No Arabic abstract
Weak emission line quasars are a rare and puzzling group of objects. In this paper we present one more object of this class found in the Sloan Digital Sky Survey (SDSS). The quasar SDSS J094533.99+100950.1, lying at z = 1.66, has practically no C IV emission line, a red continuum very similar to the second steepest of the quasar composite spectra of Richards et al., is not strongly affected by absorption and the Mg II line, although relatively weak, is strong enough to measure the black hole mass. The Eddington ratio in this object is about 0.45, and the line properties are not consistent with the trends expected at high accretion rates. We propose that the most probable explanation of the line properties in this object, and perhaps in all weak emission line quasars, is that the quasar activity has just started. A disk wind is freshly launched so the low ionization lines which form close to the disk surface are already observed but the wind has not yet reached the regions where high ionization lines or narrow line components are formed. The relatively high occurrence of such a phenomenon may additionally indicate that the quasar active phase consists of several sub-phases, each starting with a fresh build-up of the Broad Line Region.
We demonstrate a new technique for determining the physical conditions of the broad line emitting gas in quasars, using near-infrared hydrogen emission lines. Unlike higher ionisation species, hydrogen is an efficient line emitter for a very wide range of photoionisation conditions, and the observed line ratios depend strongly on the density and photoionisation state of the gas present. A locally optimally emitting cloud model of the broad emission line region was compared to measured emission lines of four nearby ($zapprox0.2$) quasars that have optical and NIR spectra of sufficient signal-to-noise to measure their Paschen lines. The model provides a good fit to three of the objects, and a fair fit to the fourth object, a ULIRG. We find that low incident ionising fluxes ($phih<10^{18}$cmsqs), and high gas densities ($ h>10^{12}$cmcu) are required to reproduce the observed hydrogen emission line ratios. This analysis demonstrates that the use of composite spectra in photoionisation modelling is inappropriate; models must be fitted to the individual spectra of quasars.
We identify a sample of 74 high-redshift quasars (z>3) with weak emission lines from the Fifth Data Release of the Sloan Digital Sky Survey and present infrared, optical, and radio observations of a subsample of four objects at z>4. These weak emission-line quasars (WLQs) constitute a prominent tail of the Lya+NV equivalent width distribution, and we compare them to quasars with more typical emission-line properties and to low-redshift active galactic nuclei with weak/absent emission lines, namely BL Lac objects. We find that WLQs exhibit hot (T~1000 K) thermal dust emission and have rest-frame 0.1-5 micron spectral energy distributions that are quite similar to those of normal quasars. The variability, polarization, and radio properties of WLQs are also different from those of BL Lacs, making continuum boosting by a relativistic jet an unlikely physical interpretation. The most probable scenario for WLQs involves broad-line region properties that are physically distinct from those of normal quasars.
We present the discovery of PSO J083.8371+11.8482, a weak emission line quasar with extreme star formation rate at $z=6.3401$. This quasar was selected from Pan-STARRS1, UHS, and unWISE photometric data. Gemini/GNIRS spectroscopy follow-up indicates a MgII-based black hole mass of $M_mathrm{BH}=left(2.0^{+0.7}_{-0.4}right)times10^9~M_odot$ and an Eddington ratio of $L_mathrm{bol}/L_mathrm{Edd}=0.5^{+0.1}_{-0.2}$, in line with actively accreting supermassive black hole (SMBH) at $zgtrsim6$. HST imaging sets strong constraint on lens-boosting, showing no relevant effect on the apparent emission. The quasar is also observed as a pure point-source with no additional emission component. The broad line region (BLR) emission is intrinsically weak and not likely caused by an intervening absorber. We found rest-frame equivalent widths of EW(Ly$alpha$+NV) $=5.7pm0.7$ Angstrom, EW(CIV) $leq5.8$ Angstrom (3-sigma upper limit), and EW(MgII) $=8.7pm0.7$ Angstrom. A small proximity zone size ($R_mathrm{p}=1.2pm0.4$ pMpc) indicates a lifetime of only $t_mathrm{Q}=10^{3.4pm0.7}$ years from the last quasar phase ignition. ALMA shows extended [CII] emission with a mild velocity gradient. The inferred far-infrared luminosity ($L_mathrm{FIR}=(1.2pm0.1)times10^{13},L_odot$) is one of the highest among all known quasar hosts at $zgtrsim6$. Dust and [CII] emissions put a constraint on the star formation rate of SFR $=900-4900~M_odot,mathrm{yr^{-1}}$, similar to that of hyper-luminous infrared galaxy. Considering the observed quasar lifetime and BLR formation timescale, the weak-line profile in the quasar spectrum is most likely caused by a BLR which is not yet fully formed rather than continuum boosting by gravitational lensing or a soft continuum due to super-Eddington accretion.
We investigate the relationship between the rest-frame equivalent width (EW) of the C IV lambda1549 broad-emission line, monochromatic luminosity at rest-frame 5100 A, and the Hbeta-based Eddington ratio in a sample of 99 ordinary quasars across the widest possible ranges of redshift (0 < z < 3.5) and bolometric luminosity (10^{44} <~ L <~ 10^{48} erg s^{-1}). We find that EW(C IV) is primarily anti-correlated with the Eddington ratio, a relation we refer to as a modified Baldwin effect (MBE), an extension of the result previously obtained for quasars at z < 0.5. Based on the MBE, weak emission line quasars (WLQs), typically showing EW(C IV) <~ 10 A, are expected to have extremely high Eddington ratios. By selecting all WLQs with archival Hbeta and C IV spectroscopic data, nine sources in total, we find that their Hbeta-based Eddington ratios are typical of ordinary quasars with similar redshifts and luminosities. Four of these WLQs can be accommodated by the MBE, but the other five deviate significantly from this relation, at the >~3 sigma level, by exhibiting C IV lines much weaker than predicted from their Hbeta-based Eddington ratios. Assuming the supermassive black-hole masses in all quasars can be determined reliably using the single-epoch Hbeta-method, our results indicate that EW(C IV) cannot depend solely on the Eddington ratio. We briefly discuss a strategy for further investigation into the roles that basic physical properties play in controlling the relative strengths of broad-emission lines in quasars.
We present Chandra observations of nine high-redshift quasars (z=2.7-5.9) discovered by the Sloan Digital Sky Survey with weak or undetectable high-ionization emission lines in their UV spectra (WLQs). Adding archival X-ray observations of six additional sources of this class has enabled us to place the strongest constraints yet on the X-ray properties of this remarkable class of AGNs. Although our data cannot rule out the possibility that the emission lines are overwhelmed by a relativistically boosted continuum, as manifested by BL Lac objects, we find that WLQs are considerably weaker in the X-ray and radio bands than the majority of BL Lacs found at much lower redshifts. If WLQs are high-redshift BL Lacs, then it is difficult to explain the lack of a large parent population of X-ray and radio bright weak-lined sources at high redshift. We also consider the possibility that WLQs are quasars with extreme properties, and in particular that the emission lines are suppressed by high accretion rates. Using joint spectral fitting of the X-ray spectra of 11 WLQs we find that the mean photon index in the hard X-ray band is consistent with those observed in typical radio-quiet AGNs with no hint of an unusually steep hard-X-ray spectrum. This result poses a challenge to the hypothesis that WLQs have extremely high accretion rates, and we discuss additional observations required to test this idea.