UNCOVER: The growth of the first massive black holes from JWST/NIRSpec -- spectroscopic redshift confirmation of an X-ray luminous AGN at z=10.1

TL;DR

The paper reports JWST/NIRSpec Prism spectroscopy that spectroscopically confirms UHZ-1 at $z=10.07$, marking the highest-redshift X-ray luminous AGN with a secure redshift to date. The rest-frame UV/optical spectrum is dominated by host-galaxy light and shows weak emission lines, consistent with heavy obscuration of the AGN, while the X-ray data indicate a Compton-thick, accreting black hole. SED fitting yields a stellar mass of roughly a few times 10^8 solar masses and a star-formation rate near 1–1.5 solar masses per year, with a very high inferred BH-to-host mass ratio that supports heavy-seed formation scenarios for SMBHs. These results imply that heavily obscured, early AGN could contribute to SMBH growth and possibly reionization, and demonstrate JWST’s power to uncover and characterize such systems through NIRSpec spectroscopy and multiwavelength analyses.

Abstract

The James Webb Space Telescope is now detecting early black holes (BHs) as they transition from "seeds" to supermassive BHs. Recently Bogdan et al. (2023) reported the detection of an X-ray luminous supermassive BH, UHZ-1, with a photometric redshift at $z > 10$. Such an extreme source at this very high redshift provides new insights on seeding and growth models for BHs given the short time available for formation and growth. Harnessing the exquisite sensitivity of JWST/NIRSpec, here we report the spectroscopic confirmation of UHZ-1 at $z = 10.073 \pm 0.002$. We find that the NIRSpec/Prism spectrum is typical of recently discovered z~10 galaxies, characterized primarily by star-formation features. We see no clear evidence of the powerful X-ray source in the rest-frame UV/optical spectrum, which may suggest heavy obscuration of the central BH, in line with the Compton-thick column density measured in the X-rays. We perform a stellar population fit simultaneously to the new NIRSpec spectroscopy and previously available photometry. The fit yields a stellar mass estimate for the host galaxy that is significantly better constrained than prior photometric estimates ($M_*\sim 1.4^{+0.3}_{-0.4} \times 10^8 M_\odot$). Given the predicted BH mass ($M_{\rm BH}\sim10^7-10^8 M_\odot$), the resulting ratio of $M_{\rm BH}/M_*$ remains two to three orders of magnitude higher than local values, thus lending support to the heavy seeding channel for the formation of supermassive BHs within the first billion years of cosmic evolution.

Figures (4)

• Figure 1: From top-left: NIRSpec/Prism MSA shutter positions for UHZ1 (see § \ref{['sec:data']}), JWST/NIRCam images of UHZ1 in filters F115W, F150W, F200W, F277W, F356, F410M and F444W (photometric measurements from these calibrated data are presented in Fig. \ref{['fig:spectra']}) and Chandra X-ray images in the 0.5-2 and 2-7 keV bands (smoothed with a 1 pixel width Gaussian filter). JWST cutout images are 1.5$"$ on a side, Chandra images are 5$"$ on a side. These are oriented in standard North-East convention.
• Figure 2: JWST/NIRSpec Prism spectroscopy of UHZ-1. Upper panel: 2D MSA Prism spectroscopy produced by msaexp. Lower panel: 1D spectral extraction in $f_{\lambda}$ (in units of $10^{-17}$erg s$^{-1}$ cm$^{-2}$$\mu$m$^{-1}$) with associated statistical uncertainties (gray shaded region). Slit-loss corrections are defined by convolution of the JWST photometry with the Prism spectrum (see § \ref{['sec:data']}). Prominent and/or expected emission features are highlighted assuming $z_{\rm spec}=10.07$ with significant $>3\sigma$ detections and non-detections labeled in red and gray, respectively. Overlaid are the JWST/NIRCam photometry (blue circles) with associated filter responses highlighted. Inset panel: Redshift probability distributions for fits to the NIRSpec spectroscopy produced by EAZY (yellow) and BAGPIPES (purple) packages.
• Figure 3: Upper panels: Gaussian line combined with low-order polynomial continua fits to the C iii] $\lambda \, 1907,1909$, [O ii] $\lambda \, 3727,3729$ doublet and [Ne iii] $\lambda \, 3869$ emission features. Lower panels: Residuals of the best fits.
• Figure 4: SPS modeling with Bagpipes. The left panel shows the observed galaxy SED and spectrum (black, showing only JWST photometry) after the application of the polynomial calibration with 1$\sigma$ errors (assuming an error floor of 5%) and the median model (red). Additionally, the median model is shown at the full resolution (gold), highlighting the predicted emission features that are washed out by the instrumental resolution. In the right panel, we show the covariant posteriors for a number of key measured parameters after accounting for the magnification of the source.