Subaru High-z Exploration of Low-Luminosity Quasars (SHELLQs) XXII. Chandra observations of narrow-line quasar candidates at z>6

K. Iwasawa, R. Gilli, F. Vito, Y. Matsuoka, M. Onoue, M. A. Strauss, N. Kashikawa, Y. Toba, K. Shimasaku, K. Inayoshi, T. Nagao, N. Kawanaka, J. D. Silverman, T. Izumi, K. Kohno, Y. Ueda

First listed 2025-05-07 | Last updated 2025-12-23

Abstract

We report on Chandra X-ray observations of four narrow-line quasar candidates at z~6, selected from the SHELLQs project, based on the Subaru Hyper Suprime-Cam survey. These objects are characterised by narrow (FWHM<310 km/s), luminous (>1e44 erg/s) Lya and faint UV continuum (M_1450 = -22 - -21), prompting us to examine whether they are obscured luminous AGN at the epoch of reionization. However, none of these objects were detected by Chandra, giving an upper limit to their rest-frame 2-10 keV luminosity (Lx) of 2e44 erg/s (2 sigma), assuming a spectral slope Gamma=2. Subsequent rest-frame optical spectroscopy of these objects by the JWST-NIRSpec, presented in a companion paper, show weak broad Balmer emission at the base of narrow cores. With the scaling relation for low-redshift AGN, the observed strong [OIII]5007 flux of these sources would predict Lx to be around 1e45 erg/s, which is well above the Chandra upper limits. These optical spectra and X-ray quietness are reminiscent of JWST-selected broad-line AGN. We attribute the weak broad Balmer emission to the broad-line regions hidden partially by optically-thick obscuring matter which also hides the optical and X-ray continuum emission from the accretion disc. Compton-thick obscuration, which would strongly suppress X-ray emission, could be due to a dense inter-stellar medium that is often present in galaxies at high redshifts. Alternatively, the same effect could be obtained from an inflated disc at the innermost radii in a supercritical accretion flow, when the disc is viewed at inclined angles.

Short digest

Chandra followed up four z>6 SHELLQs narrow-line quasar candidates chosen for luminous Lyα and faint UV continua, and found no X-ray detections, placing L2–10 keV < 2×10^44 erg/s (2σ; Γ=2). JWST/NIRSpec reveals weak broad Balmer bases beneath narrow cores, while strong [O III] would predict Lx ≈ 10^45 erg/s from low‑z scaling, in stark tension with the Chandra limits. The authors argue for heavily buried nuclei—either Compton-thick columns in dense ISM or an inclined, inflated inner disc in supercritical accretion—making these sources higher-luminosity analogs of JWST-selected broad-line AGN at reionization. The work clarifies how extreme Lyα–bright, UV-faint quasars can hide powerful engines early on.

Key figures to inspect

• Figure 1: Use the CDFs to see how the four targets sit at FWHM(Lyα) < 500 km/s with extreme EW (>250 Å), quantifying their placement among SHELLQs outliers.
• Figure 2: Inspect where the four circled targets lie as >2σ outliers in M1450 vs L(Lyα), highlighting their suppressed UV continua at fixed Lyα luminosity compared to LAEs.
• Figure 3: Examine the stacked Lyα profile’s weak red wing to gauge any broad component width and consider BLR/outflow/resonant-scattering contributions consistent with partial hiding of the BLR.
• Figure 4: Check the 1–5 keV postage stamps to confirm non-detections within the 1″ circles and assess any nearby serendipitous sources or astrometric offsets that could affect upper limits.