Heavily Reddened Quasars

• Heavily Reddened Quasars (HRQs) are hyperluminous active galactic nuclei with severe dust extinction that reddens their optical/UV emission.
• They are detected using extreme near- and mid-infrared color cuts and confirmed through spectroscopic broad-line identification and SED fitting with SMC-like extinction laws.
• HRQs represent a transient, merger-driven blow-out phase, offering critical insights into rapid SMBH growth, AGN feedback, and the co-evolution of galaxies during cosmic noon.

Heavily Reddened Quasars (HRQs) are a population of active galactic nuclei (AGN) whose optical and ultraviolet (UV) continua are strongly attenuated by dust, yielding large extinction parameters (typically $E(B-V) \gtrsim 0.5$) and prominent red colors in near-infrared (NIR) selection bands. Surveys over the past decade leveraging wide-field NIR and mid-infrared (MIR) imaging have established that HRQs are a hyperluminous ($L_{\rm bol} \sim 10^{46}–10^{48}$ erg s$^{-1}$), high-mass ($M_{\rm BH} \sim 10^9–10^{10}\,M_\odot$) population distinct from classical blue, optically-selected quasars. HRQs occupy a brief, feedback-driven "blow-out" phase, often coincident with major galaxy mergers, extreme star formation, and rapid black hole growth during cosmic noon ($z\sim1–3$) (Glikman et al., 2012, Glikman et al., 2022, Banerji et al., 2015).

1. Selection Criteria, Reddening Laws, and Survey Strategies

HRQs are identified by their extreme NIR colors, typically via $(J-K)_{\rm Vega} > 2.5$ or $(J-K)_{\rm AB} > 1.5$, and elevated MIR excesses, such as $(W1-W2)_{\rm Vega} > 0.85$ (Banerji et al., 2012, Banerji et al., 2012). These color cuts select objects with sight-line dust columns $A_V \sim 2–6$ mag, well above the range typical for UV/optical-selected QSOs. Radio matching (e.g., FIRST 1.4 GHz) and WISE MIR surveys enable both radio-selected and radio-quiet HRQ samples (Glikman et al., 2012, Glikman et al., 2022). The extinction law governing HRQs' SEDs is consistently found to resemble the Small Magellanic Cloud (SMC) curve [Gordon & Clayton 1998; Pei 1992], characterized by $k(\lambda)$ and negligible 2175 Å bump (Glikman et al., 2012, Banerji et al., 2012, Jiang et al., 2013).

Spectroscopic confirmation requires broad permitted lines (e.g., H$\alpha$, H$\beta$, C IV) to distinguish Type 1 quasars. SED-fitting procedures employ composite quasar templates, which are reddened by trial $E(B-V)$ values, with optimal fits found via $\chi^2$ minimization across optical/NIR spectra and photometry, excluding regions dominated by emission lines or noise (Glikman et al., 2012, Banerji et al., 2012, Glikman et al., 2013).

2. Physical Properties: Luminosity, Black-Hole Mass, and Demographics

HRQs span redshift $z \sim 0.1–3$ with measured $E(B-V) \sim 0.1–1.8$, reaching up to $A_V \gtrsim 6$ mag for the most extreme objects (Banerji et al., 2015, Stepney et al., 24 Jan 2026). Virial black-hole mass estimates and extinction-corrected luminosities consistently place HRQs in the hyperluminous regime, matching or exceeding the brightest unobscured QSOs. For example,

• $M_{\rm BH} \sim 10^9–10^{10}\,M_\odot$
• $L_{\rm bol} \sim 10^{47}–10^{48}$ erg s$^{-1}$ (Banerji et al., 2012, Banerji et al., 2015, Stepney et al., 2024)

Their space density, after correcting for extinction and selection biases, approaches or exceeds that of blue quasars at the highest luminosities (e.g., $M_i < –29$), but drops rapidly at lower luminosities (Banerji et al., 2012, Banerji et al., 2015). The fraction of luminous QSOs in the HRQ phase is strongly luminosity-dependent, rising to $>$40% at $\log L_{\rm bol} > 46.5$ and falling below 5% at lower $L_{\rm bol}$ (Glikman et al., 2022).

Surface densities in radio/NIR-selected samples (e.g., FIRST-2MASS, UKIDSS) are $f_{\rm HRQ} \sim 15–20\%$ above $K<15$ mag; in purely MIR-selected samples (WISE-2MASS), HRQs constitute $40\%$ at the highest $L_{\rm bol}$ (Glikman et al., 2012, Glikman et al., 2022, Glikman et al., 2013). However, deeper K-band and MIR observations are required to recover the most heavily reddened sources at high redshift.

3. Spectral Energy Distributions, Scattered Light, and Hot Dust Paucity

The rest-UV to IR SEDs of HRQs routinely display "V-shaped" structure: a strongly reddened optical/NIR continuum, suppressed hot dust emission in the 2–4 μm regime, and a blue excess in the far-UV (Stepney et al., 24 Jan 2026, Stepney et al., 2024). SED decomposition reveals that $>80\%$ of HRQs show a scattered-UV component, with the scattering fraction $F_{\rm UV}$ typically in the range $0.03–1\%$ of the intrinsic luminosity (median $\sim0.26\%$), and a significant population with $F_{\rm UV} \ll 1\%$ due to selection bias (i.e., objects with higher $F_{\rm UV}$ are excluded by $(i-K)$ color cuts) (Stepney et al., 24 Jan 2026).

The hot dust emission parameter, defined as $$f_{\rm hot} = (L_{\rm Dust}/L_{\rm Disk})|_{2\,\mu{\rm m}}$$, is suppressed in HRQs compared to blue QSOs ($\langle f_{\rm hot} \rangle_{\rm HRQ}=1.6 \pm 0.8$ vs. $4.4 \pm 2.1$ for blue QSOs; $p=2\times10^{-30}$), indicating a reduced covering factor of torus dust or partial depletion—consistent with AGN feedback evacuating the inner obscuring medium (Stepney et al., 24 Jan 2026, Stepney et al., 2024). The UV scattering fraction anti-correlates with the line-of-sight extinction $E(B-V)$ and correlates with the hot dust emission, favoring a scenario in which the scattering medium is the sublimation-radius dust (Stepney et al., 24 Jan 2026).

A subset of HRQs with strong V-shaped SEDs and minimal hot dust emission match the criteria for JWST "Little Red Dots" (LRDs), representing cosmic-noon analogues of these rare high-$z$ objects and further demonstrating the evolutionary and feedback-driven interpretation (Stepney et al., 2024, Stepney et al., 24 Jan 2026).

4. Host Galaxies: Star Formation, ISM Properties, and Mechanisms of Obscuration

ALMA and SCUBA-2 observations of HRQs show that these systems reside in gas-rich host galaxies, often undergoing major mergers, and exhibit intense star formation rates (SFRs $\sim 10^2–10^4\,M_\odot\,{\rm yr}^{-1}$). Molecular gas reservoirs are massive ($M_{\rm gas}\sim10^{10}–10^{11}\,M_\odot$), and dust masses reach $M_{\rm dust}\sim10^8–10^9\,M_\odot$ (Banerji et al., 2016, Banerji et al., 2018, Wethers et al., 2020).

In the rest-UV and H$\alpha$ band, unobscured SFRs of $25–365\,M_\odot\,{\rm yr}^{-1}$ (mean $130\pm95\,M_\odot\,{\rm yr}^{-1}$) are detected in HRQ host galaxies, with a positive correlation with bolometric luminosity, indicating coeval bulge/bh growth (Wethers et al., 2018). Spatially, star formation regions lie within $6–8\,$kpc of the nucleus, and many HRQs are unresolved at $<6\,$kpc scale in IFU data (Alaghband-Zadeh et al., 2016, Banerji et al., 2021). In the FIR/submm, stacked analyses limit average SFRs in undetected HRQs to $<880\,M_\odot\,{\rm yr}^{-1}$ (Wethers et al., 2020).

Gas-phase properties inferred from PDR models show extreme densities ($n_{\rm H}\sim 10^5–10^6\,{\rm cm}^{-3}$) and high enrichment (evidenced by elevated [CI], CO, and far-infrared luminosities), consistent with major merger triggers and rapid burst phases (Banerji et al., 2018, Banerji et al., 2016). ALMA CO imaging confirms both disk and close-separation merger morphologies, with major variation in gas fraction ($f_{\rm gas}\sim6\%–45\%$), signaling diversity in evolutionary stage (Banerji et al., 2021).

5. Multi-phase Outflows, AGN Feedback, and the Evolutionary "Blow-out" Scenario

Spectroscopic studies reveal that HRQs host powerful ionized outflows, with [O III]$\lambda5008$ line profiles showing broad wings to $\sim2500\,{\rm km\,s}^{-1}$ in $40\%$ of cases (Temple et al., 2019, Stepney et al., 2024). However, when matched in luminosity and redshift, the [O III] kinematics of HRQs are statistically indistinguishable from those of blue quasars. This supports a model in which feedback-driven winds persist for some time after the dust cocoon disperses (Temple et al., 2019).

Rest-UV spectra detect multiple narrow and broad absorption systems (C IV, N V, Si IV) at up to $v>1000\,{\rm km\,s}^{-1}$; broad-line blueshifts and kinetic power estimates ($\dot{E}_k \sim 0.001\,L_{\rm Bol}$) confirm multi-phase AGN-driven winds (Stepney et al., 2024). X-ray studies corroborate this, with HRQs found in the so-called "forbidden" $N_{\rm H}$-$\lambda_{\rm Edd}$ region, where radiation pressure is expelling dusty gas on $10^2$ pc scales. The inferred blow-out duration is of order $2\times10^5$ years, and HRQs comprise $\sim15–20\%$ of the luminous quasar duty cycle ($\sim10^6$ years) (Lansbury et al., 2019, Glikman et al., 2012).

6. Dust and Gas Geometry: Host-scale (ISM) vs. Nuclear Obscuration

Combined SED, IFU, and X-ray analyses indicate that the obscuring medium in HRQs is predominantly located on galactic scales, outside the narrow-line region, rather than in the classical AGN torus or broad-line region (Temple et al., 2019, Banerji et al., 2016, LaMassa et al., 2016). Dust-to-gas ratios in HRQs are intermediate between local ISM and AGN values, and NuSTAR hard X-ray spectroscopy reveals patchy or clumpy geometry, with global columns $N_{\rm H, S}\gg N_{\rm H,Z}$ (LaMassa et al., 2016, Lansbury et al., 2019). Partial covering effects (e.g., in damped Lyman-alpha (DLA) systems) can even turn luminous quasars optically invisible, biasing the demographics of HRQs and cosmic metal budgets (Krogager et al., 2015).

A fraction of HRQs exhibit anomalously steep UV-only reddening curves, best explained by silicate dust population truncated at small grain sizes (e.g., IRAS 14026+4341; $dn(a)/da\propto a^{-1.4}$ to $a_{max}=70$ nm), likely a result of AGN feedback or in-situ dust formation (Jiang et al., 2013).

7. Implications for Quasar/Host Co-Evolution and Cosmic AGN Census

HRQs occupy a critical, obscured stage in the life cycle of the most massive black holes and their host galaxies. Their high Eddington ratios, frequent major mergers, and evidence for simultaneous rapid star formation and SMBH accretion render them pivotal for understanding feedback-regulated black hole growth, the transition from dust-enshrouded starbursts to UV-luminous quasars, and determining the completeness of AGN census at peak cosmic accretion epochs (Glikman et al., 2012, Banerji et al., 2016, Banerji et al., 2015).

The evolutionary sequence: major merger $\rightarrow$ starburst/ULIRG $\rightarrow$ HRQ (blow-out/feedback phase) $\rightarrow$ unobscured blue quasar, is robustly supported by SEDs, host properties, and outflow diagnostics. The HRQ phase is brief compared to the unobscured quasar lifetime ($\sim15$–$20\%$), but is disproportionately represented among the highest luminosity, fastest-growing QSOs, showing a flatter luminosity function and a duty cycle of $\sim10^6$ years per object (Banerji et al., 2015, Glikman et al., 2012, Glikman et al., 2022).

The prevalence of both unobscured host star formation (detected in the rest-UV) and extremely high obscured SFRs (seen in FIR/sub-mm) implies that HRQs must be included to reconcile AGN-driven feedback and total star-formation histories in galaxy evolution models. Their multi-phase feedback links small-scale nuclear phenomena to kpc-scale ISM evolution and provides stringent constraints for hydrodynamical simulations of SMBH/galaxy co-growth (Wethers et al., 2018, Alaghband-Zadeh et al., 2016, Banerji et al., 2018).

---

Survey	HRQ selection cuts	Typical $E(B-V)$ range	Duty cycle / Fraction	Reference
FIRST–2MASS	$J-K_s\geq1.7$; $R-K_s\geq4.0$	$0.1–1.5$	$15–20\%$ (to $K<15$)	(Glikman et al., 2012)
UKIDSS/VHS/WISE	$(J-K)_{\rm Vega}>2.5$	$2–6$ mag ($A_V$)	$>$40% at high $L_{\rm bol}$	(Banerji et al., 2015, Glikman et al., 2022)
WISE–2MASS	$W$ color, $E(B-V)>0.25$	$0.25–1.5$	$>40\%$ @ high $L_{\rm bol}$	(Glikman et al., 2022)
ALMA/SCUBA-2	$K<18.4$, $(J-K)_{\rm Vega}$	$A_V\sim2–6$	Obscured SFR $\sim600$–$4500\,M_\odot\,yr^{-1}$	(Wethers et al., 2020)

Widespread multi-wavelength coverage and high-resolution spectroscopy have now positioned HRQs as the canonical transitional AGN population, required for any complete theory of SMBH and host galaxy co-evolution at cosmic noon and beyond.