The Disk Plus (Failed) Wind System of 3C 47: A Story of Accretion Disks and Binary Black Holes

TL;DR

The paper investigates how accretion-disk emission shapes broad-line profiles in Population B AGN, using a disk-plus-failed-wind model to reconcile low- and high-ionization lines, with 3C 47 as a case study. It analyzes a large, low-redshift AGN sample, showing that redward asymmetries in Hβ scale with black hole mass in a way consistent with gravitational redshift for $M_{\mathrm{BH}} \gtrsim 5\times 10^{8}\,M_{\odot}$, and that radio-loud and radio-quiet Pop B objects converge when controlling $M_{\mathrm{BH}}$ and $L/L_{\mathrm{Edd}}$. The work highlights double-peaked emitters as rare disk-dominated systems that may be truncated by a secondary black hole, and demonstrates that a symmetric, virialized disk plus a wind-like component can account for the CIV 1549 and other UV lines. Altogether, the study proposes a unified BLR framework in which inner disk emission dominates the low-ionization lines while a failed wind shapes high-ionization features, with significant implications for SMBH binary demographics and quasar evolution in mature, starved systems.

Abstract

[Abridged] Optically thick, geometrically thin accretion disks around supermassive black holes are thought to contribute to broad-line emission in type-1 active galactic nuclei (AGN). However, observed emission line profiles most often deviate from those expected from a rotating disk. This report examines the role of accretion disks in broad-line emission of Population B AGN characterized by relatively low accretion rates in which broad lines show large redward asymmetry both in H$β$ and Mg II$λ$ 2800. An unbiased comparison matching black hole mass and Eddington ratio suggests that the most powerful radio-loud quasars show the highest red-ward asymmetries in H$β$. These shifts can be accounted for by gravitational and transverse redshift effects, especially for black hole masses larger than $\approx$10$^{8.7}$ M$_\odot$. The analysis of the extremely jetted quasar 3C 47 adds another piece to the puzzle: not only are the low ionization profiles of 3C 47 well-described by a relativistic Keplerian accretion disk model, with line emission in the range 100 - 1,000 gravitational radii, but also the high-ionization line profiles can be understood as a combination of disk plus a failed wind contribution that is in turn hiding the disk emission. Constraints on radio properties and line profile variability suggest that 3C 47 might involve the presence of a second black hole with secondary-to-primary mass ratio $\sim$ 0.5. We conjecture that the double peakers - type-1 AGN with Balmer line profiles consistent with accretion disk emission - might have their emission truncated by the sweeping effect of a second black hole. In non-starving systems, the disk signal is plausibly masked by additional line emission, rendering the disk contribution harder to detect.

Figures (7)

• Figure 1: The quasar main sequence optical plane defined by the parameters $R_\mathrm{{Fe}{\textsc{ii}}}$ and FWHM H$\beta$, with a subdivision into Population A and B, extreme Population B, and spectral types. The numbers below each spectral type identification are the prevalence of AGN in each spectral type, followed by the prevalence of radio-intermediate and RL AGN in each spectral bin. The shading is proportional to the surface density of AGN in the sample of zamfiretal10, and has been expanded at low density values to better represent of sparsely occupied spectral bins. The location of 3C 47 in the main sequence (discussed in Section \ref{['3c47']}) is at an extreme of extreme Population B (the area shaded rose encompassing the B1${^{++}}$ and B1$^{+++}$ spectral types).
• Figure 2: Left panel: sketch of a typical Balmer line profile (H$\beta$) of a Population B AGN with the centroids at different fractional intensity marked, with an emphasis on the $c(\frac{1}{4})$ analyzed in this work. The thin red line traces the putative "very broad component" that might ultimately be traced back to accretion disk emission, and is represented by a red-shifted Gaussian of FWHM $\sim 10,000$ km s$^{-1}$. Right: Distribution of H$\beta$ centroid $c(\frac{1}{4})$ for Population B RQ (blue) and RL (red), for the sample of zamfiretal10. The shaded area is the uncertainty range at a $\pm 2 \sigma$ confidence level.
• Figure 3: Behavior of centroid at $\frac{1}{4}$ peak line intensity of H$\beta$ as a function of black hole mass, for RQ (top) and RL (bottom) AGN of the terefemengistueetal23 sample. Data points are median centroid values in ranges of 0.5 dex; error bars are sample semi-interquartile ranges.
• Figure 4: Sketch illustrating the differences expected for gravitational redshift (left) and redshift due to gas with an infalling component of its velocity where the approaching side of the inflow is hidden/obscured (right). The photon gravitational redshift from an annulus at a given distance from the central black hole is independent of the velocity fields. In contrast, in the case of infall, the blue and red colors of the arrows represent the Doppler shifts (blue-shift and red-shift) caused by the radial component of the gas velocity field. The gray cloud indicates that the approaching side of the gas yielding a blueshift must be obscured to produce a net redshift in the emission line profile.
• Figure 5: Interpretation of the C iv$\lambda$1549 line blend. The thick black line is the accretion disk profile as derived from the H$\beta$ fit, scaled to the radial velocities of C iv$\lambda$1549 and He ii$\lambda$1640. The blue thick line represents the failed wind emission, in the form of a strong, symmetric line component plus a faint wing whose shape is highly uncertain. Orange line represent semi-broad components of He ii$\lambda$1640, O iii]$\lambda 1663$, Niv]$\lambda 1486$, Alii]$\lambda 1670$. Ordinate is specific flux in arbitrary units. The bump on the blue side of C iv$\lambda$1549 at $\approx -12,000$ km s$^{-1}$ is likely associated with the outflow, as the intensity of Niv]$\lambda 1486$ is expected to be just a few hundreds the one of C iv$\lambda$1549 from photoionization simulations. The narrow component of the C iv$\lambda$1549 profile is suppressed by narrow absorption lines clustered close to the rest-frame of the quasar. Adapted from Fig. 8 in terefemengistueetal24.