The INTEGRAL/IBIS AGN catalogue I: X-ray absorption properties versus optical classification

In this work we present the most comprehensive INTEGRAL AGN sample which lists 272 objects. Here we mainly use this sample to study the absorption properties of active galaxies, to probe new AGN classes and to test the AGN unification scheme. We find that half (48%) of the sample is absorbed while the fraction of Compton thick AGN is small (~7%). In line with our previous analysis, we have however shown that when the bias towards heavily absorbed objects which are lost if weak and at large distance is removed, as it is possible in the local Universe, the above fractions increase to become 80% and 17%. We also find that absorption is a function of source luminosity, which implies some evolution in the obscuration properties of AGN. Few peculiar classes, so far poorly studied in the hard X-ray band, have been detected and studied for the first time such as 5 XBONG, 5 type 2 QSOs and 11 LINERs. In terms of optical classification, our sample contains 57% of type 1 and 43% of type 2 AGN; this subdivision is similar to that found in X-rays if unabsorbed versus absorbed objects are considered, suggesting that the match between optical and X-ray classification is overall good. Only a small percentage of sources (12%) does not fulfill the expectation of the unified theory as we find 22 type 1 AGN which are absorbed and 10 type 2 AGN which are unabsorbed. Studying in depth these outliers we found that most of the absorbed type 1 AGN have X-ray spectra characterized by either complex or warm/ionized absorption more likely due to ionized gas located in an accretion disk wind or in the biconical structure associated to the central nucleus, therefore unrelated to the toroidal structure. Among 10 type 2 AGN which resulted to be unabsorbed, at most 3-4% is still eligible to be classified as a “true” type 2 AGN.

💡 Research Summary

This paper presents the most extensive hard‑X‑ray AGN catalogue obtained with INTEGRAL/IBIS, comprising 272 active galaxies. The authors use the sample to investigate X‑ray absorption properties, to explore poorly studied AGN subclasses, and to test the predictions of the AGN unified scheme.

First, the authors determine that 48 % of the whole sample shows significant absorption (column density N H > 10^22 cm⁻²) and only about 7 % are Compton‑thick (N H > 1.5 × 10^24 cm⁻²). However, hard‑X‑ray surveys are biased against distant, faint, heavily absorbed sources. By restricting the analysis to the local Universe (z < 0.05) where the survey is essentially complete, they correct for this bias and find that the true fractions rise dramatically: roughly 80 % of AGN are absorbed and 17 % are Compton‑thick. This demonstrates that the hard‑X‑ray band uncovers a much larger hidden AGN population than soft‑X‑ray surveys.

A key result is the clear anti‑correlation between absorption and intrinsic luminosity. High‑luminosity AGN (L_2‑10 keV > 10^44 erg s⁻¹) are far less likely to be absorbed, consistent with the “receding torus” scenario in which the dusty torus opening angle widens as the radiation pressure of the central engine increases.

Optically, the catalogue contains 57 % type 1 and 43 % type 2 AGN. When the X‑ray classification (unabsorbed vs. absorbed) is compared with the optical one, the correspondence is excellent: the fractions of type 1/2 match the fractions of unabsorbed/absorbed to within a few percent, confirming the overall validity of the unified model.

Nevertheless, 12 % of the sources are outliers. Twenty‑two type 1 AGN exhibit significant X‑ray absorption. Detailed spectral fitting reveals that most of these objects possess complex or warm (ionised) absorbers, often modelled with multiple column components or ionised gas. The authors argue that such absorption likely originates in an accretion‑disk wind or in a biconical ionised outflow rather than in the classical dusty torus, explaining why the optical classification remains type 1.

Conversely, ten type 2 AGN appear X‑ray unabsorbed. Only a small subset (≈3–4 % of the total sample) could be genuine “true type 2” AGN—objects lacking a torus altogether. The remainder are probably mis‑classifications, or they may be experiencing temporary reductions in line‑of‑sight column density.

The hard‑X‑ray selection also enables the first systematic detection of several rare classes: five X‑ray bright optically normal galaxies (XBONG), five type 2 quasars, and eleven low‑ionisation nuclear emission‑region galaxies (LINERs). Their inclusion broadens the demographic picture of the AGN population in the hard X‑ray sky.

In summary, the INTEGRAL/IBIS AGN catalogue confirms the basic premises of the unified scheme—optical type correlates with X‑ray absorption—while highlighting important refinements. Absorption is strongly luminosity‑dependent, the torus is not the sole absorber (disk winds and ionised outflows play a major role), and a non‑negligible fraction of AGN are missed by soft‑X‑ray surveys due to bias. The work underscores the necessity of hard‑X‑ray observations for a complete census of active galaxies and provides a robust benchmark for future population‑synthesis models of the X‑ray background.