On the nature of high X-ray luminosities in SDSS galaxies

Surveys have revealed a class of object displaying both high X-ray luminosities (Lx > 10^42 erg/s), and a lack of a discernible active galactic nucleus (AGN) in the optical band. If these sources are powered by star formation activity alone, they would be the most extreme X-ray luminosity star forming galaxies known. We have investigated the mechanism driving the X-ray luminosities of such galaxies by studying the X-ray emission of three moderate redshift (z ~ 0.1) examples of this class, selected from a cross-correlation of the SDSS-DR5 and 2XMMp-DR0 catalogues. X-ray spatial and long-term variability diagnostics of these sources suggest that they are compact X-ray emitters. This result is supported by the detection of rapid short term variability in an observation of one of the sources. The X-ray spectra of all three sources are best fitted with a simple absorbed power-law model, thus betraying no significant signs of star formation. These results indicate that the X-ray emission is powered by AGN activity. But why do these sources not display optical AGN signatures? We show that the most likely explanation is that the optical AGN emission lines are being diluted by star formation signatures from within their host galaxies.

💡 Research Summary

The paper addresses a puzzling class of galaxies that exhibit very high X‑ray luminosities (L_X > 10^42 erg s⁻¹) yet show no obvious active‑galactic‑nucleus (AGN) signatures in their optical spectra. If the X‑ray output were powered solely by star formation, these objects would represent the most extreme star‑forming galaxies known in X‑rays. To determine the true origin of the emission, the authors selected three moderate‑redshift (z ≈ 0.1) examples from a cross‑match of the Sloan Digital Sky Survey Data Release 5 (SDSS‑DR5) and the 2XMMp‑DR0 catalogues.

First, they examined the spatial extent of the X‑ray sources using XMM‑Newton EPIC images. All three objects appear point‑like, with no detectable extension beyond the instrument’s point‑spread function, indicating that the X‑ray emission is confined to the galactic nucleus rather than being distributed across the host.

Next, they investigated variability on both long and short timescales. By comparing observations separated by several years, they found flux changes of 30 % or more for two of the galaxies, a level of variability that is typical of AGN but not of diffuse star‑formation‑related X‑ray emission. In one source, intra‑observation variability on the order of hours was detected, providing further evidence for a compact, rapidly changing emitter.

Spectral analysis was performed with XSPEC, fitting each spectrum with an absorbed power‑law model. The best‑fit photon indices (Γ ≈ 1.8–2.0) and modest intrinsic column densities (N_H ≈ 10^21–10^22 cm⁻²) are consistent with unobscured or mildly obscured AGN. Adding thermal plasma components (kT ≈ 0.5 keV) or searching for Fe Kα emission lines—features commonly associated with star‑forming galaxies—did not improve the fits, and any thermal contribution was constrained to less than ~5 % of the total flux.

Optical spectra from SDSS place these objects in the star‑forming region of the BPT diagnostic diagram. The authors argue that the lack of AGN emission lines is not due to the absence of an active nucleus but rather to dilution: the 3‑arcsecond SDSS fiber captures light from the entire host galaxy, and the strong star‑formation continuum and nebular lines overwhelm the relatively weak AGN signatures. Supporting this, the host galaxies have colors and star‑formation rates (~10 M_⊙ yr⁻¹) indicative of vigorous star formation that can mask nuclear line emission.

In summary, the combination of point‑like X‑ray morphology, significant short‑ and long‑term variability, and power‑law dominated spectra strongly points to low‑luminosity AGN as the power source for these high‑X‑ray‑luminosity galaxies. The apparent optical “quietness” is best explained by line dilution from intense star formation within the host. This work demonstrates that X‑ray variability and spectral diagnostics are essential tools for uncovering hidden AGN that optical surveys alone may miss, and it highlights the need for systematic multi‑wavelength follow‑up of similar objects in future large surveys.