Oct 2025, Giri et al., 2024). The imminent deployment of new instrumentation for deep 21-cm, FRB, LIM, and galaxy surveys, coupled with advanced multi-physics simulations, is expected to drive the field toward percent-level constraints on reionization’s chronology and governing physics (Alvarez et al., 2019, Koopmans et al., 2015).