EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 571 (November 2014) A&A, 571 (2014) A34 Full HTML
Free Access
Issue
A&A
Volume 571, November 2014
Article Number A34
Number of page(s) 10
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/201424791
Published online 04 November 2014

© ESO, 2014

1. Introduction

The existence of a close link between super massive black holes (SMBHs) and their host galaxies has found support in many works over the last twenty years, starting from the seminal paper by Kormendy & Richstone (1995). Several relations in the local Universe between the mass of the black holes and the properties of the host galaxies (e.g., their velocity dispersion; Gebhardt et al. 2000; Ferrarese & Merritt 2000) are signatures of this connection. Most active galactic nuclei (AGN)-galaxy co-evolution models predict the existence of a dust-enshrouded phase characterized by rapid SMBH growth and active star formation, largely triggered by multiple galaxy mergers and encounters (e.g., Silk & Rees 1998; Di Matteo et al. 2005; Menci et al. 2008; Zubovas & King 2012; Lamastra et al. 2013). This phase is associated with obscured AGN growth in strongly star-forming galaxies, as already shown a decade ago by Alexander et al. (2005) using sub-millimeter selection and deep X-ray data. Then massive AGN-driven outflows blow away most of the cold gas reservoir; the AGN become unobscured (i.e., Type 1) and, when the phase of significant nuclear accretion is over because of the limited amount of remaining available gas, a population of what are known as red-and-dead gas-poor elliptical galaxies is what is left (e.g., Hopkins et al. 2008; Cattaneo et al. 2009). If, on the one hand, this scenario seems to provide a valid explanation for the most luminous and massive systems, for which a large fraction of the mass is assembled in relatively short periods (≈10−100 Myr) of intense nuclear and star-forming activity, on the other hand the bulk of galaxies and SMBHs are likely to grow their mass in a secular (i.e., smooth) mode over timescales of Gyrs (e.g., Daddi et al. 2007a; Hickox et al. 2009; Villforth et al. 2014, and references therein).

Although there is a large quantity of evidence for feedback processes in terms of outflows in molecular and neutral/ionized gas, at both low and high redshift (e.g., Nesvadba et al. 2008; Feruglio et al. 2010; Maiolino et al. 2012; Harrison et al. 2012, 2014; Cicone et al. 2014; Brusa et al. 2014), including signatures in the X-ray band (e.g., Chartas et al. 2002; Tombesi et al. 2012, and references therein), this picture still needs to be confirmed in most of its aspects. In particular, we need to investigate the crucial phase during which large amounts of gas are funneled to the center, thus inducing both obscured accretion and star formation (e.g., Treister et al. 2010). Type 2 (obscured) AGN are evidence of and, at the same time, main actors in this phase; finding them across cosmic time is therefore crucial to place constraints on AGN vs. galaxy co-evolution models (see, e.g., the reviews by Merloni & Heinz 2013 and Vignali 2014) as well as on X-ray background synthesis models (XRB; e.g., Gilli et al. 2007; Ballantyne 2009; Treister et al. 2009; Akylas et al. 2012; Shi et al. 2013; see also Moretti et al. 2012).

A complete census of absorbed AGN, especially of the most heavily obscured (called Compton thick, having column densities above 1.5 × 1024 cm-2; see Comastri 2004 for a review) up to high redshift cannot be achieved using either a single selection or a single observing-band method. Although the hard X-ray band up to ≈100–150 keV is potentially appropriate to provide an almost unbiased census of obscured AGN, since both Compton-thin (with NH< 1.5 × 1024 cm-2) and Compton-thick AGN (at least the sources having column densities up to ≈1025 cm-2) can be detected by hard X-ray instruments, the sensitivity of Swift/BAT and INTEGRAL/IBIS surveys (≈10-11erg cm-2 s-1; e.g., Tueller et al. 2008; Beckmann et al. 2009; Burlon et al. 2011; Ajello et al. 2012; Vasudevan et al. 2013) is such to limit this kind of investigations to the local Universe. Interesting prospects might come from the much higher sensitivity (a factor of ≈100) hard X-ray imaging instrument onboard NuSTAR (e.g., Alexander et al. 2013; Lansbury et al. 2014).

A proper evaluation of obscured AGN at higher redshifts and at fainter X-ray fluxes requires deep X-ray surveys with Chandra and XMM-Newton (e.g., Tozzi et al. 2006; Georgantopoulos et al. 2009, 2013; Comastri et al. 2011; Feruglio et al. 2011; Gilli et al. 2011, 2014; Iwasawa et al. 2012a; Brightman & Ueda 2012; Vito et al. 2013; Buchner et al. 2014; Brightman et al. 2014). However, even the deepest X-ray exposures currently available miss a significant number of very obscured AGN, especially the most heavily Compton-thick AGN at low-medium redshift (because of the energy bandpass, poor counting statistics and sensitivity), hence a not negligible fraction of the accretion power in the Universe, and convey the idea that a multi-wavelength approach to the search for obscured AGN is mandatory.

The availability of deep radio, optical and mid-infrared (mid-IR) data has recently allowed the definition of complementary methods to pick up obscured AGN up to high redshifts. In particular, sources with weak emission in the optical band (due to dust extinction) and relatively bright mid-IR emission (due to the disk emission being thermally reprocessed by the AGN torus) can be considered good obscured AGN candidates (e.g., Martínez-Sansigre et al. 2005; Houck et al. 2005; Weedman et al. 2006; Dey et al. 2008), unless a significant contribution to the mid-IR comes from star formation processes (PAH features and continuum emission). This selection of obscured AGN, albeit not complete, represents a step forward with respect to mid-IR color-color diagrams, where separating heavily obscured AGN from the remaining AGN populations is not trivial (e.g., Lacy et al. 2004, 2013; Stern et al. 2005; Donley et al. 2012; Castelló-Mor et al. 2013; Mateos et al. 2013). However, the selection of obscured AGN by means of long-wavelength data needs to be corroborated by X-ray information, as plenty of works have shown in the last decade (e.g., Polletta et al. 2006; Fiore et al. 2008; Lanzuisi et al. 2009; Severgnini et al. 2012; Del Moro et al. 2013). This observational approach has led many authors to place constraints, for the first time, to the space density of Compton-thick AGN at high redshifts (z ≈ 2−3; Daddi et al. 2007b; Fiore et al. 2009; Bauer et al. 2010; Alexander et al. 2011).

Obscured AGN can be selected also on the basis of narrow, high-ionization emission lines, like [O iii]5007 Å and [Ne v]3426 Å. These lines are produced in the narrow-line region (NLR), so they do not suffer from the effects of attenuation ascribed to the torus obscuring material (although some extinction within the NLR may be present). As such, these lines can be considered good proxies of the nuclear emission from the AGN. Previous works have shown that coupling AGN selection in one of these lines with X-ray observations may be effective in finding highly reliable Compton-thick AGN candidates (e.g., Vignali et al. 2006, 2010; Ptak et al. 2006; LaMassa et al. 2009; Lamastra et al. 2009; Gilli et al. 2010, hereafter G10; Jia et al. 2013; but see also Trouille & Barger 2010). Once again, X-ray information is crucial to infer the nature of such sources, at least on a statistical ground, with additional enlightening indications coming from the mid-IR emission (e.g., Vignali et al. 2010; Goulding et al. 2011; see also Gandhi et al. 2009).

In this paper we present the X-ray properties of a sample of Type 2 AGN selected from the 20k-zCOSMOS Bright survey (Mignoli et al. 2013, hereafter M13). The optical selection and X-ray (mostly Chandra) coverage are shown in Sect. 2. Results from X-ray spectroscopy and stacking analyses are reported in Sect. 3, along with the estimated fraction of Compton-thick AGN. Their space density at z ≈ 0.8 is then computed in Sect. 4, where current results are also compared with expectations based on AGN synthesis models of the XRB. Conclusions are then reported in Sect. 5.

Hereafter we adopt a cosmology with H0 = 70 km s-1 Mpc-1, ΩM = 0.3 and ΩΛ = 0.7.

2. The sample: optical selection and Chandra coverage

The selection of the [Ne v] Type 2 AGN sample has been extensively described and discussed in M13. In the following, we will report on the main issues related to the optical selection process, the X-ray coverage, the fraction of X-ray detections, and the X-ray flux estimates needed to place sources in the X/NeV flux ratio diagnostic histogram (G10).

Type 2 AGN were selected from the zCOSMOS spectroscopic survey (Lilly et al. 2007, 2009), which was designed to observe ≈20 000 galaxies at I-band magnitudes brighter than 22.5 over the entire ≈2 square degree Cosmic Evolution Survey (COSMOS) field (Scoville et al. 2007), at a medium resolution (R ≈ 600) and in the 5500–9700 Å spectral range with the VLT instrument VIMOS. Hereafter, this spectroscopic survey will be referred to as the 20k-zCOSMOS Bright survey.

Ninety-four Type 2 AGN were selected from the 20k-zCOSMOS survey in the redshift range ≈0.65–1.20. The lower redshift boundary guarantees that both [Ne v]3346 Å and [Ne v]3426 Å emission lines fall within the spectral coverage; the choice of the upper boundary is due to the strong fringing of VIMOS spectra at long wavelengths, which makes the process of line detection more difficult and less reliable. The Type 2 classification was based on accurate analysis of all of the optical spectra, showing only narrow features. The presence of significant emission from broad emission-line components is ruled out, as clearly shown in the composite spectrum of the Type 2 AGN sample in M13 (see their Fig. 2). The rest-frame equivalent width (EW) of the [Ne v] emission line spans the range ≈2−35Å, with a median value of ≈8 Å.

Seventy-one of these [Ne v]-selected Type 2 AGN fall in the Chandra-COSMOS mosaic (≈ 0.9 square degrees). Two of these sources fall in a bad position of the mosaic and were therefore discarded. The choice of focusing on Chandra instead of XMM-Newton to derive the X-ray properties of this sample is due to the typically deeper exposure in the Chandra mosaic (≈160 ks over the inner 0.45 square degrees – thus reaching fainter flux limits than XMM-COSMOS by a factor of ≈3.5 – and an outer region of ≈0.4 square degrees with an effective exposure of up to ≈80 ks; see Puccetti et al. 2009; Elvis et al. 2009; Civano et al. 2012, hereafter C12), and to its better performances when investigation concerns X-ray faint sources. The low and “smooth” background level of Chandra observations is also an important ingredient for stacking analyses.

Twenty-three [Ne v]-selected Type 2 AGN were detected by Chandra (see Table 1) within 1.2′′ from the optical position. The median displacement between the X-ray and the optical counterpart, 0.46′′, and the maximum likelihood technique presented by C12 assure the reliability of the associations. We have visually inspected the four sources with a counterpart at a distance larger than 1 arcsec and found they are good matches.

To characterize the sources in terms of obscuration, we used the X/NeV flux ratio diagram (as in G10), where X corresponds to the rest-frame 2–10 keV flux with no absorption correction, and NeV is the flux of the emission line derived from the zCOSMOS spectra after applying aperture corrections, whose median value corresponds to a factor of ≈1.5 (see Sect. 6 of M13). The rest-frame 2–10 keV band corresponds to the observed-frame ≈1–5.5 keV band. We have extracted the 23 [NeV]-selected source information from the multiple pointings of the Chandra-COSMOS mosaic using the procedure outlined in M13 and derived X-ray photometry in the chosen energy interval. Because of the tiles in the C-COSMOS field, every source can be observed in more than one pointing (up to six) and at different locations within the ACIS-I field of view, hence the source count distribution is characterized by different (i.e., multiple) point spread function (PSF) sizes and shapes. To properly account for these effects, we have used the Acis Extract software (Broos et al. 2010), which extracts the source counts from each observation using the 90% of the encircled energy fraction at 1.5 keV at the source position and then corrects for the chosen aperture. At this stage, counts were extracted in the rest-frame 2–10 keV for each source and divided by the effective (i.e., “vignetting-corrected”) exposure time at each source position (averaged over an area of 5-arcsec radius to minimize the possible effects of small-scale fluctuations in the timemaps) to obtain a count rate. The X-ray flux has then been obtained by a simple count rate to flux conversion, assuming a powerlaw model with photon index Γ = 1.4 modified by Galactic absorption. The corresponding X/NeV flux ratios are plotted in Fig. 1 (red filled region).

Along with photometry, X-ray spectra were also extracted and fitted with a simple model (a powerlaw modified by Galactic absorption), because of the limited photon statistics for most of these sources (see Sect. 3.1). This X-ray spectral analysis allowed us to derive another estimate for the rest-frame 2–10 keV fluxes (see Table 1, where two X/NeV columns are shown, corresponding to the two methods used to derive X-ray fluxes). The X/NeV flux ratios corresponding to this basic X-ray spectral analysis are plotted in Fig. 1 as a blue dotted histogram. Details on the X-ray properties of these 23 Type 2 AGN are presented in Sect. 3.1.

thumbnail Fig. 1

Distribution of the rest-frame 2–10 keV (with no correction for absorption) to [NeV] luminosity ratio (X/NeV) of the 69 zCOSMOS Type 2 AGN with reliable X-ray photometry. X-ray detections are reported as either red (filled) or blue (dotted) histograms; in particular, the red histogram refers to X-ray photometry obtained from the Acis Extract analysis carried out in the rest-frame 2–10 keV band, while the blue histogram reports the X-ray flux values derived from direct X-ray spectral fitting in the same energy band (see text for details). The black histogram refers to the X-ray undetected [Ne v]-selected Type 2 AGN. The vertical dashed line shows the threshold defined by G10 for Compton-thick (leftward direction) and Compton-thin (rightward direction) sources. All of the X-ray detected sources are likely Compton thin using the X/NeV threshold of 15 adopted by G10. The filled symbols represent the average X/NeV value obtained from the X-ray stacking analysis (black circle: stack of the 46 X-ray undetected Type 2 AGN; green triangle: stack of the 22 sources with “nominal” upper limit on the X/NeV ratio <30; magenta square: stack of the 24 sources with “nominal” upper limit on the X/NeV ratio >30; see Sect. 3.2 for details). Their y-axis position is arbitrary.

Table 1

List of X-ray detected [NeV]-selected Type 2 AGN.

A similar, Acis Extract-based procedure has been adopted for the X-ray photometry of the remaining 46 [Ne v]-selected Type 2 AGN with no X-ray detection in the C12 catalog (and no clear X-ray emission in the 0.5–7 keV images). To derive X-ray flux upper limits, at each source optical position counts were extracted taking into account the PSF size and shape in the observed band equivalent to the rest-frame 2–10 keV energy interval. One-sigma count upper limits have then been converted into count rates using the exposures derived from the timemaps at each source position, and count rates were then converted into X-ray flux upper limits assuming a powerlaw with photon index Γ = −0.4, which is broadly consistent with a reflection-dominated spectrum. These values were then used to derive the X/NeV flux ratio upper limits reported in Table 2, similarly to the analysis presented in M13. We note that Monte Carlo runs were then used to properly interpret the results obtained for the X-ray undetected Type 2 AGN given their limited photon statistics (see Sect. 3.2 for details).

The different behavior in the X-ray band between the sample of 23 X-ray detected and that of 46 X-ray undetected Type 2 AGN cannot be ascribed to a difference in the exposure time; the former sample has an average exposure of ≈120 ks, while the latter is characterized by a slightly larger exposure, ≈128 ks. Furthermore, the two samples have comparable average redshift (z = 0.85 vs. z = 0.84) and similar [Ne v] flux (3.1 × 10-17 vs. 2.8 × 10-17erg cm-2 s-1). Therefore, we motivate our choice of a different photon index for the count rate to flux conversion for the two samples under the hypothesis that X-ray undetected Type 2 AGN may suffer from heavier absorption than X-ray detected Type 2 AGN.

As a final remark, we note that the present sample of Type 2 AGN is far from being complete, as extensively discussed by M13. In particular, there are other ≈160 zCOSMOS galaxies in the same redshift interval having the properties of Type 2 AGN (either from optical spectroscopy or X-ray analysis or both) without the presence of [Ne v] emission. The mean optical extinction of this sample (derived from the Hβ/Hγ flux ratio, ⟨ E(B − V) ⟩ = 0.27 ) is higher than that derived for the [Ne v]-selected Type 2 AGN discussed in this paper (⟨ E(B − V) ⟩ = 0.18); the higher extinction is likely causing the lack of appreciable [Ne v] emission in the “additional” sample of Type 2 AGN (see Fig. 6 of M13). Summarizing, the AGN selection based on [Ne v] emission can be considered clean (because of the ionization potential of this line, unambiguously ascribed to accretion power) but not complete.

3. X-ray results

3.1. X-ray detected [NeV]-selected Type 2 AGN

For the 23 X-ray detected Type 2 AGN we were able to perform a basic X-ray spectral analysis assuming a powerlaw model. The main limitation of this analysis is due to the paucity of X-ray photons available in our data. The median number of net (i.e., background-subtracted) counts is 60. Two sources have ten counts or fewer, two sources have 140–150 counts (CID = 456 and 522, according to C12 source nomenclature), and two sources have 300–350 counts (CID = 110 and 339).

On the one hand, fitting all the X-ray data together with the same model (accounting for the different redshifts and fluxes of the sources) would produce biased results, given the presence of four AGN with more than 100 counts. On the other hand, fitting all the spectra individually using unbinned data and C-statistic (Cash 1979) within xspec (Version 12.7.0; Arnaud 1996) would not constrain properly the X-ray spectral parameters (Γ, NH and normalization) for most of the 23 sources. Therefore, we decided to fit the four sources with most counts using χ2 statistics and binned data (10–15 counts per bin); the remaining spectra were “summed” to produce a rest-frame 1–10 keV stacked spectrum (with a spectral resolution of ≈0.5 keV) – following the prescriptions extensively described in Iwasawa et al. (2012a,b) – and then fitted (again, adopting the χ2 statistics).

Errors are quoted at the 90% confidence level for one parameter of interest (Δχ2 = 2.71; Avni 1976). Solar abundances (from Anders & Grevesse 1989) and Galactic absorption (NH = 1.7 × 1020 cm-2; Kalberla et al. 2005) were adopted in all models.

CID339 and CID110 are characterized by single powerlaws with limited absorption at the source redshift (NH=3.8-2.8+3.2×1021\hbox{$N_{\rm H}=3.8^{+3.2}_{-2.8}\times10^{21}$} cm-2 and NH< 3.4 × 1021 cm-2, respectively, once a “canonical” AGN photon index Γ = 1.8 – e.g., Piconcelli et al. (2005) – is assumed; see panels (a) and (b) in Fig. 2); we note that the best-fitting Γ for the two sources (≈1.6–1.7) are very close to the adopted value. An emission line at rest-frame energy ≈6.9 keV is tentatively detected (≈2.5σ) in CID339. The other two relatively bright sources, CID522 and CID456, appear more absorbed, being characterized by Γ ≈ 0.9−1.2 if no absorption is included. Fixing the photon index to 1.8 and including obscuration in the fit produce a column density of 2.7-2.0+2.3 × 1022\hbox{$2.7^{+2.3}_{-2.0}\,\times\,10^{22}$} cm-2 and 3.1-1.0+1.4 × 1023\hbox{$3.1^{+1.4}_{-1.0}\,\times\,10^{23}$} cm-2, respectively (see panels (c) and (d) in Fig. 2).

Table 2

List of X-ray undetected [NeV]-selected Type 2 AGN.

thumbnail Fig. 2

Rest-frame X-ray spectra of the four Type 2 AGN with most counts (i.e., above 140 counts; CID339 a), CID110 b), CID522 c), and CID456 d)), characterized by increasing absorption from top to bottom, and stacked X-ray spectrum of the 19 Type 2 AGN with limited photon statistics (panel e)), showing an emission line at an energy consistent with iron Kα. An emission line is also apparent in CID339. X-ray spectra are arbitrarily scaled for visual clarity.

The 1–10 keV stacked spectrum of the remaining sources is characterized by a flat slope and an emission feature which, at first glance, is consistent with iron emission (see the X-ray stacked spectrum in panel (e) of Fig. 2). If no absorption is assumed, Γ = 0.8 ± 0.1 is derived. Similarly to the sources above, the photon index was fixed to 1.8, and the derived column density is NH=2.2-0.5+0.6×1022\hbox{$N_{\rm H}=2.2^{+0.6}_{-0.5}\times10^{22}$} cm-2. This column density is consistent with the broad range of X/NeV (hence absorption) encompassed by these sources (see the filled red and open blue histograms in Fig. 1, and Table 1). The emission-line energy is 6.4 ± 0.2 keV, in agreement with either neutral or mildly ionized iron, and its equivalent width is EW = 0.7 ± 0.4 keV. Since a typical signature of strong obscuration is the presence of a strong (EW~> 1 keV) FeKα emission line on top of a flat X-ray continuum (e.g., Comastri et al. 2011; Feruglio et al. 2011; Iwasawa et al. 2012a), we might interpret the EW of the iron line in the stacked spectrum as suggestive of the presence of a few sources with heavy obscuration.

The X-ray spectral results obtained from the stacked spectrum were confirmed by fitting the spectra simultaneously (i.e., all of the sources were fitted at the same time using the same spectral model, leaving the normalization of each source free to vary; here C-statistic is adopted). Assuming Γ = 1.8, we obtain NH = (2.2 ± 0.4) × 1022 cm-2 (consistent with the analysis reported above) but the iron line does not emerge as strongly as in the stacked spectrum described above.

Among the analyzed X-ray spectra, of particular interest is the case of source CID1019, corresponding to the zCOSMOS object 813250 at z = 0.7302 (≈30 background-subtracted counts). A single powerlaw provides Γ ≈ −0.5, which is definitely an indication of heavy absorption and is consistent with a reflection-dominated spectrum below 10 keV, despite the relatively high X/NeV ratio (≈40−90, see Table 1). The tentative presence of an iron Kα emission at the rest-frame energy of ≈6.4 keV with EW ≈ 1.7 keV further supports this scenario (see the X-ray spectrum in Fig. 3).

Eighteen of the 23 Type 2 AGN were detected also by XMM-Newton. Typically, the signal-to-noise ratios in Chandra spectra are higher than in XMM-Newton spectra (see Mainieri et al. 2007) because of the much lower background guaranteed by the former satellite. Basic X-ray spectral analysis of XMM-Newton data confirms the Chandra results in terms of spectral properties. Interestingly, there are three sources showing X-ray flux variability by at least a factor of two, with CID1230 being a factor of ≈6 fainter in Chandra than in XMM-Newton. Unfortunately, the limited number of counts (≈30 net counts in the 0.5–7 keV band in Chandra, and ≈130 in XMM-Newton in the same energy band) prevented us from investigating the nature of this difference properly.

thumbnail Fig. 3

Observed-frame X-ray spectrum of CID1019 (zCOSMOS_ID = 813250) at z = 0.7302. The flat X-ray continuum and the tentative iron Kα emission line at rest-frame energy of ≈6.4 keV are indicative of strong, likely Compton-thick absorption.

3.2. [NeV]-selected Type 2 AGN with no X-ray detection

Results for the sample of 46 X-ray undetected [Ne v]-selected Type 2 AGN have been obtained by X-ray stacking and X/NeV ratio analyses. Because of the limited photon statistics, the former approach is not highly effective in constraining the average spectral properties of the sources. The fitted photon distribution (130 counts in the observed-frame 0.5–7 keV band) is consistent with a powerlaw with Γ = 0.9 ± 0.5, indicative of obscuration (see the rest-frame 1–10 keV spectrum in Fig. 4). The derived average flux and luminosity of this sample in the rest-frame 2–10 keV band are ≈2.5 × 10-16erg cm-2 s-1and ≈8.4 × 1041 erg s-1, respectively.

A step forward in the analysis of this sample consists of placing their X/NeV upper limits in Fig. 1 (open histogram). At face value, 39 of these sources have upper limits below 100, i.e., consistent with column densities possibly exceeding 1023 cm-2 (see G10). Nine of these upper limits are located in the region likely populated by Compton-thick AGN (i.e., at X/Ne<15; G10). The average X/NeV ratio for these 46 sources is 9.8 (filled circle in Fig. 1).

To estimate the fraction of Compton-thick AGN in our sample as most reliably as possible, we divided the sample of 46 sources into two subsamples of similar size: 22 sources with “nominal” X/NeV ratio <30 (hereafter referred to as the “faint” subsample) and 24 sources with “nominal” X/NeV ratio >30 (the “bright” sample), as also described in M13. Stacked X-ray photometry for the former subsample indicates a 1.6σ detection in the rest-frame 2–10 keV band, compared to the 5.8σ detection for the latter subsample. The derived average X/NeV ratios are 3.6 and 36 for the “faint” and “bright” subsample, respectively (shown as filled triangle and square in Fig. 1, respectively).

thumbnail Fig. 4

Rest-frame 1–10 keV stacked spectrum of the 46 Type 2 AGN with no individual X-ray detection. The limited-quality spectrum is flat, suggestive of absorption, but apparently no significant iron Kα emission line is present (EW< 1.5 keV).

We then stacked the total sample of 46 X-ray undetected Type 2 AGN, the “faint” and “bright” subsamples; we obtained an average number of 2.6, 1.1 and 3.9 net counts per source, respectively. After verifying that the count distribution of the 46 sources follows a Poisson distribution (see details in M13), we performed 10 000 Monte Carlo runs, each time randomly extracting 46 count values from the Poisson distribution and converting these counts into X-ray fluxes (following the prescriptions reported in Sect. 2). Then these X-ray fluxes were randomly associated with the [Ne v] fluxes to produce X/NeV flux ratios. As a result of the Monte Carlo runs, an average number of 29.4 sources (2.7 rms) were found at X/NeV<15 and are entitled to be considered reliable Compton-thick AGN candidates, bringing the fraction of Compton-thick AGN in the present sample from 13 ± 5%, i.e., 9/69 originally observed at X/NeV<15, to 43 ± 6% (i.e., 29.4/69, where the reported error is just statistical, and systematic errors related to source selection and stacking analysis are likely much higher). This fraction compares well with XRB synthesis model predictions (≈50%) by Gilli et al. (2007).

As a consistency check, we ran spectral simulations within xspec to verify whether the estimated fraction of Compton-thick AGN is consistent with the slope of the stacked spectrum of the 46 X-ray undetected sources (Γ = 0.9 ± 0.5). We assumed that the 30 sources with the most stringent limits on the X/NeV ratio are the Compton-thick candidates, whereas the remaining 16 objects are Compton thin. We assumed a pure reflection template spectrum (pexrav in xspec) for Compton-thick AGN and an absorbed powerlaw for Compton-thin AGN. A soft scattered component was also added in both templates, which mainly contributes at energies below 2 keV rest-frame (see Gilli et al. 2007 and G10 for details). The X-ray flux normalization of each Compton-thin source was obtained by assuming the real source redshift, an intrinsic 2–10 keV rest-frame luminosity of L2−10 keV = 400 × L[NeV], where L[NeV] is the measured [Ne v] line luminosity, and a column density estimated through the NH vs. X/NeV model relation of G10. Based on the measured ratio between the median of the X/NeV upper limits of the sources in the “bright” sample and the X/NeV value derived from their stacking, we assumed that each object has an X/NeV ratio a factor of 2 lower than its measured limit. A similar procedure was adopted to normalize the spectra of Compton-thick candidates, based on the “faint” sample. The flux normalization of the pexrav spectrum was obtained by assuming that it produces ≈2% of the intrinsic power in the 2–10 keV rest-frame energy band. All spectra were simulated by using an average instrumental response for C-COSMOS sources and by considering for each source its effective exposure in the C-COSMOS mosaic. The 46 spectra were finally co-added using the addspec ftool. The simulated stacked spectrum contains ~110 photons in the 0.5−7 keV band, and a simple powerlaw fit returns a photon index of Γ = 0.40 ± 0.50. This value is consistent (within 1σ) with that measured in the real stacked spectrum.

4. The space density of obscured and Compton-thick AGN at z ≈ 0.8

thumbnail Fig. 5

Distribution of the 69 [Ne v]-selected Type 2 AGN in the [Ne v]l˜ine luminosity vs. redshift plane. The 23 objects detected in the X-rays are marked with a cross. The black dashed lines show the boundaries of the three bins used in the computation of the [Ne v] luminosity function. A total of 50 AGN fall in these three luminosity bins. The blue dashed curve shows the [Ne v] flux limit (in units oferg cm-2 s-1) of the zCOSMOS survey.

Based on the standard 1 /Vmax method (Schmidt 1968), we first derived the luminosity function of [Ne v]-selected Type 2 AGN at z ≈ 0.8 and then the space density of Compton-thick objects among them. The distribution of our sources in the [Ne v] luminosity vs. redshift plane is shown in Fig. 5. For the luminosity function computation we only considered the 50 objects in the redshift range z = 0.65−1.01 and in the luminosity range logL[NeV] ≈ 40.6−41.3, which were further subdivided into three luminosity bins (see Fig. 5). We were forced to choose a single, rather broad redshift range to keep a reasonable number of objects in each luminosity bin. The median redshift of this 50-object sample is z = 0.83, and does not vary significantly across the three luminosity bins. The chosen luminosity thresholds guarantee the best completeness level above the [Ne v] flux limit of the zCOSMOS survey (f[NeV]> 7.5 × 10-18erg cm-2 s-1 at ≈5σ).

Sources belonging to our sample were first selected to have IAB< 22.5 as targets for zCOSMOS optical spectroscopy (see Lilly et al. 2009). Therefore, in the 1 /Vmax computation, we considered as Vmax the minimum among V[NeV], VIAB and Vzmax, where V[NeV] and VIAB and Vzmax are the maximum comoving volumes within which each object would still be included in the sample based on the [Ne v] flux cut, IAB magnitude cut, and our adopted redshift cut (z = 1.01), respectively. In addition, a correction that accounts for both the sampling and success rates of zCOSMOS optical spectroscopy needs to be introduced (see, e.g., Bongiorno et al. 2010, hereafter B10). In particular, by defining the target sampling rate (TSR) as the fraction of sources observed in the spectroscopic survey compared to the total number of objects in the parent photometric catalog, and the spectroscopy success rate (SSR) as the fraction of spectroscopically targeted objects for which a secure spectroscopic identification was obtained, trends of TSR and SSR as a function of object magnitude and redshift were determined (Zucca et al. 2009; Bolzonella et al. 2010). A weight can therefore be defined for each object in zCOSMOS as wi = 1 / (TSRi × SSRi). For reference, in our sample wi ranges from 1.06 to 2.56 and has a mean (median) value of 1.87 (1.93). The [Ne v] luminosity function in the considered redshift bin is therefore given by Φ(L)=1ΔlogL∑iwiVmax,i,\begin{equation} \Phi(L) = \frac{1}{\Delta \log L} \sum_i \frac{w_i}{V_{{\rm max}, i}}, \label{eqphi} \end{equation}(1)and the associated statistical uncertainty is given by (Marshall et al. 1983) σφ=1ΔlogL∑iwi2Vmax,i2·\begin{equation} \sigma_{\phi} = \frac{1}{\Delta \log L} \sqrt{\sum_i \frac{w_i^2}{V_{{\rm max}, i}^2}}\cdot \end{equation}(2)For comparison with literature results, the resulting [Ne v] luminosity function was converted into an [O iii] luminosity function using the relation L[OIII] = 9.1 × L[NeV], as derived from objects in our sample that are in the redshift range where both lines can be observed. This relation is also consistent with the average [Ne v]/[O iii] ratio derived from the zCOSMOS Type 2 AGN composite spectrum (M13) and from other literature samples (G10).

thumbnail Fig. 6

[O iii]-converted luminosity function of [Ne v]-selected Type 2 AGN at z = 0.65−1.01 in zCOSMOS (light blue triangles; this work) compared with the [O iii] luminosity function of zCOSMOS Type 2 AGN in a similar redshift range selected through the “blue” line diagnostic diagram (black circles; B10). Open symbols highlight those datapoints affected by significant incompleteness (see text and B10). Small numbers indicate how many objects were used to derive each datapoint.

In Fig. 6 we compare the [O iii]-converted luminosity function of our [Ne v]-selected Type 2 AGN at z = 0.65−1.01 with that derived by B10 for zCOSMOS Type 2 AGN selected by means of the [O iii]/Hβ vs. [O ii]/Hβ diagnostic diagram (Lamareille et al. 2004) in a similar redshift range (z = 0.50−0.92). With the exception of the datapoint in the faintest luminosity bin, which may suffer from some residual incompleteness (see below for a more complete discussion), the luminosity function of [Ne v]-selected Type 2 AGN at z ≈ 0.8 appears to be a factor of ~2 higher than that of Type 2 AGN in the same redshift and luminosity range derived by B10. This confirms the result of M13, who showed that a large fraction of Type 2 AGN, selected through their [Ne v] emission, is in fact missed by the optical classification based on the “blue” [O iii]/Hβ vs. [O ii]/Hβ diagnostic diagram. As also discussed by B10 for their [O iii] luminosity function, we note that the drop observed in our luminosity function at the faintest luminosities (see Fig. 6) may result from incompleteness caused by the two-step selection of our sample. In fact, the zCOSMOS magnitude cut at I = 22.5 may exclude from the luminosity function objects that would otherwise be in based on their [Ne v] flux. We computed the [Ne v] luminosity at which the luminosity function may suffer from severe incompleteness as follows. Based on the Type 2 AGN composite spectrum of M13, we computed the 3426 Å continuum luminosity corresponding to IAB = 22.5 at z = 1.01 (the upper redshift boundary adopted in the luminosity function computation). By combining this value with the [Ne v] median EW of the objects included in the luminosity function, we derived the line luminosity at which the magnitude cut would remove approximately half of the objects. This corresponds to L[OIII ] ,half = 1.15 × 108 L⊙ and shows that our luminosity function is largely incomplete in the lowest luminosity bin.

thumbnail Fig. 7

Space density of Compton-thick AGN samples with different intrinsic 2–10 keV luminosities compared with the XRB model predictions (Gilli et al. 2007). Datapoints have to be compared with the corresponding color curves; luminosity thresholds are as labeled. References to the observed densities, sorted by increasing luminosity threshold, and by increasing redshift for the same luminosity threshold, are as follows: L11: Luo et al. (2011); D07: Daddi et al. (2007b); V14: this work; F09: Fiore et al. (2009); A08: Alexander et al. (2008); V10: Vignali et al. (2010); P06: Polletta et al. (2006); B10: Bauer et al. (2010).

Finally, we combined the two datapoints in our luminosity function that are most complete (i.e., those at logL[NeV]> 40.9 or, equivalently, L[OIII]> 1.9 × 108 L⊙) to evaluate the space density of Compton-thick AGN in our sample. First we computed the space density of the 24 AGN in the range ΔlogL[NeV] = 40.9−41.3 at z = 0.65−1.01 (by using Eq. (1) and then multiplying back by the Δlog L term), and then we multiplied this density by the fraction of Compton-thick objects fThick as derived in Sect. 3.2 (i.e., fThick = 0.43). The resulting space density of Compton-thick AGN at z = 0.83 as derived from our sample is then ΦThick = (9.1 ± 2.1) × 10-6 Mpc-3. To compare this result with the predictions from AGN synthesis models of the XRB and literature results, we converted the [Ne v] luminosity threshold of logLNeV> 40.9 into an X-ray luminosity threshold. In particular, since we are interested in the intrinsic AGN power, we used the relation logL2−10 keV = 2.6 + log L[NeV] as derived by G10 for unobscured AGN, where L2−10 keV is the intrinsic, rest-frame 2–10 keV luminosity. The derived space density of Compton-thick AGN at z = 0.83 with logL2−10 keV> 43.5 is shown in Fig. 7 along with some measurements of the density of Compton-thick AGN in different luminosity and redshift ranges reported by other works and with the expectations from the XRB model of Gilli et al. (2007) (see also Alexander et al. 2011 for an additional estimate at z ≈ 2; and Della Ceca et al. 2008 and Severgnini et al. 2012 for measurements at z< 0.06). Our measurement is in good agreement with both the model expectations for logL2−10 keV> 43.5 and the space density measured by Fiore et al. (2009) for objects in a similar redshift and luminosity range selected through their mid-IR excess in their SED. Since, as discussed earlier, even a mild extinction in the NLR can suppress [Ne v] emission, we regard our selection technique for Compton-thick AGN as clean but not complete (see the extended discussion on this issue in M13). This is why we plot our measured space density with an upward arrow in Fig. 7 (see also Vignali et al. 2010 who discuss a similar issue on the space density of [O iii]-selected Type 2 AGN at z ≈ 0.3−0.8).

5. Conclusions

According to XRB synthesis models, a significant fraction of the XRB emission at 20 keV (the so-called “missing XRB”) is produced by Seyfert-like Compton-thick objects at z ≈ 0.5−1.0 and with 2–10 keV intrinsic luminosities <1044 erg s-1 (see Fig. 2 of Gilli 2013). To this goal, we have selected a sample of 69 Type 2 AGN drawn from the zCOSMOS-Bright Survey on the basis of the presence of the [Ne v]3426 Å emission line at z = 0.65−1.20, a reliable tracer of AGN accretion. In this paper we have presented the X-ray properties of these sources (through individual analyses and stacking), and estimated the fraction of Compton-thick AGN on the basis of the X/NeV ratio (see G10). Finally, we have computed the space density of Compton-thick AGN at z ≈ 0.8 and compared it with previous results and Gilli et al. (2007) XRB model expectations. The main results are as follows:

  • Twenty-three of the 69 [Ne v]-selectedType 2 AGN were detected in the Chandra-COSMOS observations. Their spectral properties suggest a widerange of absorption, going from limited obscuration for fewsources among those with most X-ray photons toheavy obscuration, although formally none of these objects hasX/NeV ratio falling in the Compton-thick AGN region.

  • Using X-ray stacking analysis for the individually X-ray undetected Type 2 AGN coupled with Monte Carlo simulations provides a fraction of ≈43% of Compton-thick AGN in the sample of 69 [Ne v]-selected Type 2 AGN.

  • The derived space density of Compton-thick AGN at z = 0.83 is ΦThick = (9.1 ± 2.1) × 10-6 Mpc-3. This value is in good agreement with XRB model expectations for logL2−10 keV> 43.5 AGN and with the previously measured space density for Compton-thick objects in a similar redshift and luminosity range (Fiore et al. 2009).

  • The [O iii] luminosity function, “converted” from the measured [Ne v] luminosity function, is a factor of ≈ 2 above that computed by B10 for Type 2 AGN selected in the zCOSMOS survey at similar redshifts using the [O iii]/Hβ vs. [O ii]/Hβ diagnostic diagram. This result indicates that any optical selection method based on spectroscopy and diagnostic diagrams is not complete in finding obscured AGN. Therefore, to search for the most heavily obscured objects, a “panchromatic” observing strategy is the way to go (e.g., Delvecchio et al., in prep.).

Acknowledgments

C.V. thanks M. Bolzonella for useful discussion. Financial contribution from “PRIN–INAF 2011” and “PRIN–INAF 2012” is acknowledged. K.I. thanks support from Spanish Ministerio de Cíencia e Innovacíon (MICINN) through the grant (AYA2010-21782-C03-01).

References

  1. Ajello, M., Alexander, D. M., Greiner, J., et al. 2012, ApJ, 749, 21 [NASA ADS] [CrossRef] [Google Scholar]
  2. Akylas, A., Georgakakis, A., Georgantopoulos, I., Brightman, M., & Nandra, K. 2012, A&A, 546, A98 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  3. Alexander, D. M., Smail, I., Bauer, F. E., et al. 2005, Nature, 434, 738 [NASA ADS] [CrossRef] [PubMed] [Google Scholar]
  4. Alexander, D. M., Brandt, W. N., Smail, I., et al. 2008, AJ, 135, 1968 [NASA ADS] [CrossRef] [Google Scholar]
  5. Alexander, D. M., Bauer, F. E., Brandt, W. N., et al. 2011, ApJ, 738, 44 [NASA ADS] [CrossRef] [Google Scholar]
  6. Alexander, D. M., Stern, D., Del Moro, A., et al. 2013, ApJ, 773, 125 [NASA ADS] [CrossRef] [Google Scholar]
  7. Anders, E., & Grevesse, N. 1989, Geochim. Cosmochim. Acta, 53, 197 [Google Scholar]
  8. Arnaud, K. A. 1996 in Astronomical Data Analysis Software and Systems V, ASP Conf. Ser., 101, 17 [Google Scholar]
  9. Avni, Y. 1976, ApJ, 210, 642 [NASA ADS] [CrossRef] [Google Scholar]
  10. Ballantyne, D. R. 2009, ApJ, 698, 1033 [NASA ADS] [CrossRef] [Google Scholar]
  11. Bauer, F. E., Yan, L., Sajina, A., & Alexander, D. M. 2010, ApJ, 710, 212 [NASA ADS] [CrossRef] [Google Scholar]
  12. Beckmann, V., Soldi, S., Ricci, C., et al. 2009, A&A, 505, 417 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  13. Bolzonella, M., Kovač, K., Pozzetti, L., et al. 2010, A&A, 524, A76 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  14. Bongiorno, A., Mignoli, M., Zamorani, G., et al. 2010, A&A, 510, A56 (B10) [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  15. Brightman, M., & Ueda, Y. 2012, MNRAS, 423, 702 [NASA ADS] [CrossRef] [Google Scholar]
  16. Brightman, M., Nandra, K., Salvato, M., et al. 2014, MNRAS, 443, 1999 [NASA ADS] [CrossRef] [Google Scholar]
  17. Broos, P. S., Townsley, L. K., Feigelson, E. D., et al. 2010 ApJ, 714, 1582 [NASA ADS] [CrossRef] [Google Scholar]
  18. Brusa, M., Civano, F., Comastri, A., et al. 2010, ApJ, 716, 348 [NASA ADS] [CrossRef] [Google Scholar]
  19. Brusa, M., Bongiorno, A., Cresci, G., et al. 2014, MNRAS, in press [arXiv:1409.1615] [Google Scholar]
  20. Buchner, J., Georgakakis, A., Nandra, K., et al. 2014, A&A, 564, A125 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  21. Burlon, D., Ajello, M., Greiner, J., et al. 2011, ApJ, 728, 58 [NASA ADS] [CrossRef] [Google Scholar]
  22. Cappelluti, N., Brusa, M., Hasinger, G., et al. 2009, A&A, 497, 635 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  23. Cash, W. 1979, ApJ, 228, 939 [NASA ADS] [CrossRef] [Google Scholar]
  24. Castelló-Mor, N., Carrera, F. J., Alonso-Herrero, A., et al. 2013, A&A, 556, A114 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  25. Cattaneo, A., Faber, S. M., Binney, J., et al. 2009, Nature, 460, 213 [NASA ADS] [CrossRef] [PubMed] [Google Scholar]
  26. Chartas, G., Brandt, W. N., Gallagher, S. C., & Garmire, G. P. 2002, ApJ, 579, 169 [NASA ADS] [CrossRef] [Google Scholar]
  27. Cicone, C., Maiolino, R., Sturm, E., et al. 2014, A&A, 562, A21 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  28. Civano, F., Elvis, M., Brusa, M., et al. 2012, ApJS, 201, 30 (C12) [NASA ADS] [CrossRef] [Google Scholar]
  29. Comastri, A. 2004, Supermassive Black Holes in the Distant Universe, in Astrophysics and Space Science Library, eds. A. J. Barger (Kluwer), 308, 245 [Google Scholar]
  30. Comastri, A., Ranalli, P., Iwasawa, K., et al. 2011, A&A, 526, L9 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  31. Daddi, E., Dickinson, M., Morrison, G., et al. 2007a, ApJ, 670, 156 [NASA ADS] [CrossRef] [Google Scholar]
  32. Daddi, E., Alexander, D. M., Dickinson, M., et al. 2007b, ApJ, 670, 173 [NASA ADS] [CrossRef] [Google Scholar]
  33. Della Ceca, R., Caccianiga, A., Severgnini, P., et al. 2008, A&A, 487, 119 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  34. Del Moro, A., Alexander, D. M., Mullaney, J. R., et al. 2013, A&A, 549, A59 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  35. Dey, A., Soifer, B. T., Desai, V., et al. 2008, ApJ, 677, 943 [NASA ADS] [CrossRef] [Google Scholar]
  36. Di Matteo, T., Springel, V., & Hernquist, L. 2005, Nature, 433, 604 [NASA ADS] [CrossRef] [PubMed] [Google Scholar]
  37. Donley, J. L., Koekemoer, A. M., Brusa, M., et al. 2012, ApJ, 748, 142 [NASA ADS] [CrossRef] [Google Scholar]
  38. Elvis, M., Civano, F., Vignali, C., et al. 2009, ApJS, 184, 158 [NASA ADS] [CrossRef] [Google Scholar]
  39. Ferrarese, L., & Merritt, D. 2000, ApJ, 539, L9 [NASA ADS] [CrossRef] [Google Scholar]
  40. Feruglio, C., Maiolino, R., Piconcelli, E., et al. 2010, A&A, 518, L155 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  41. Feruglio, C., Daddi, E., Fiore, F., et al. 2011, ApJ, 729, L4 [NASA ADS] [CrossRef] [Google Scholar]
  42. Fiore, F., Grazian, A., Santini, P., et al. 2008, ApJ, 672, 94 [NASA ADS] [CrossRef] [Google Scholar]
  43. Fiore, F., Puccetti, S., Brusa, M., et al. 2009, ApJ, 693, 447 [NASA ADS] [CrossRef] [Google Scholar]
  44. Gandhi, P., Horst, H., Smette, A., et al. 2009, A&A, 502, 457 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  45. Gebhardt, K., Bender, R., Bower, G., et al. 2000, ApJ, 539, L13 [NASA ADS] [CrossRef] [Google Scholar]
  46. Georgantopoulos, I., Akylas, A., Georgakakis, A., & Rowan-Robinson, M. 2009, A&A, 507, 747 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  47. Georgantopoulos, I., Comastri, A., Vignali, C., et al. 2013, A&A, 555, A43 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  48. Gilli, R. 2013, Mem. Soc. Astron. It., 84, 647 [NASA ADS] [Google Scholar]
  49. Gilli, R., Comastri, A., & Hasinger, G. 2007, A&A, 463, 79 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  50. Gilli, R., Vignali, C., Mignoli, M., et al. 2010, A&A, 519, A92 (G10) [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  51. Gilli, R., Su, J., Norman, C., et al. 2011, ApJ, 730, L28 [NASA ADS] [CrossRef] [Google Scholar]
  52. Gilli, R., Norman, C., Vignali, C., et al. 2014, A&A, 562, A67 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  53. Goulding, A. D., Alexander, D. M., Mullaney, J. R., et al. 2011, MNRAS, 411, 1231 [NASA ADS] [CrossRef] [Google Scholar]
  54. Harrison, C. M., Alexander, D. M., Swinbank, A. M., et al. 2012, MNRAS, 426, 1073 [NASA ADS] [CrossRef] [Google Scholar]
  55. Harrison, C. M., Alexander, D. M., Mullaney, J. R., & Swinbank, A. M. 2014, MNRAS, 441, 3306 [Google Scholar]
  56. Hickox, R. C., Jones, C., Forman, W. R., et al. 2009, ApJ, 696, 891 [Google Scholar]
  57. Hopkins, P. F., Hernquist, L., Cox, T. J., & Kereš, D. 2008, ApJS, 175, 356 [NASA ADS] [CrossRef] [Google Scholar]
  58. Houck, J. R., Soifer, B. T., Weedman, D., et al. 2005, ApJ, 622, L105 [NASA ADS] [CrossRef] [Google Scholar]
  59. Jia, J., Ptak, A., Heckman, T., & Zakamska, N. L. 2013, ApJ, 777, 27 [NASA ADS] [CrossRef] [Google Scholar]
  60. Kalberla, P. M. W., Burton, W. B., Hartmann, D., et al. 2005, A&A, 440, 775 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  61. Kormendy, J., & Richstone, D. 1995, ARA&A, 33, 581 [NASA ADS] [CrossRef] [Google Scholar]
  62. Iwasawa, K., Gilli, R., Vignali, C., et al. 2012a, A&A, 546, A84 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  63. Iwasawa, K., Mainieri, V., Brusa, M., et al. 2012b, A&A, 537, A86 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  64. Lacy, M., Storrie-Lombardi, L. J., Sajina, A., et al. 2004, ApJS, 154, 166 [NASA ADS] [CrossRef] [Google Scholar]
  65. Lacy, M., Ridgway, S. E., Gates, E. L., et al. 2013, ApJS, 208, 24 [NASA ADS] [CrossRef] [Google Scholar]
  66. Lamareille, F., Mouhcine, M., Contini, T., Lewis, I., & Maddox, S. 2004, MNRAS, 350, 396 [NASA ADS] [CrossRef] [Google Scholar]
  67. LaMassa, S. M., Heckman, T. M., Ptak, A., et al. 2009, ApJ, 705, 568 [NASA ADS] [CrossRef] [Google Scholar]
  68. Lamastra, A., Bianchi, S., Matt, G., et al. 2009, A&A, 504, 73 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  69. Lamastra, A., Menci, N., Fiore, F., et al. 2013, A&A, 559, A56 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  70. Lansbury, G. B., Alexander, D. M., Del Moro, A., et al. 2014, ApJ, 785, 17 [NASA ADS] [CrossRef] [Google Scholar]
  71. Lanzuisi, G., Piconcelli, E., Fiore, F., et al. 2009, A&A, 498, 67 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  72. Lilly, S.J., Le Févre, O., Renzini, A., et al. 2007, ApJS, 172, 70 [NASA ADS] [CrossRef] [Google Scholar]
  73. Lilly, S.J., Le Brun, V., Maier, C., et al. 2009, ApJS, 184, 218 [Google Scholar]
  74. Luo, B., Brandt, W. N., Xue, Y. Q., et al. 2011, ApJ, 740, 37 [NASA ADS] [CrossRef] [Google Scholar]
  75. Mainieri, V., Hasinger, G., Cappelluti, N., et al. 2007, ApJS, 172, 368 [NASA ADS] [CrossRef] [Google Scholar]
  76. Maiolino, R., Gallerani, S., Neri, R., et al. 2012, MNRAS, 425, L66 [NASA ADS] [CrossRef] [Google Scholar]
  77. Martínez-Sansigre, A., Rawlings, S., Lacy, M., et al. 2005, Nature, 436, 666 [NASA ADS] [CrossRef] [Google Scholar]
  78. Marshall, H. L., Tananbaum, H., Avni, Y., & Zamorani, G. 1983, ApJ, 269, 35 [NASA ADS] [CrossRef] [Google Scholar]
  79. Mateos, S., Alonso-Herrero, A., Carrera, F. J., et al. 2013, MNRAS, 434, 941 [NASA ADS] [CrossRef] [Google Scholar]
  80. Menci, N., Fiore, F., Puccetti, S., & Cavaliere, A. 2008, ApJ, 686, 219 [NASA ADS] [CrossRef] [Google Scholar]
  81. Merloni, A., & Heinz, S. 2013, Evolution of Active Galactic Nuclei, in Planets, Stars and Stellar Systems. Vol. 6: Extragalactic Astronomy and Cosmology, 503 (Springer) [Google Scholar]
  82. Mignoli, M., Vignali, C., Gilli, R., et al. 2013, A&A, 556, A29 (M13) [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  83. Moretti, A., Vattakunnel, S., Tozzi, P., et al. 2012, A&A, 548, A87 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  84. Nesvadba, N. P. H., Lehnert, M. D., De Breuck, C., Gilbert, A. M., & van Breugel, W. 2008, A&A, 491, 407 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  85. Piconcelli, E., Jimenez-Bailón, E., Guainazzi, M., et al. 2005, A&A, 432, 15 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  86. Polletta, M. d. C., Wilkes, B. J., Siana, B., et al. 2006, ApJ, 642, 673 [NASA ADS] [CrossRef] [Google Scholar]
  87. Ptak, A., Zakamska, N. L., Strauss, M. A., et al. 2006, ApJ, 637, 147 [NASA ADS] [CrossRef] [Google Scholar]
  88. Puccetti, S., Vignali, C., Cappelluti, N., et al. 2009, ApJS, 185, 586 [NASA ADS] [CrossRef] [Google Scholar]
  89. Schmidt, M. 1968, ApJ, 151, 393 [NASA ADS] [CrossRef] [Google Scholar]
  90. Scoville, N., Abraham, R.G., Aussel, H., et al. 2007, ApJS, 172, 38 [NASA ADS] [CrossRef] [Google Scholar]
  91. Severgnini, P., Caccianiga, A., & Della Ceca, R. 2012, A&A, 542, A46 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  92. Shi, Y., Helou, G., & Armus, L. 2013, ApJ, 777, 6 [NASA ADS] [CrossRef] [Google Scholar]
  93. Silk, J., & Rees, M. J. 1998, A&A, 331, L1 [NASA ADS] [Google Scholar]
  94. Stern, D., Eisenhardt, P., Gorjian, V., et al. 2005, ApJ, 631, 163 [NASA ADS] [CrossRef] [Google Scholar]
  95. Tombesi, F., Cappi, M., Reeves, J. N., & Braito, V. 2012, MNRAS, 422, L1 [NASA ADS] [CrossRef] [Google Scholar]
  96. Tozzi, P., Gilli, R., Mainieri, V., et al. 2006, A&A, 451, 457 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  97. Treister, E., Urry, C. M., & Virani, S. 2009, ApJ, 696, 110 [NASA ADS] [CrossRef] [Google Scholar]
  98. Treister, E., Natarajan, P., Sanders, D. B., et al. 2010, Science, 328, 600 [NASA ADS] [CrossRef] [Google Scholar]
  99. Trouille, L., & Barger, A. J. 2010, ApJ, 722, 212 [NASA ADS] [CrossRef] [Google Scholar]
  100. Tueller, J., Mushotzky, R. F., Barthelmy, S., et al. 2008, ApJ, 681, 113 [NASA ADS] [CrossRef] [Google Scholar]
  101. Vasudevan, R. V., Brandt, W. N., Mushotzky, R. F., et al. 2013, ApJ, 763, 111 [NASA ADS] [CrossRef] [Google Scholar]
  102. Vignali, C. 2014, in Multiwavelength AGN Surveys and Studies, eds. A. Mickaelian & D. Sanders, IAU Sympos., 304, 132 [Google Scholar]
  103. Vignali, C., Alexander, D. M., & Comastri, A. 2006, MNRAS, 373, 321 [NASA ADS] [CrossRef] [Google Scholar]
  104. Vignali, C., Alexander, D. M., Gilli, R., & Pozzi, F. 2010, MNRAS, 404, 48 [NASA ADS] [Google Scholar]
  105. Villforth, C., Hamann, F., Rosario, D. J., et al. 2014, MNRAS, 439, 3342 [NASA ADS] [CrossRef] [Google Scholar]
  106. Vito, F., Vignali, C., Gilli, R., et al. 2013, MNRAS, 428, 354 [NASA ADS] [CrossRef] [Google Scholar]
  107. Weedman, D. W., Soifer, B. T., Hao, L., et al. 2006, ApJ, 651, 101 [NASA ADS] [CrossRef] [Google Scholar]
  108. Zubovas, K., & King, A. R. 2012, MNRAS, 426, 2751 [NASA ADS] [CrossRef] [Google Scholar]
  109. Zucca, E., Bardelli, S., Bolzonella, M., et al. 2009, A&A, 508, 1217 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]

All Tables

Table 1

List of X-ray detected [NeV]-selected Type 2 AGN.

In the text
Table 2

List of X-ray undetected [NeV]-selected Type 2 AGN.

In the text

All Figures

thumbnail Fig. 1

Distribution of the rest-frame 2–10 keV (with no correction for absorption) to [NeV] luminosity ratio (X/NeV) of the 69 zCOSMOS Type 2 AGN with reliable X-ray photometry. X-ray detections are reported as either red (filled) or blue (dotted) histograms; in particular, the red histogram refers to X-ray photometry obtained from the Acis Extract analysis carried out in the rest-frame 2–10 keV band, while the blue histogram reports the X-ray flux values derived from direct X-ray spectral fitting in the same energy band (see text for details). The black histogram refers to the X-ray undetected [Ne v]-selected Type 2 AGN. The vertical dashed line shows the threshold defined by G10 for Compton-thick (leftward direction) and Compton-thin (rightward direction) sources. All of the X-ray detected sources are likely Compton thin using the X/NeV threshold of 15 adopted by G10. The filled symbols represent the average X/NeV value obtained from the X-ray stacking analysis (black circle: stack of the 46 X-ray undetected Type 2 AGN; green triangle: stack of the 22 sources with “nominal” upper limit on the X/NeV ratio <30; magenta square: stack of the 24 sources with “nominal” upper limit on the X/NeV ratio >30; see Sect. 3.2 for details). Their y-axis position is arbitrary.
In the text
thumbnail Fig. 2

Rest-frame X-ray spectra of the four Type 2 AGN with most counts (i.e., above 140 counts; CID339 a), CID110 b), CID522 c), and CID456 d)), characterized by increasing absorption from top to bottom, and stacked X-ray spectrum of the 19 Type 2 AGN with limited photon statistics (panel e)), showing an emission line at an energy consistent with iron Kα. An emission line is also apparent in CID339. X-ray spectra are arbitrarily scaled for visual clarity.
In the text
thumbnail Fig. 3

Observed-frame X-ray spectrum of CID1019 (zCOSMOS_ID = 813250) at z = 0.7302. The flat X-ray continuum and the tentative iron Kα emission line at rest-frame energy of ≈6.4 keV are indicative of strong, likely Compton-thick absorption.
In the text
thumbnail Fig. 4

Rest-frame 1–10 keV stacked spectrum of the 46 Type 2 AGN with no individual X-ray detection. The limited-quality spectrum is flat, suggestive of absorption, but apparently no significant iron Kα emission line is present (EW< 1.5 keV).
In the text
thumbnail Fig. 5

Distribution of the 69 [Ne v]-selected Type 2 AGN in the [Ne v]l˜ine luminosity vs. redshift plane. The 23 objects detected in the X-rays are marked with a cross. The black dashed lines show the boundaries of the three bins used in the computation of the [Ne v] luminosity function. A total of 50 AGN fall in these three luminosity bins. The blue dashed curve shows the [Ne v] flux limit (in units oferg cm-2 s-1) of the zCOSMOS survey.
In the text
thumbnail Fig. 6

[O iii]-converted luminosity function of [Ne v]-selected Type 2 AGN at z = 0.65−1.01 in zCOSMOS (light blue triangles; this work) compared with the [O iii] luminosity function of zCOSMOS Type 2 AGN in a similar redshift range selected through the “blue” line diagnostic diagram (black circles; B10). Open symbols highlight those datapoints affected by significant incompleteness (see text and B10). Small numbers indicate how many objects were used to derive each datapoint.
In the text
thumbnail Fig. 7

Space density of Compton-thick AGN samples with different intrinsic 2–10 keV luminosities compared with the XRB model predictions (Gilli et al. 2007). Datapoints have to be compared with the corresponding color curves; luminosity thresholds are as labeled. References to the observed densities, sorted by increasing luminosity threshold, and by increasing redshift for the same luminosity threshold, are as follows: L11: Luo et al. (2011); D07: Daddi et al. (2007b); V14: this work; F09: Fiore et al. (2009); A08: Alexander et al. (2008); V10: Vignali et al. (2010); P06: Polletta et al. (2006); B10: Bauer et al. (2010).
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.