The X-ray properties of the most luminous quasars with strong emission-line outflows

TL;DR

The paper investigates whether the presence of strong UV-emission-line winds in the most luminous quasars is tied to intrinsic X-ray weakness or absorption by analyzing a UV-selected sample of 10 high-luminosity, radio-quiet quasars with blueshifted C\ IV lines using Chandra data. The X-ray spectra generally show steep continua with $\Gamma\gtrsim1.7$, and eight of ten objects are consistent with the standard $L_{\rm X}$–$L_{\rm UV}$ relation, while two are X-ray weak (one potentially obscured), indicating no strong, universal X-ray deficiency associated with winds. A tentative 2$\sigma$ indication suggests a link between extreme wind velocities ($>3000$ km s$^{-1}$) and suppressed X-ray flux, but the small sample size prevents a definitive claim. The results underscore the complexity of the wind–X-ray relationship in the most luminous quasars and call for larger, simultaneous UV–X-ray campaigns to disentangle intrinsic properties from variability and orientation effects. Key methodological points include uniform Chandra/X-ray analysis, careful UV spectral modelling of C\ IV with MFICA reconstructions to confirm winds, and CXO/Archival cross-checks with a control-like supplementary sample.

Abstract

Strong outflows from active galactic nuclei are frequently observed in objects with lower coronal X-ray luminosity. This intrinsic X-ray weakness is considered a requirement for the formation of radiatively driven winds. To obtain an unbiased view on the connection between X-ray emission and the presence of powerful winds in the most luminous quasar phase, we present an X-ray analysis of a sample of extremely luminous, radio-quiet quasars with signatures of strong outflows in their rest-frame ultraviolet (UV) emission spectra. We study the $Chandra$ X-ray spectral properties of 10 objects, selected from the Sloan Digital Sky Survey Data Release 16 quasar catalogue based on their UV luminosities and ${\rm C}_{\rm IV}$ emission line blueshifts, comparing them to typical optically blue quasars. Our analysis reveals that seven out of 10 quasars in our sample have photon indices $Γ>1.7$. Only two out of 10 objects exhibiting outflows with velocities exceeding 1400 km/s are X-ray 'weak', consistent with the fraction of X-ray 'weak' objects generally observed in quasar populations. Notably, one of the objects identified as X-ray 'weak' is likely an intrinsically X-ray 'normal' quasar that is heavily obscured. We observe a tentative indication at a $\sim$2$σ$ confidence level that the correlation between the excessively low X-ray flux level and the presence of ${\rm C}_{\rm IV}$ emission-line outflows might emerge at wind velocities greater than 3000 km/s. Our study provides additional evidence that the relationship between X-ray emission and the presence of winds is intricate. Our findings emphasise the need for X-ray observations of a larger sample of UV-selected quasars with confirmed strong emission-line outflows to unravel the nuanced interplay between winds and X-ray emission.

Figures (10)

• Figure 1: Shift of centroid of C iv$\lambda$1549 Å line in units of km/s, as function of logarithmic luminosity $\nu L_\nu$ at 2500 Å, available in 2022ApJS..263...42W. The sample consists of all the quasars in the SDSS DR16 sample with redshift in the $z=1.8$ -- $2.2$ interval, colour-coded from red to black by the density of the underlying points. Empty cyan circles indicate 45 objects pre-selected for the one-by-one modelling of the C iv line with multiple components. Filled semi-transparent blue circles show a subset of 10 objects with the confirmed strongly blueshifted component, the main dataset analysed in this paper. Violet circles show four supplementary archival sources.
• Figure 2: Example of analysis of C iv spectral region of SDSS spectra. The fitted lines are reported as labels. The components employed in the fit are colour-coded, as shown in the legend. The reported C iv outflow velocity is measured as the shift of the outflow component with respect to the BLR component. The dashed vertical lines indicate the expected position for the fitted lines according to the redshift reported in the catalogue. The shaded light grey regions are telluric bands, narrow absorption lines, or bad pixels and are therefore excluded from the fit. See Appendix \ref{['optspectrall']} for the analysis of other sources.
• Figure 3: C iv emission space with hexagons colour-coded by median He ii EW for 66810 quasars in SDSS DR16 sample with redshift in $z=1.8$ -- $2.2$ interval, having reliable C iv and He ii measurements from MFICA reconstructions. Only hexagons with five or more quasars are plotted. Circles colour-coded by corresponding He ii EW show the main dataset of 10 objects and four supplementary archival sources.
• Figure 4: Example of analysis of Chandra X-ray spectra. To enhance visual clarity, the spectrum in the plot is binned to ensure at least 1.5 counts per energy channel. Black crosses represent the observational data with corresponding errors, red line represents the best-fit model. See Appendix \ref{['Xspectrall']} for the analysis of other sources.
• Figure 5: Rest-frame monochromatic luminosities $L_{\mathrm{X}}$ against $L_{\mathrm{UV}}$ for 10 quasars in our main dataset and four additional quasars from Chandra and XMM-Newton archives, colour-coded by C iv outflow velocity in units of km/s, measured in our modelling of C iv line as the shift of the outflow component with respect to the BLR component. Colour-coded and empty black squares show modelling of J111800.50+195853.4 with a free and fixed photon index, respectively. Downward arrows indicate non-detection, and the corresponding squares represent upper limits. Light red dots represent the sample of about 2000 quasars from 2020AA...642A.150L, with the relative regression line in red, for which the slope $\gamma$, with its error, and the dispersion $\delta$ are specified. The dashed and dotted lines trace the $1\sigma$ and $3\sigma$ dispersion, respectively.