SDSS J013127.34−032100.1: A NEWLY DISCOVERED RADIO-LOUD QUASAR AT z = 5.18 WITH EXTREMELY HIGH LUMINOSITY

Identifying a sample of high-redshift, radio-loud quasars provides important clues about the early universe and potentially can probe the formation of massive galaxies. Jiang et al. (2007a) found that the radio-loud fraction of quasars is a strong function of both redshift and optical luminosity, such that the radio-loud fraction decreases rapidly with increasing redshift and decreasing luminosity. Although the expected number of radio-loud sources with redshifts smaller than 3 is fairly consistent with the optically selected quasars from the Sloan Digital Sky Survey (SDSS), when it comes to high-redshift cases, there is an apparent deficit of SDSS radio-loud quasars (Volonteri et al. 2011). It still remains to be seen whether the small number of high-redshift radio-loud quasars discovered to date is due to selection effects owing to current limitations of the high redshift identification techniques or whether these quasars are intrinsically rare. Radio-emitting, high-redshift objects may be particularly valuable in addressing questions with respect to early cosmological evolution. These rare, distant, and powerful objects provide important insights into the activity of supermassive black holes (SMBHs) at early cosmological epochs and into the physical properties of their environments. Rapid galaxy formation and starburst activity may go hand-in-hand with the radio-active phase of accreting SMBHs (Frey et al. 2010).

Recent studies (e.g., Barth et al. 2003; Jiang et al. 2007b; Kurk et al. 2009; Willott et al. 2010) show that high-redshift quasars with M_BH > 10⁹ M_☉ are accreting close to the Eddington limit (L_bol/L_Edd ≈ 1), while at lower redshifts (2 < z < 3) the most luminous quasars (L_bol > 10⁴⁷ erg s⁻¹) are characterized by L_bol/L_Edd ≈ 0.25 with a dispersion of 0.23 dex (e.g., Shen et al. 2008). Furthermore, De Rosa et al. (2011) found that the distribution of observed Eddington ratios for high-redshift quasars is significantly different than that of a luminosity-matched comparison sample of SDSS quasars at lower redshifts (0.35 < z < 2.25): the z > 4 sources are accreting significantly faster than the lower redshift ones, which implies that the masses of these BHs at high redshifts are typically lower than those of the lower redshift quasars. High-redshift quasars, especially highly radio-loud ones, may have intrinsically different properties compared to lower redshift ones.

We report the discovery and first observational results of an extremely luminous and radio-loud quasar, SDSS J0131−0321, at z = 5.18. It appears similar to another radio-loud source, SDSS J0741+2520 (McGreer et al. 2009), at nearly the same redshift, but J0131−0321 is more luminous at optical and radio wavelengths. The high luminosity of J0131−0321 may be due to super-Eddington accretion. Throughout this Letter, luminosity distances are calculated using a Λ-dominated flat cosmology with H₀ = 70 km s⁻¹Mpc⁻¹, Ω_m = 0.3 and Ω_Λ = 0.7.

J0131−0321 was first selected as a candidate high-redshift quasar according to optical-IR selection criteria based on SDSS and Wide-Field Infrared Survey Explorer (WISE) photometric data (Wu et al. 2012). The first low-resolution optical spectrum of this source was obtained with the Lijiang 2.4 m telescope (LJT; see Fan et al. 2014) using the Yunnan Fainter Object Spectrograph and Camera (YFOSC; Zhang et al. 2014) on 2013 November 25 (UT). With a 25 minute exposure, this spectrum, taken with the YFOSCG3 grism (λ/Δλ ∼ 680 at 7000 Å with a 10 slit) in long-slit mode, yields, through the identification of Lyα emission, a redshift close to 5.2. It has a signal-to-noise ratio (S/N) of 8.3 pixel⁻¹ in the continuum. Due to the relatively low S/N and a cut-off at 9000 Å in the first spectrum, another spectrum was obtained with a more red-sensitive grism (YFOSCG5, λ/Δλ ≈ 550 at 7000 Å with a 10 slit) by the LJT during a photometric night on 2014 February 20 (UT). The average seeing (12) during the second observation was better than the first.

In order to obtain more reliable estimates of the continuum luminosity and SMBH mass of J0131−0321, we obtained subsequent observations in the optical and near-IR bands with higher resolution spectroscopy using the Magellan Echelette (MagE) and FIRE spectrographs on 2014 January 3, 16, and 18, (UT) with 90 and 50 minutes of integration time, respectively. The MagE optical spectrum provides a resolution of λ/Δλ = 4100 with a 1'' slit, and the FIRE spectrum covers 8200 Å to 25000 Å at a spectral resolution of λ/Δλ = 6000. These higher resolution spectra lead to a more accurate redshift of z = 5.18 ± 0.01 based on the Lyα and Mg ii lines. The raw data are processed in a standard way, including bias, flat correction, sky subtraction, flux calibration, and telluric correction. The region of overlap with the optical spectrum is used to calibrate and normalize the near-IR flux more accurately. All the optical spectra plotted in Figure 1 are scaled to the SDSS photometric magnitude, and the spectrum taken with YFOSCG5 on a photometric night is considered to be the standard for recalibrating the other two spectra. For the purposes of display and comparison with the YFOSC spectra, the MagE spectrum has been smoothed with a 5 pixel boxcar filter.

Zoom In Zoom Out Reset image size

Photometric points from the radio to the UV/optical bands were collected from multiple surveys, including Faint Images of the Radio Images of the Sky at Twenty-Centimeters (FIRST), Two Micron All Sky Survey (2MASS), WISE, and SDSS (Becker et al. 1995; Skrutskie et al. 2006; Wright et al. 2010; Adelman et al. 2000). The radio flux density of J0131−0321 is 33 mJy at 1.4 GHz (observed-frame). The source appears core-dominated, as no extended structure is visible on the 5'' resolution scale of FIRST.

For simplicity, we use the UV/optical continuum luminosity as a measure of the energy output of the quasar. We corrected the photometric magnitude for Galactic reddening using the E(B − V) values from Schlafly & Finkbeiner (2011). In Figure 2, the blue solid line is the mean spectral energy distribution (SED) for local radio-loud quasars (Elvis et al. 1994). The photometric data points of J0131−0321 are roughly consistent with the upward-scaled template for local radio-loud quasars (the top dashed line), except for the stronger Lyα absorption seen in the bluest bands of J0131−0321. Six additional high-redshift (z ≈ 5; see Table 1) radio-loud quasars are shown for comparison. It is apparent that J0131−0321 is the most luminous among them. Compared with other less luminous quasars at z ≈ 5 tabulated in Table 1, one may easily postulate that these highly luminous quasars are little affected by the UV/optical obscuration along the line of sight or significant jet contributions or lensing effects may exist. In addition, it is noteworthy that J0011+1446 (Schneider et al. 2005) and J0741+2520 (McGreer et al. 2009) also have high UV/optical luminosities comparable to J0131−0321 except with lower radio fluxes. These extremely luminous radio-loud quasars at z > 5 will be important for exploring the powering mechanisms behind high-redshift radio-loud quasars. The origin of the high UV/optical luminosities of these quasars may be due to nonthermal emission boosted by the jet or perhaps gravitational lensing. Since the best angular resolution for all the photometric data available is no better than 1'', in order to investigate the origin of the extremely high luminosity, additional high-resolution images are needed.

Zoom In Zoom Out Reset image size

The radio-loudness parameter R, defined as the ratio of the rest-frame flux densities in the radio (5 GHz) to optical (4400 Å) bands (Kellermann et al. 1989), strongly correlates with the mass-accretion rate (Ho 2002). Supposing that the power-law continuum of J0131-0321 between the optical and radio bands follows f(ν)∝ν^−0.5, the radio flux density at rest-frame 5 GHz is 6.9×10⁻²⁶ erg s⁻¹cm⁻² Hz⁻¹ and the flux density at 4400 Å is 6.6 × 10⁻²⁸ erg s⁻¹cm⁻² Hz⁻¹. Thus, the radio-loudness of J0131−0321 can be derived to be R ≈ 100.

Recent investigations (e.g., Momjian et al. 2004; Frey et al. 2010, 2011; Cao et al. 2014) based on very long baseline interferometry (VLBI) of radio-loud quasars at high redshifts (z > 4) have revealed a population of steep-spectrum, compact sources plausibly related to very young radio quasars associated with early galaxy formation activity. It is unclear whether J0131−0321 is related to this class of objects. Additional follow-up VLBI radio observations would be helpful as well as X-ray data that can be combined with radio and optical diagnostics (see Wu et al. 2013).

Despite the fact that a time lag for the Mg ii line has been measured only in a very few cases of reverberation mapping of active galactic nuclei (AGNs; Reichert et al. 1994; Metzroth et al. 2006), the line width of Mg ii is known to correlate well with that of Hβ in single-epoch spectra (Shen et al. 2008; McGill et al. 2008; Wang et al. 2009), suggesting that using the Mg ii line can yield virial mass estimates consistent with those based on the Balmer lines. It has been further confirmed that the Mg ii and Balmer lines share similar kinematics because they deliver mutually consistent SMBH mass estimates with minimal internal scatter (⩽0.1 dex) using the latest virial mass estimators (Ho et al. 2012). The empirical scaling relation usually adopted to estimate SMBH masses from the Mg ii line and its associated continuum at 3000 Å is as follows:

$$\begin{eqnarray} \log \left(\frac{M_{\rm BH,vir}}{M_\odot }\right)&=&a+b\cdot \log \left(\frac{\lambda L_{\lambda }}{10^{44}\, {\rm erg\, s^{-1}}}\right)\nonumber \\ &&+2\cdot \log \left(\frac{\rm FWHM}{\rm km\, s^{-1}}\right). \end{eqnarray} \tag{ 1 }$$

For the Mg ii line, the typical values of a and b are 0.50 and 0.62 (McLure & Dunlop 2004), respectively. This equation is calibrated from local AGNs based on broad emission-line widths and continuum luminosities. Figure 3 presents the Magellan/FIRE near-IR spectrum of J0131−0321 and corresponding fitted profiles of the Mg ii line, pseudo-continuum, and Fe ii template. Considering the rest-frame 3000 Å continuum is just red-shifted to the position of an absorption window, both global fitting and emission-line fitting are used to obtain the continuum slope and FWHM of the Mg ii line. As a result, the 3000 Å luminosity for J0131−0321 is estimated to be 2.1$^{+0.4}_{-0.3}\times 10^{47}$ erg s⁻¹. Assuming an empirical conversion factor between a 3000 Å continuum luminosity and a bolometric luminosity of L_bol = 5.18 × L₃₀₀₀ (e.g., Jin et al. 2012; Richards et al. 2006), it is easy to get the approximate bolometric luminosity of J0131−0321 ($L_{\rm bol} = 1.1^{+0.2}_{-0.2} \times 10^{48}$ erg s⁻¹), which is nearly 4.5 times greater than in the most distant quasar ULASJ1120+0641 at z = 7.085 (Mortlock et al. 2011). The BH mass is estimated to be 2.7$^{+0.5}_{-0.4} \times 10^9 \, M_\odot$ according to Equation (1). Assuming isotropic emission, the Eddington luminosity L_Edd = 1.3 × 10³⁸M_BH/M_☉ erg s⁻¹ = 3.5 × 10⁴⁷ erg s⁻¹, and the Eddington ratio is L_bol/L_Edd = 3.14. While in the recent research of Trakhtenbrot & Netzer (2012), they find that the Mg ii estimator may cause an underestimation of the BH mass by ∼0.25 dex. Therefore, when using the latest relation, the BH mass of J0131−0321 is ∼4.0×10⁹ M_☉, and the corresponding L_Edd is ∼5.2 × 10⁴⁷ erg s⁻¹, which means the bolometric luminosity is still larger than the Eddington luminosity. Measurement errors are small, only ∼0.1 dex. The main uncertainties are dominated by systematic errors, which amount to at least 0.4 dex, and can be summarized as follows.

• 1.  
  The scatter in the virial estimate of BH mass for the SDSS sub-sample is already ∼0.34 dex (Trakhtenbrot & Netzer 2012). We must further assume that these mass estimators can be applied to quasars at high redshifts(z > 5) and high luminosity (L_bol > 10⁴⁸ erg s⁻¹) cases where they are completely untested.
• 2.  
  There is evidence indicating that super-Eddington accreting BHs may exhibit a somewhat different R_BLR − L relation than sub-Eddington sources (Du et al. 2014).
• 3.  
  Due to the limitations of the currently available data, the flux measurements are not very accurate and the assumed bolometric correction may not be suitable for this quasar.
• 4.  
  We assume isotropic radiation from this quasar, while for radio-loud quasars, the measured bolometric luminosity may be biased somewhat due to nonthermal contributions and selection effects.

Zoom In Zoom Out Reset image size

All the considerations mentioned above should be taken into account when calculating the BH mass and Eddington ratio. However, taken at face value, J0131−0321 appears to be a candidate for one of the few super-Eddington quasars known at z > 5.

In the subsample of radio-loud quasars at z ≳ 4, there are six objects with BH mass and Eddington ratio estimates (see Table 2) in the catalog of Shen & Liu (2011). In this table, all of the objects have relatively high Eddington ratios and J0131−0321 is the only one with an Eddington ratio larger than 2 at z > 5. Except for GB 1508+5714, these objects have M_BH > 10⁹ M_☉. When considering spinning BHs, which have a high radiative efficiency, it is very difficult for such large BH masses to form even from seeds with masses ∼10³–10⁵ M_☉ (e.g., Ghisellini et al. 2014). At the same time, the relative dominance between the radio and optical bands seems to evolve strongly with redshift (e.g., Singal et al. 2013), which results in significant changes in the observed frequency of radio-loud quasars with redshift and UV luminosity (e.g., Jiang et al. 2007a). Furthermore, as the degree of radio-loudness increases with increasing radio luminosity, it would be possible to explore the connections between the efficiency of the formation of relativistic jets and accretion power, which may, in turn, depend on the combination of the evolving accretion rate and BH spin (see Tchekhovskoy et al. 2010). It is worth noting that, apart from J0131−0321, the BH masses of the other six objects listed in Table 2 are all based on the C iv emission line, which could be highly uncertain due to nonvirial kinematics in the high-ionization C iv line region (Shen & Liu 2012). The BH mass of J0131−0321 is based on Mg ii, and thus should be more reliable. For radio-loud sources in general, if we cannot eliminate jet contamination, use of the size–continuum luminosity relation might overestimate the size of the broad-line region (BLR) and hence the BH mass (Wu et al. 2004). Broad emission lines, however, can be seen clearly in all spectra, which suggests that the beamed relativistic jets may have a small contribution to the UV/optical continua in these sources.

The observational data for J0131−0321 are currently very limited. It would be particularly valuable to do multi-wavelength follow-up observations with more dedicated and advanced facilities, such as VLBI for the radio regime and Chandra or XMM-Newton for X-ray energies. These future observations will enable us to investigate further the evolution of radio-loud quasars and to explore their powering mechanisms, which are still unclear.

We acknowledge the support of the staff of the LJT. Funding for the LJT has been provided by CAS and the People's Government of Yunnan Province. X.B.W. is grateful for the support from the NSFC grants 11033001 and 11373008. J.M.B. is supported by the NSFC grants 11133006 and 11361140347. X.B.W., J.M.B., and L.C.H. are all supported by the Strategic Priority Research Program "The Emergence of Cosmological Structures" of the Chinese Academy of Sciences (grant No. XDB09000000). W.N.B. is grateful for support from NASA ADP grant NAG5-13035. This Letter includes data gathered with the 6.5 meter Magellan Telescopes located at Las Campanas Observatory, Chile. The MagE spectral reduction was kindly provided by George Becker. We also acknowledge the use of the quasar spectral fitting code kindly provided by Yue Shen.