Do Some AGN Lack X-ray Emission?

TL;DR

This study investigates whether metal-poor dwarf galaxies host AGN with IMBH/SMBH seeds by combining new Chandra X-ray data and multi-epoch optical spectroscopy for seven broad-H$\ alpha$ emitters. The targets show persistent broad H$\ alpha$ emission but exhibit a pronounced X-ray deficit relative to established AGN luminosity relations, with $L_{\rm 2-10\,keV}$ typically $1$–$2$ dex below expectations. The lack of strong high-ionization lines, coupled with MIR colors that can be produced by star-forming activity, leaves open whether these sources are truly accreting black holes or extreme stellar phenomena; the broad lines themselves may not reliably indicate BH mass in this regime. The results imply either X-ray/UV–weak AGN activity in metal-poor dwarfs or the need to reassess broad-line diagnostics, motivating deeper, high-resolution follow-up across X-ray, optical, and radio wavelengths.

Abstract

$Context:$ Intermediate-Mass Black Holes (IMBHs) are thought to be the seeds of early Supermassive Black Holes (SMBHs). While $\gtrsim$100 IMBH and small SMBH candidates have been identified in recent years, few have been robustly confirmed to date, leaving their number density in considerable doubt. Placing firmer constraints both on the methods used to identify and confirm IMBHs/SMBHs, as well as characterizing the range of host environments that IMBHs/SMBHs likely inhabit is therefore of considerable interest and importance. Additionally, finding significant numbers of IMBHs in metal-poor systems would be particularly intriguing, since such systems may represent local analogs of primordial galaxies, and therefore could provide clues of early accretion processes. $Aims:$ Here we study in detail several candidate Active Galactic Nuclei (AGN) found in metal-poor hosts. $Methods:$ We utilize new X-ray and optical observations to characterize these metal-poor AGN candidates and compare them against known AGN luminosity relations and well-characterized IMBH/SMBH samples. $Results:$ Despite having clear broad optical emission lines that are long-lived ($\gtrsim$10--13\,yr), these candidate AGN appear to lack associated strong X-ray and hard UV emission, lying at least 1--2 dex off the known AGN correlations. If they are IMBHs/SMBHs, our constraints imply that they either are not actively accreting, their accretion disks are fully obscured along our line-of-sight, or their accretion disks are not producing characteristic high energy emission. Alternatively, if they are not AGN, then their luminous broad emission lines imply production by extreme stellar processes. The latter would have profound implications on the applicability of broad lines for mass estimates of massive black holes.

Figures (6)

• Figure 1: [O iii]/H$\beta$ against [N ii]/H$\alpha$ narrow line emission ratios. Vertical and horizontal dotted lines mark the boundaries of the three main classes of emission line nuclei from ho. The dashed curve shows the demarcation between starburst galaxies and AGNs defined by kewley01. The solid black curve shows the demarcation between star-forming H ii regions and starburst galaxies from kauffmann, while the solid red, blue and green curves denote the AGN-H ii separation from kewley13, corrected to redshifts of $z$$=$0.1, 0.2 and 0.3, respectively. Filled red and green triangles represent the I08 and I07 objects, respectively. Open circles represent the GH04 objects. The I08 and I07 galaxies lie on the left side near the boundary separating AGN from star-forming regions. Their location is consistent with the stellar origin, but non-thermal radiation is not excluded. It is important to note that metal-poor AGN will have weaker [N ii] and [O iii] emission, such that they migrate from the right side of the diagram to the left groves, making the standard diagnostic more ambiguous. Thus, the origin of the ionising process cannot be determined by this diagram alone.
• Figure 2: Hard X-ray (2--10 keV) emission compared to broad H$\alpha$ emission (left) and [O iii] emission (right). The I08 and I07 objects are denoted by red and green filled triangles, respectively. The best-fit regressions for AGN from P06 are shown as solid black lines, while the individual AGN used by P06 are shown as open black circles. The dotted line shows the hard X-ray (2--10 keV) -- star formation rate relation presented by ranalli and kennicutt. Additional objects from GH04 and the literature are denoted by filled blue circles and filled grey triangles, respectively. Downward arrows denote X-ray upper limits for all samples.
• Figure 3:
• Figure 4: Time evolution of the broad H$\alpha$ luminosity for the I08 objects, as well as several luminous type IIn SNe kuncarayaktiaretxagastritzingersmithjencson and the transient event SDSS1133 koss. The I08 objects are roughly constant over periods of 10--13 yr; there is some variability between epochs at the $\sim$50% level, but this could be due to calibration uncertainties. All error bars are smaller than the symbols.
• Figure 5: Optical spectra for galaxies J0045+1339 and J1222+3602 obtained using the LBT telescope on UT 2013 October 27 and UT 2015 May 18, respectively. The broad H$\alpha$ flux does not present considerable variations compared to previous measurements.