The X-ray$-$UV Luminosity Relation of eROSITA Quasars

TL;DR

This study analyzes the X-ray–UV luminosity relation in unobscured quasars detected by eROSITA using two samples (eFEDS and eRASS1) to test redshift invariance and cosmological utility. The relation is modeled as $L_X ∝ L_{UV}^{γ}$, with measured slopes around $γ≈0.62–0.63$ and intrinsic dispersion $δ≈0.20–0.22$ dex, showing no significant evolution up to $z≲3$ (eFEDS) or $z≲1.5$ (eRASS1). Careful sample cleaning (dust/host contamination, BAL/RL removal, absorption handling, and Eddington-bias correction) yields robust, instrument-independent results, reinforcing the potential of quasars as standardizable candles. Future eROSITA releases are expected to deliver ~78,000 QSOs, enabling a competitive QSO Hubble diagram and powerful cosmological tests beyond the current sample size.

Abstract

The non-linear relation between the UV and X-ray luminosity in quasars has been studied for decades. However, as we lack a comprehensive model able to explain it, its investigation still relies on observational efforts. This work focuses on optically selected quasars detected by eROSITA. We present the properties of the sources collected in the eROSITA early data release (eFEDS) and those resulting from the first six months of eROSITA all-sky survey (eRASS1). We focus on the subset of quasars bright enough in the optical/UV band to avoid an ''Eddington bias'' towards X-ray brighter-than-average spectral energy distributions. The final samples include 1,248 and 519 sources for eFEDS and eRASS1, up to redshift $z\approx3$ and $z\approx1.5$, respectively. We found that the X-ray$-$UV luminosity relation shows no significant evolution with redshift, and its slope is in perfect agreement with previous compilations of quasar samples. The intrinsic dispersion of the relation is about 0.2 dex, which is small enough for possible cosmological applications. However, the limited redshift range and statistics of the current samples do not allow us to obtain significant cosmological constraints yet. We show how this is going to change with the future releases of the eROSITA data.

Figures (7)

• Figure 1: Flux in the 0.5--2 keV band versus redshift for the source in the eFEDS (red circles) and the eRASS1 (blue squares) parent samples. Given the high density of objects, contours with matching colors have been added to highlight the sources' distributions.
• Figure 2: Flux in the 0.5--2 keV band versus redshift for the source in the eFEDS (red circles) and the eRASS1 (blue squares) final samples.
• Figure 3: X-ray--UV relation. In red circles and blue squares are eFEDS and eRASS1 data, respectively. The best fits for the samples are reported by the dashed and dash-dotted lines. The gray dots represent the eRASS and eFEDS parent sample, before any filtering has been applied.
• Figure 4: Parameters of the flux-flux relation as a function of redshift. The top panel shows the slope ($\gamma$), the gray line indicates the standard $0.6$ value for the slope, and the bottom panel shows the intrinsic dispersion ($\delta$). In blue and red are indicated, respectively, eFEDS and eRASS1 data.
• Figure 5: Hubble diagram for the simulated eRASS:4.5 sample. The density plot represents the individual sources, while averaged values are indicated by black data points. In lime are shown SNe Ia from scolnic18. The solid green line shows the cosmographic model.