A publishing partnership

The following article is Free article

A NEW SAMPLE OF OBSCURED AGNs SELECTED FROM THE XMM-NEWTON AND AKARI SURVEYS

, , , , , , and

Published 2015 November 12 © 2015. The American Astronomical Society. All rights reserved.
, , Citation Yuichi Terashima et al 2015 ApJ 814 11DOI 10.1088/0004-637X/814/1/11

0004-637X/814/1/11

ABSTRACT

We report a new sample of obscured active galactic nuclei (AGNs) selected from the XMM-Newton serendipitous source and AKARI point-source catalogs. We match X-ray sources with infrared (18 and 90 μm) sources located at to create a sample consisting of 173 objects. Their optical classifications and absorption column densities measured by X-ray spectra are compiled and study efficient selection criteria to find obscured AGNs. We apply the criteria (1) X-ray hardness ratio defined by using the 2–4.5 keV and 4.5–12 keV bands > −0.1 and (2) EPIC-PN count rate (CR) in the 0.2–12 keV to infrared flux ratio CR/ or CR/ where F18 and F90 are infrared fluxes at 18 and 90 μm in Jy, respectively, to search for obscured AGNs. X-ray spectra of 48 candidates, for which no X-ray results have been published, are analyzed and X-ray evidence for the presence of obscured AGNs such as a convex shape X-ray spectrum indicative of absorption of ∼ 1022–24 cm−2, a very flat continuum, or a strong Fe–K emission line with an equivalent width of is found in 26 objects. Six of them are classified as Compton-thick AGNs, and four are represented by either Compton-thin or Compton-thick spectral models. The success rate of finding obscured AGNs combining our analysis and the literature is 92% if the 18 μm condition is used. Of the 26 objects, 4 are optically classified as an H ii nucleus and are new "elusive AGNs" in which star formation activity likely overwhelms AGN emission in the optical and infrared bands.

Export citation and abstractBibTeXRIS

1. INTRODUCTION

Multiwavelength observations have been finding various populations of active galactic nuclei (AGNs). The population of obscured AGNs, which constitute a large fraction of AGNs (e.g., Fabian 2004; Gilli et al. 2007), among various classes, are believed to be important in various aspects including the origin of the Cosmic X-ray background (XRB; e.g., Gilli et al. 2007), the connection between obscuring matter and star formation activity in the host galaxies (e.g., Wada & Norman 2002), and the evolutional paths of AGNs (e.g., Sanders et al. 1988; Hopkins et al. 2006; Alexander & Hickox 2012). Modern large area surveys at various wavebands are indeed utilized to find a large number of hidden AGNs and to elucidate their nature.

Obscured AGNs are found by X-ray emission transmitted through obscuring matter, optical line emission from extended and ionized regions, infrared emission from dust heated by central AGNs, and so on. Hard X-ray surveys are one effective way to find obscured AGNs because the photoelectric cross section decreases as X-ray energy increases, and transmitted X-rays can be observed. Indeed, more than a dozen Compton-thick AGNs, which are absorbed by a hydrogen column density greater than cm−2, show transmitted hard X-rays above 10 keV (Comastri et al. 2004 and references therein; Burlon et al. 2011). The all-sky and sensitive hard X-ray surveys conducted by the Swift Burst Alert Telescope (BAT) and INTEGRAL provide an unprecedented opportunity to search for the heavily absorbed population. While obscured AGNs have been found in these surveys, the number of Compton-thick sources are not as many as expected from other studies using, e.g., optical emission-line-selected samples (Gilli et al. 2007) or prediction from the synthesis models of the XRB (Burlon et al. 2011). This bias is due to the attenuation of X-rays even at energies above 10 keV for Compton-thick cases (Wilman & Fabian 1999; Ikeda et al. 2009; Murphy & Yaqoob 2009; Brightman & Nandra 2011a). Burlon et al. (2011) estimated a correction factor of about four for Compton-thick fraction in their Swift/BAT sample. Thus, hard X-rays are still biased against heavily absorbed AGNs.

Infrared emission from warm dust heated by the central source is also employed to find obscured AGNs. Mid-infrared (MIR) emission indeed traces the power of the AGNs regardless of whether they are type 1 or 2, and the attenuation of MIR in type 2 AGNs is not large (Horst et al. 2008; Gandhi et al. 2009; Ichikawa et al. 2012; Matsuta et al. 2012). A weakness of utilizing infrared emission, however, is that emission from dust heated by stars cannot be separated if spatial resolution is not sufficiently high, and therefore the infrared-selected sample contains a non-negligible fraction of non-AGN galaxies. For example, 13 and 18 out of 126 galaxies selected at 12 μm are optically classified as H ii and H ii/AGN composite galaxies, respectively (Brightman & Nandra 2011b).

The combination of X-ray and infrared selection is a useful way to select a heavily obscured AGN population and is applied to deep (Fiore et al. 2008, 2009) or wide surveys (Mateos et al. 2012; Severgnini et al. 2012; Rovilos et al. 2014) to overcome the biases in the selections using only hard X-ray or infrared emission. The X-ray to infrared flux ratios, X-ray hardness, infrared colors, and so on are utilized in these selections. Among the techniques employed, we extend the method used by Severgnini et al. (2012). They used 25 μm fluxes (F25) measured in the IRAS Point Source Catalog (PSC) and X-ray data taken from the XMM-Newton serendipitous source catalog (2XMM catalog; Watson et al. 2009) and made a diagnostic plot of X-ray hardness ratio (HR4) and X-ray to infrared flux ratio (F(2−12 keV)/), where HR4 is defined by using X-ray count rates (CRs) in 2–4.5 keV CR(2–4.5 keV) and in 4.5–12 keV CR(4.5–12 keV) as

They defined the region for candidates of Compton-thick AGNs as and HR4 > −0.2, and built a sample consisting of 43 candidates. For absorbed sources, the X-ray to infrared ratio becomes small since X-rays below 12 keV are attenuated by photoelectric absorption. Absorbed sources show flatter X-ray spectra and therefore larger values of hardness ratios are expected. Thus, their criteria are expected to work to select heavily absorbed sources. About 84% of the objects in their sample are confirmed as Compton-thick AGNs and 20% are newly discovered ones. Thus, the combination of wide field survey data in the infrared and X-ray bands is promising in the search for heavily obscured AGNs.

In this paper, we combine the infrared all-sky survey data obtained with AKARI (Murakami et al. 2007) and the 2XMM catalog. We construct diagnostic diagrams to classify activity in galaxies and to search for obscured AGNs. We selected 48 candidates for obscured AGNs and analyzed their X-ray spectra. This paper is organized as follows. Section 2 describes the selection method of X-ray and infrared sources. Diagnostic diagrams to classify the selected sources are presented in Section 3. Results of X-ray spectral analysis are shown in Section 4. Section 5 discusses the results and summaries are given in Section 6. We adopt H0 = 70 km s−1, = 0.3, and = 0.7 throughout this paper.

2. THE SAMPLE

2.1. XMM-Newton and AKARI Catalogs

We combine two large area survey data in the X-ray and infrared bands. We used the XMM-Newton Serendipitous source catalog Data Release 3 (2XMMi-DR3), which contains 262,902 unique X-ray sources. The median X-ray flux in 0.2–12 keV is erg s−1 cm−2. The typical positional uncertainty is 1farcs5 (1σ) (Watson et al. 2009). The entries listed in this catalog (CRs in 0.2–12 keV and HR4) are used to create diagnostic diagrams and select candidates for obscured AGNs. The data from EPIC-PN, which has a larger effective area than those of EPIC-MOS, was used throughout the analysis. We use sources located at the Galactic latitude in 2902 observations with usable PN data. Among these observations, 2686 and 216 observations are taken with full window and large window modes, respectively. There are 2062 and 139 unique fields taken with the full window and large window modes, respectively. The total number of unique EPIC-PN sources at is 150,799. In the following analysis, sources with EPIC-PN counts in 0.2–12 keV greater than 60 counts (60,851 unique sources) are used.

AKARI PSCs were used as infrared data. AKARI surveyed most of the sky with the two instruments, the Infrared Camera (IRC; Onaka et al. 2007; Ishihara et al. 2010) and the Far-Infrared Surveyor (FIS; Kawada et al. 2007). The bandpasses of the IRC are centered at 9 and 18 μm, while the band centers of the FIS are 65, 90, 140, and 160 μm. We used 18 and 90 μm measurements as mid- and far-infrared data. The 18 μm is chosen to avoid silicate features, which may affect continuum measurements, in the 9 μm bandpass of the IRC. The 90 μm band is used among the fir-infrared bands because this band is most sensitive. We use only data with the quality flag of FQUAL = 3, which means flux measurements are most reliable (Yamamura et al. 2010). There are 43,865 (18 μm) and 62,326 (90 μm) sources with FQUAL = 3 located at

2.2. Cross Correlation of the X-Ray and Infrared Catalogs

We compared the positions of X-ray sources in the 2XMM-DR3 and infrared sources in the AKARI PSCs and made a list of X-ray and infrared sources. We first listed the FIS sources within 20'' of the X-ray source positions. If there are multiple sources within the circular region around an X-ray source, the source closest to the X-ray position is assumed to be a counterpart and is used in the following analysis. If one source is observed more than twice with XMM-Newton, only the data with the largest number of counts in 0.2–12 keV is used. We then matched the IRC and X-ray sources in a similar way using a matching radius of 10''. Finally, the FIS and IRC source lists are combined. The combined list contains 253 sources. The histograms of the separation between the X-ray and infrared positions are shown in Figure 1 for the 18 and 90 μm sources. The separations for the 18 μm sources are more concentrated within a small radius from the X-ray positions compared to the 90 μm sources as expected from the positional accuracy of the IRC and FIS sources ( 6'' and 18'', respectively, for faint sources).

Figure 1. Refer to the following caption and surrounding text.

Figure 1. Distribution of the separation between the X-ray and infrared positions. Solid line: 18 μm sources. Dotted–dashed line: 90 μm sources.

Standard image High-resolution image

We searched for the most probable counterpart of the X-ray sources using the NASA/IPAC Extragalactic Database (NED) and SIMBAD Astronomical Database. Stars, H ii regions, planetary nebulae, young stellar objects, and ultraluminous X-ray sources are excluded from the sample, resulting in 173 objects. The sample thus obtained is summarized in Table 1, where the XMM source name, alternative name, redshift, infrared flux densities, X-ray CRs in 0.2−12 keV, and hardness ratios (HR4) are shown. Redshifts are available for 171 of the 173 sources. Their distribution is shown in Figure 2 as a histogram. The distribution of X-ray CRs in 0.2–12 keV for all the EPIC-PN sources located at with 0.2–12 keV counts greater than 60 and the 171 objects in our sample is shown in Figure 3. Comparison of these histograms indicates that X-ray brighter objects tend to have a possible infrared counterpart detected in both the 18 and 90 μm bands. None of X-ray sources fainter than 0.004 counts s−1 in 0.2–12 keV are matched with infrared sources. We also compared distributions of infrared flux densities. The solid histograms in Figure 4 are distributions of infrared flux densities for 90 μm sources having 18 μm source(s) within 20'' and their nearest neighbor 18 μm sources (3535 objects in total) located at with FQUAL = 3. The dotted–dashed histograms are those for the 173 objects matched with EPIC-PN sources. A clear difference before and after matching with X-ray sources is shown as the double peak structure in the histogram for the 18 μm sources. The peak at a higher infrared flux density is likely to be composed of Galactic sources because the peak is more enhanced if we use objects at low Galactic longitude (e.g., ) and because the peak disappeared after excluding Galactic sources in our selection procedure. The distribution of 90 μm fluxes for X-ray-matched objects are flatter than that for all the 90 μm sources with nearby 18 μm sources. This fact indicates that objects brighter at 90 μm tend to have a greater probability of being matched with an X-ray source.

Figure 2. Refer to the following caption and surrounding text.

Figure 2. Distribution of redshifts for the matched sample using X-ray, 18, and 90 μm data shown in Table 1.

Standard image High-resolution image
Figure 3. Refer to the following caption and surrounding text.

Figure 3. Distribution of X-ray count rates in 0.2–12 keV. Solid line: all the EPIC-PN sources located at with 0.2–12 keV counts greater than 60. Dotted–dashed line: 173 objects in our sample after matching with the 18 and 90 μm sources and excluding the Galactic and ultraluminous X-ray sources.

Standard image High-resolution image
Figure 4. Refer to the following caption and surrounding text.

Figure 4. (Left) Distribution of 18 μm flux densities. (Right) Distribution of 90 μm flux densities. Solid line: nearest neighbor 18 μm sources to 90 μm sources within 20'' located at Dotted–dashed line: 173 objects in our sample after matching with the X-ray sources and excluding the Galactic and ultraluminous X-ray sources.

Standard image High-resolution image

Table 1.  XMM + AKARI 18 and 90 μm Sample

XMM Name Other Name Redshift CR HR4 F18 F90 CR/F18 CR/F90 Class References X-ray References
      (s−1)   (Jy) (Jy)         (1022 cm−2) Class  
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
2XMM J004847.1+315724 NGC 262/Mrk 348 0.0150 2.389 0.402 0.593 0.736 4.031 3.2486 Sy2 1 Thin 1
2XMMi J005329.8-084604 NGC 291 0.0190 0.034 0.522 0.204 3.135 0.166 0.0108 Sy2 2 >100a Thick 2
2XMM J005334.9+124136 Mrk 1502 0.0589 8.677 −0.512 0.714 1.964 12.144 4.4174 Sy1 1 b Un 1
2XMMi J005356.2-703804 IRAS 00521-7054 0.0689 0.281 0.013 0.595 0.635 0.472 0.4431 Sy2 1 7.2 ± 0.3 Thin 3
2XMM J010516.8-582615 ESO 113-G010 0.0257 4.071 −0.416 0.207 1.460 19.663 2.7877 Sy1.8 3 <0.0072 Un 4
2XMM J011127.5-380500 NGC 424 0.0118 0.411 0.463 1.293 1.181 0.318 0.3481 Sy1.8 4 Thick 5
2XMM J011607.1+330522 NGC 449/Mrk 1 0.0159 0.097 −0.101 0.531 1.395 0.183 0.0698 Sy2 1 >100a Thick 6
2XMM J012732.4+191044 Mrk 359 0.0174 5.229 −0.342 0.187 1.282 27.917 4.0794 Sy1 1 b Un 7
2XMM J013331.1+354006 NGC 591/Mrk 1157 0.0152 0.108 0.370 0.226 2.024 0.479 0.0536 Sy2 5 >100a Thick 8
2XMMi J013805.3-125210 IRAS 01356-1307 0.0402 0.068 0.064 0.291 0.860 0.234 0.0790 Sy2 1

Notes. Column 1: XMM name; Column 2: Alternative name; Column 3: Redshift; Column 4: EPIC-PN count rate in the 0.2–12 keV band; Column 5: Hardness ratio HR4; Column 6: Flux density at 18 μm in units of Jy; Column 7: Flux density at 90 μm in units of Jy; Column 8: EPIC-PN count rate in 0.2–12 keV to 18 μm flux density ratio; Column 9: EPIC-PN count rate in 0.2–12 keV to 90 μm flux density ratio; Column 10: Optical classification. Sy—Seyfert, L—LINER, T—transition object, H ii—H ii nucleus, BL Lac—BL Lac object, Unclass—unclassified, 1—type 1, 2—type 2, and a fractional number denotes intermediate types; Column 11: Reference for optical classification; Column 12: Absorption column density in units of 1022 cm−2; Column 13: Classification of absorption measured with X-ray. Un—unabsorbed or slightly absorbed ( cm−2), Thin—Compton-thin ( cm−2), Thick—Compton-thick ( cm−2). If only the lower limit of cm−2 is obtained, such objects are regarded as Compton-thick; Column 14: Reference for absorption column density.

aCompton-thick candidates judged from a large Fe–K line EW (>700 eV) and/or flat continuum. bX-ray spectrum shows no evidence for significant absorption. cThere are two IRC sources corresponding to NGC 3690E and NGC 3690W. The nucleus of NGC 3690W is likely to be an obscured AGN (Zezas et al. 2003), and the 18 μm flux of NGC 3690W is shown. FIS does not resolve the two sources and the 90 μm flux is for the single FIS source. dCompton-thin candidates in Severgnini et al. (2012).

References. Optical classification: (1) de Grijp et al. (1992), (2) Kewley et al. (2001), (3) Pietsch et al. (1998), (4) Murayama et al. (1998), (5) Veilleux & Osterbrock (1987), (6) Veilleux et al. (1999), (7) Ho et al. (1997), (8) Veilleux et al. (1995), (9) Gonçalves et al. (1999), (10) Shuder & Osterbrock (1981), (11) Schulz et al. (1994), (12) Yuan et al. (2010), (13) Adams & Weedman (1975), (14) Wu et al. (1998), (15) Shuder (1980), (16) Prieto et al. (1986), (17) Maia et al. (2003), (18) Masetti et al. (2006), (19) Miller et al. (1992), (20) Landi et al. (2007), (21) Boller et al. (1993), (22) Gallo et al. (2006), (23) Pastoriza (1979), (24) Masetti et al. (2008), (25) Buchanan et al. (2006), (26) Moran et al. (1996), (27) Phillips et al. (1979), (28) Véron-Cetty & Véron (1986a), (29) Véron-Cetty & Véron (1986b), (30) Morris & Ward (1988), (31) Huchra et al. (1982), (32) Brightman & Nandra (2011b), (33) Coziol et al. (2000), (34) Burlon et al. (2011), (35) Martínez et al. (2010), (36) Djorgovski et al. (1990), (37) 6dF; (38) Pereira-Santaella. et al. (2011), (39) Sargent (1972), (40) Keel et al. (1985), (41) Storchi-Bergmann et al. (1995), (42) Baan et al. (1998), (43) García-Marín et al. (2006) (44) Corbett et al. (2003), (45) Izotov et al. (1996). X-ray: (1) Brightman & Nandra (2011a), (2) LaMassa et al. (2009), (3) Ricci et al. (2014a), (4) Gallo et al. (2006), (5) Burlon et al. (2011), (6) Singh et al. (2011), (7) Fukazawa et al. (2011)l (8) Guainazzi et al. (2005a), (9) Teng & Veilleux (2010), (10) Winter et al. (2009), (11) Severgnini et al. (2012), (12) Churazov et al. (2003), (13) Pfefferkorn et al. (2001), (14) Evans et al. (2007), (15) Awaki et al. (2008), (16) Awaki et al. (2009), (17) Panessa et al. (2008), (18) Saez et al. (2012), (19) Maiolino et al. (2003), (20) Tilak et al. (2008), (21) Greenhill et al. (2008), (22) Malizia et al. (2007), (23) LaMassa et al. (2014), (24) Pounds et al. (2004), (25) González-Martín et al. (2009), (26) Terashima et al. (2002), (27) Fabian et al. (2013), (28) Noguchi et al. (2009), (29) Rodriguez et al. (2010), (30) Guainazzi et al. (2004), (31) De Rosa et al. (2008), (32) Hiroi et al. (2013), PKS 1549-79 (33) Ricci et al. (2010), (34) Piconcelli et al. (2007), (35) Singh et al. (2011), (36) Pereira-Santaella. et al. (2011), (37) Trinchieri et al. (2008) (38) Jiménez-Bailón et al. (2007), (39) Franceschini et al. (2003), (40) Panessa et al. (2005), (41) Della Ceca et al. (2002), (42) Seta et al. (2009).

Only a portion of this table is shown here to demonstrate its form and content. A machine-readable version of the full table is available.

Download table as:  DataTypeset image

We summarize the X-ray and infrared luminosities for the X-ray-IR matched sample in Table 2. The distribution of the luminosities are shown in Figures 5 and 6 as shaded histograms. Distances to nearby objects with a redshift parameter smaller than 0.003 are taken from the literature shown in Table 2, except for Mrk 59 for which a redshift-independent distance is not available. Distances to objects at z > 0.003 and Mrk 59 are calculated from source redshifts and the assumed cosmology. Observed X-ray luminosities in the 2–10 keV band (source rest frame) are obtained from the literature or our analysis presented in Section 4 if X-ray spectra clearly show the presence of an AGN. Most X-ray fluxes or luminosities are taken from the literature, like for absorption column densities. References are given in Table 2 only if a reference different from that for is used. When we use published fluxes in 2–10 keV to calculate luminosities, we made a simple K-correction by assuming a simple power law with a photon index of 1.8. Since most of the objects in our samples are at a low redshift , the assumption on the spectral shape only slightly affects the correction. If a photon index of 1.4 is assumed instead, the luminosity would be lower by 10% at z = 0.3. This amount is much smaller for lower redshift sources. In some cases, we obtained observed luminosities from the literature and converted them to our assumed cosmology. Infrared luminosities ( for 18 and/or 90 μm) are calculated from the AKARI measurement of infrared fluxes. We applied a K-correction by assuming a template spectral energy distribution for Seyfert 2s by Poletta et al. (2007). Again, because of low redshifts for our sample, the amount of the correction is relatively small. A correction factor is at most about 10% for a 18 μm luminosity at z = 0.3.

Figure 5. Refer to the following caption and surrounding text.

Figure 5. Distribution of observed X-ray luminosities in the 2–10 keV band for objects showing AGN activity for the matched sample using X-ray, 18, and 90 μm data shown in Table 1 (shaded histogram), the 90 μm sample (hatched histogram), and the 18 μm (open histogram) sample.

Standard image High-resolution image
Figure 6. Refer to the following caption and surrounding text.

Figure 6. Distribution of infrared luminosities at 18 μm (left) and 90 μm (right) for the matched sample using X-ray, 18, and 90 μm data shown in Table 1 (shaded histogram), the 90 μm sample (hatched histogram), and the 18 μm (open histogram) samples.

Standard image High-resolution image

Table 2.  X-Ray and Infrared Luminosities

Name References Note References
      (18 μm) (90 μm)    
(1) (2) (3) (4) (5) (6) (7)
NGC 262/Mrk 348 43.149 43.702 43.092
NGC 291 41.007 43.447 43.929
Mrk 1502 43.599 45.009 44.733
IRAS 00521-7054 43.408 45.074 44.384
ESO 113-G010 42.580 43.722 43.864
NGC 424 41.398 43.829 43.087
NGC 449/Mrk 1 40.659 43.705 43.421
Mrk 359 40.542 43.332 43.464
NGC 591/Mrk 1157 41.013 43.294 43.543
IRAS 01356-1307 44.271 44.031

Notes. Column 1: Name; Column 2: Logarithm of observed X-ray luminosity in the 2–10 keV band; Column 3: Reference for X-ray luminosity; Column 4: Logarithm of at 18 μm; Column 5: Logarithm of at 90 μm; Column 6: Distance of nearby objects used to calculate luminosities. Column 7: Reference for distance.

aX-ray spectra are analyzed in this paper, and no clear evidence for the presence of an obscured AGN is found.

References. X-ray: (1) De Rosa et al. (2012), (2) Brightman & Nandra (2011a), (3) Rivers et al. (2014), (4) Patrick et al. (2012), (5) Corral et al. (2011), (6) Ueda et al. (2005), (7) Malizia et al. (2007), (8) Levenson et al. (2006), (9) Winter et al. (2012), (10) Matt et al. (2004), (11) Ponti et al. (2010), (12) Teng & Veilleux (2010) (13) This work, (14) Harrison et al. (2003), (15) Akylas & Georgantopoulos (2009); (16) Shu et al. (2011), (17) Markowitz et al. (2007). Distance: (1) Jensen et al. (2003), (2) Yoshii et al. (2014) (3) Ekholm et al. (2000), (4) Bose & Kumar (2014), (5) Springob et al. (2009), (6) Cortes et al. (2008), (7) Wu et al. (2014), (8) Terry et al. (2002), (9) Dalcanton et al. (2009), (10) Parodi et al. (2000), (11) Jacobs et al. (2009), (12) McQuinn et al. (2010), (13) Takats & Vinko (2012), (14) Tully et al. (2009), (15) Benedict et al. (2009), (16) Mouhcine et al. (2005).

Only a portion of this table is shown here to demonstrate its form and content. A machine-readable version of the full table is available.

Download table as:  DataTypeset image

We compiled optical classifications from the literature or spectra in the archives. The sources are classified into Seyfert, Low-Ionization Nuclear Emission line Region (LINER), H ii nucleus, transition between LINER and H ii, BL Lac object, or normal galaxy. Seyfert, LINER, H ii, and, transition objects are defined based on the location of the optical emission line ratios ([N ii]λ6584/Hα, [S ii]λλ6716, 6730/Hα, and [O iii]λ5007/Hβ) on the excitation diagrams. Among various definitions of the boundary among the classes (Baldwin et al. 1981; Veilleux & Osterbrock 1987; Ho et al. 1997; Kewley et al. 2006), we adopted criteria of Ho et al. (1997) because many objects in our sample are contained in the Ho et al. sample in which stellar absorption lines were carefully treated in measuring emission line fluxes. Some objects are in the boundary region of two activity classes on the excitation diagrams, or use of different emission lines results in different classifications. For such ambiguous cases, both activity classes are shown such as Seyfert/LINER. The types (1, 1.2, 1.5, 1.8, 1.9, and 2) of Seyferts, LINERs, and transition objects are also shown if available in the literature. The classifications for some objects are not published in the literature, and we classified their optical spectra from the archives of the Sloan Digital Sky Survey, the 6dF Galaxy survey, or the Updated Zwicky Catalog (Falco et al. 1999), if available. If optical spectra show only absorption lines, they are classified as a normal galaxy. No classifications are available for some of the objects in the sample. They are denoted as "unclassified." These classifications are shown in Table 1 and summarized as histograms in Figure 7. In the histograms, Seyfert 1, 1.2, and 1.5 are treated as "Seyfert 1," while Seyfert 1.8, 1.9, and 2 are regarded as "Seyfert 2."

Figure 7. Refer to the following caption and surrounding text.

Figure 7. Summary of optical classifications for the matched sample using X-ray, 18, and 90 μm data shown in Table 1. Sy1: Seyfert 1, 1.2, and 1.5; Sy2: Seyfert 1.8, 1.9, and 2; L: LINER; T: transition object between LINER and H ii nucleus; H ii: H ii nucleus; Normal: normal galaxy; BL: BL Lac object; Un: Unclassified.

Standard image High-resolution image

Since one of our aims is to search for obscured AGNs, we compiled absorption column densities () measured using X-ray spectra from the literature, as shown in Table 1. values are shown for Seyferts and LINERs in which X-ray emission is dominated by AGNs. Some galactic nuclei classified as H ii show evidence for the presence of an AGN. Their values are also shown. If there are multiple published results, we put priority on the results of systematic analysis of a large sample, and results based on better quality of data. For objects showing a signature of heavy absorption exceeding ∼1024 cm−2, we use results based on wide-band spectra covering hard X-rays above 10 keV whenever possible. Objects showing a strong Fe–K emission line with an equivalent width (EW) exceeding 700 eV and/or a very flat spectral slope are regarded as > 1024 cm−2 even if only X-ray spectra below 10 keV are available. The boundary of the EW (700 eV) was chosen based on analysis of X-ray spectra for a large sample of AGNs (Guainazzi et al. 2005a; Fukazawa et al. 2011) and theoretical predictions (Awaki et al. 1991; Leahy & Creighton 1993; Ghisellini et al. 1994; Ikeda et al. 2009; Murphy & Yaqoob 2009; Brightman & Nandra 2011a). If an value is not explicitly presented in the literature and if X-ray spectral shape does not show a clear signature of absorption, is regarded as small as noted in Table 1.

3. DIAGNOSTIC DIAGRAMS

3.1. Hardness and X-Ray/Infrared Ratio

We first made a diagnostic diagram using the hardness ratio HR4 and the ratio between the X-ray CR in 0.2–12 keV and the infrared (18 or 90 μm) flux density. These diagrams are essentially the same as that used by Severgnini et al. (2012). We used X-ray CRs instead of X-ray fluxes because X-ray CRs are values directly derived from observational data without any assumptions on the X-ray spectral shape. Our diagrams are shown in the left and right panels of Figure 8 for 18 and 90 μm, respectively. Different symbols are used to represent optical classifications of the activity. Seyfert 1, 1.2, and 1.5 are denoted as "Seyfert 1," while Seyfert 1.8, 1.9, and 2 are shown as "Seyfert 2" in the diagrams.

Figure 8. Refer to the following caption and surrounding text.

Figure 8. Hardness ratio (HR4) vs. count rate ()/infrared flux (F18 or F90) ratio diagram. Different symbols represent optical classifications; open circles: Seyfert 1, filled circles: Seyfert 2, open squares: LINER, open stars: transition object, filled stars: H ii nucleus, asterisk: BL Lac object, filled triangles: unclassified. (Left) Diagram using 18 μm flux as infrared flux. (Right) Diagram using 90 μm flux as infrared flux.

Standard image High-resolution image

Seyfert 1s are located in the upper left part of the diagram, while Seyfert 2s tend to be located in the lower right. The larger values of hardness ratios and lower X-ray counts relative to infrared fluxes of Seyfert 2s are due to the suppression of lower energy X-rays via photoelectric absorption. H ii galaxies are located in the lower left side. The location of H ii galaxies indicates that X-rays are relatively weak compared to AGNs at a given infrared power. In order to examine the effect of absorption, we defined three groups sorted by and plotted them using different symbols in Figure 9. The ordinate and abscissa are same as in Figure 8. The open circles, open squares, filled circles, and filled triangles represent the groups of < 1022 cm−2, = 1022–23 cm−2, = 1023–24 cm−2, and > 1024 cm−2, respectively. As expected, objects with lower and higher tend to be located at around the upper left and lower right part of the diagrams, respectively.

Figure 9. Refer to the following caption and surrounding text.

Figure 9. Hardness ratio (HR4) vs. count rate ()/infrared flux (F18 or F90) diagram is the same as Figure 2 but only AGNs with measured X-ray spectra are plotted. Different symbols represent measured from X-ray spectra; open circles: cm−2, open squares: 1022 cm−2< < 1023 cm−2, filled circles: 1023 cm−2< < 1024 cm−2, filled triangles: cm−2. The solid lines are tracks expected for various column densities. EPIC-PN count rate (0.2–12 keV) to infrared flux (18 or 90 μm) ratio of 10 is assumed at = 0 cm−2. The plus signs are marked at = (0, 0.5, 1, 5, 10, 50, 100, 150) cm−2 from the upper left-most to the lower right-most point. See the text for details of the assumed spectral model.

Standard image High-resolution image

We made tracks of expected hardness ratios and X-ray/infrared ratios for various values of . A power-law spectrum with a photon index of 1.8 was assumed as incident emission. One percent of the incident emission is assumed to appear as scattered emission keeping the spectral shape. The Galactic absorption of 2 × 1020 cm−2 is assumed as a representative value. The expected values were calculated for various intrinsic column densities from = 0 to 1.5 × 1024 cm−2. The track thus calculated is shown as the solid line in Figure 9, where the values of and are assumed to be 10 when = 0 cm−2. The plus signs are marked at = (0, 0.5, 1, 5, 10, 50, 100, 150) × 1022 cm−2 from the upper left-most to the lowest point. The trend shows that X-ray counts are suppressed and that hardness ratio becomes larger for larger values up to ∼ 5 × 1023 cm−2. If is larger than ∼ 5 × 1023 cm−2, the flux from the scattered emission becomes more significant relative to the absorbed power-law emission, and the hardness ratio becomes small. The above considerations suggest that if objects appeared in the lower right part of the diagram, they are candidates for obscured AGNs. Some of the objects in this region have no published X-ray spectra. We study the X-ray spectra of such candidates for obscured AGNs selected from that region ( < 0.1 or < 1) and HR4 > −0.1 in Section 4.

A disadvantage of using CRs instead of fluxes is that they depend on the instrument used in observations. We calculate the conversion factors from X-ray fluxes to observed CRs. The same spectral shape used to derive the track shown in Figure 9 is assumed. The on-axis response for the EPIC-PN and an integration radius of 60'', which contains 95% of the total flux, are used. The expected CRs for an observed flux in 2–12 keV of 10−14 erg s−1 cm−2 are shown as a function of the absorption column density in Figure 10.

Figure 10. Refer to the following caption and surrounding text.

Figure 10. Flux to count rate conversion factor for various values. The expected count rates in 0.2–12 keV for a source with a flux of 10−14 erg s−1 cm−2 in 2–12 keV are shown. See the text for details of the assumed spectral model.

Standard image High-resolution image

3.2. Infrared Color and X-Ray/Infrared Ratio

Figure 11 shows a diagram versus infrared flux ratios (). This diagram supplements Figure 8 in dividing AGNs and H ii nuclei. H ii nuclei show smaller and ratios compared to AGNs and are located in the lower left part of the diagram; most H ii nuclei have ratios and while only several AGNs are found in this region. AGNs tend to show warmer infrared colors compared to H ii nuclei, and the distributions of the color for Seyfert 1 and Seyfert 2 are almost identical. These results imply that the dust in AGNs are warmer than those in H ii nuclei because of the presence of a hard heating source, confirming earlier studies (e.g., Wu et al. 2009). Comparison between type 1 and type 2 AGNs shows that Mid-IR emission from warm dust near the central engine is visible in both types of AGNs. Thus, we confirmed earlier results based on ground-based observations (Gandhi et al. 2009), or a combination of AKARI and hard X-ray surveys (Ichikawa et al. 2012; Matsuta et al. 2012).

Figure 11. Refer to the following caption and surrounding text.

Figure 11. Count rate ()/infrared flux (F90) ratio vs. infrared flux ratio () diagram. The symbols are same as in Figure 8.

Standard image High-resolution image

4. X-RAY SPECTRA

4.1. The Sample and Data Reduction

As described in Section 3.1, objects in the lower right part of the diagram HR4 versus CR/infrared flux ratio are candidates for obscured AGNs. In order to explore the nature of the candidates, we compiled X-ray results from the literature and analyzed X-ray spectra of objects for which no published results are available.

We first select objects satisfying ( or ) and HR4 > −0.1 from the matched sample using X-ray, 18, and 90 μm. This sample consists of 85 objects and is denoted as the 18 + 90 μm sample hereafter. In addition to this sample, we selected objects satisfying and HR4 > −0.1, where we required the condition FQUAL = 3 only for 90 μm data to increase the size of the sample. 84 objects are selected by these conditions. The list of objects in this 90 μm sample is shown in Table 3. We also selected objects with FQUAL = 3 at 18 μm, FQUAL 3 at 90 μm, and HR4 > −0.1. This 18 μm sample consisting of 10 objects is shown in Table 4. The probable counterparts, redshifts, optical classifications, and absorption column densities determined from X-ray spectra taken from the literature are shown in the tables. Redshifts are available for 83, 62, and 9 objects in the 18 + 90 μm, 90, and 18 μm samples, respectively. The distributions of the redshifts are shown in Figure 12. We excluded stellar and off-nuclear sources from the sample as in the sample shown in Table 1. We inspected the X-ray data of objects with no published X-ray results as candidates for detailed studies. Then the following cases were excluded; objects located in bright diffuse emission of a cluster of galaxies or an early-type galaxy, in the outskirt of the point-spread function of a bright source, on or near the gap between CCD chips, in crowded X-ray source regions such as star-forming regions.

Figure 12. Refer to the following caption and surrounding text.

Figure 12. Distribution of redshifts for the 18 + 90 μm (shaded histogram), 90 μm (hatched histogram), and 18 μm (open histogram) samples.

Standard image High-resolution image

Table 3.  XMM + AKARI 90 μm Sample

XMM Name Other Name Redshift CR HR4 F90 CR/F90 Class References X-ray References
                  (1022 cm−2) Class  
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
2XMM J001106.5−120626 NGC 17/NGC 34/Mrk 938 0.0196 0.087 0.255 17.041 0.005103 Sy2 1 Thin 1
2XMM J001110.4−120114 NGC 35 0.0199 0.016 0.517 1.647 0.009424 Sy1 2
2XMM J002034.9−705526 IRAS 00182−7112 0.327 0.025 0.145 1.064 0.023476 Sy2/L 3 >100a Thick 2
2XMM J004128.0+252958 NGC 214 0.0151 0.037 0.224 3.261 0.011204 Sy2/L 4 Thin 1
2XMM J012445.0+320957 UGC 959/Mrk 991 0.0351 0.040 −0.088 1.760 0.022500 Sy/L 5
2XMMi J013357.6−362935 NGC 612 0.0298 0.136 0.882 2.604 0.052230 Sy2 6 Thin 3
2XMM J051109.0−342335 ESO 362−G008 0.0157 0.059 0.436 0.671 0.088454 Sy2 7 >100b Thick 4
2XMMi J081040.3+481233 2MASX J08104028+4812335 0.0776 0.007 0.451 0.614 0.011751 Sy2 8 Thin 5
2XMM J081937.8+210653 UGC 4332 0.0185 0.017 0.689 1.229 0.013501 Sy2 9 c Un 6
2XMM J094036.4+033437 NGC 2960/Mrk 1419 0.0165 0.028 0.356 1.051 0.026780 Sy2/H ii 10 100a Thick 7
02XMM J101738.5+214119 NGC 3185 0.00406 0.020 0.233 2.030 0.009795 Sy2 11 0.2 Un 8
2XMMi J103234.6+650228 NGC 3259 0.00562 0.014 0.163 1.261 0.010738 Sy1 12 4.3 Thin 9
2XMM J114503.8+195825 NGC 3861 0.0170 0.033 0.196 0.667 0.048966 Sy2/L 13 Thin 10
2XMMi J114721.6+522659 Mrk 1457 0.0486 0.046 0.461 0.798 0.057386 Sy2 14 Thin 11
2XMM J125400.8+101112 IRAS 12514+1027 0.319 0.006 0.233 0.788 0.007252 Sy2 9 Thin 12
2XMMi J132348.4+431804 CGCG 218−007 0.0273 0.086 0.892 1.280 0.067179 Sy2 14 Thin 11
2XMM J133739.8+390917 UGC 8621 0.0201 0.033 0.598 1.254 0.026568 Sy2 16
2XMMi J143722.1+363403 NGC 5695/Mrk 686 0.0141 0.035 −0.045 0.730 0.047974 Sy2 14 c Un 11
2XMM J203306.1−020128 NGC 6926 0.0196 0.079 0.203 9.982 0.007898 Sy2 1
2XMMi J231637.4−021951 NGC 7566 0.0265 0.044 0.693 0.624 0.070063 Sy2 18
2XMM J002126.4−083925 IRAS 00188−0856 0.128 0.005 0.127 2.660 0.001765 L 19 c Un 13
2XMM J010700.8−801828 ESO 013−G012 0.0168 0.018 −0.075 2.517 0.007030 L 20 c Un 7
2XMM J040234.4+714222 UGC 2916 0.0151 0.018 0.016 1.593 0.011068 L/Sy2 21
2XMMi J083529.1+250138 NGC 2611 0.0175 0.013 0.183 2.401 0.005347 L 13
2XMM J100259.0+022034 CGCG 036−024 0.0442 0.012 0.708 0.943 0.012687 L 21
2XMM J101805.6+214955 NGC 3190 0.00424 0.057 0.217 4.699 0.012137 L2 11
2XMM J103702.5−273354 NGC 3312 0.00963 0.069 0.426 1.090 0.063618 L 9 c Un 14
2XMM J110521.8+381402 CGCG 213−027 0.0285 0.024 0.495 1.475 0.016395 L 13
2XMM J115050.9+550836 NGC 3916 0.0191 0.004 0.138 0.872 0.005108 L 13
2XMM J135321.7+402151 NGC 5350/Mrk 1485 0.00774 0.035 0.005 2.734 0.012628 L 22
2XMM J150947.0+570002 NGC 5879 0.00258 0.030 0.474 3.882 0.007652 L2 11 Thin 15
2XMM J033627.6−345832 NGC 1380 0.00626 0.080 0.233 1.511 0.052724 T2 23 c Un 16
2XMM J115348.5+521943 NGC 3953 0.00351 0.199 −0.020 3.444 0.057765 T2 11
2XMMi J124957.9+051841 NGC 4713 0.00218 0.002 0.014 4.872 0.000494 T2 11
2XMM J005656.8−011236 UGC 587 0.0500 0.030 0.410 0.738 0.041171 H ii 13
2XMM J022536.4−050011 2MASX J02253645−0500123 0.0540 0.012 0.301 0.856 0.013848 H ii 6
2XMMi J033930.5−183136 NGC 1402/IRAS 03372−1841 0.0143 0.026 0.034 3.285 0.007953 H ii/L 24
2XMMi J082326.2−045455 NGC 2585 0.0229 0.012 0.612 2.237 0.005539 H ii 25
2XMM J085312.7+162617 SDSS J085312.35+162619.4 0.0638 0.040 0.183 0.830 0.047895 H ii 13
2XMM J102142.6+130654 IRAS 10190+1322 0.0766 0.006 −0.067 4.277 0.001427 H ii 19 c Un 13
2XMM J103712.7−274103 NGC 3314 0.032 0.055 1.995 0.016232 H ii 6 c Un 14
2XMMi J105650.9+065421 UGC 6046 0.0220 0.011 −0.049 0.617 0.017312 H ii 13
2XMM J123653.5+275143 NGC 4559A 0.0251 0.008 0.736 0.817 0.010298 H ii 13
2XMM J124119.3+331111 IRAS 12388+3327 0.0949 0.007 0.117 0.798 0.008567 H ii 13
2XMMi J134651.1+074716 IRAS 13443+0802NE 0.135 0.012 0.181 1.493 0.007963 H ii 19
2XMMi J143413.4+020908 2MASX J14341353+0209088 0.0774 0.005 −0.076 0.707 0.006451 H ii 13
2XMMi J222003.2−201852 2MASX J22200323−2018531 0.0514 0.011 0.429 1.025 0.010935 H ii 6
2XMM J225653.1−363313 IC 5264 0.00647 0.018 −0.094 0.632 0.028225 H ii 6
2XMMi J231505.7−021928 NGC 7546 0.0279 0.005 1.000 0.424 0.011167 H ii 6
2XMM J011119.5−455556 ESO 243−G051 0.0216 0.015 0.611 1.040 0.014610 normal 27
2XMM J143529.6+484430 NGC 5689 0.00721 0.028 0.652 0.724 0.038658 normal 21
2XMM J004330.4−180107 0.014 −0.044 0.367 0.037142 Unclass
2XMMi J005433.8+461239 IRAS 00517+4556 0.013 0.008 0.816 0.015659 Unclass
2XMM J005638.9−011740 CGCG 384−040 0.0502 0.009 0.044 1.173 0.007688 Unclass
2XMM J015253.9+360310 NGC 710 0.0205 0.051 0.138 2.152 0.023561 Unclass
2XMM J022402.5−044137 Arp 54 0.0433 0.015 0.547 0.637 0.024201 Unclass
2XMM J030843.4+405142 2MASX J03084352+4051419 0.018 0.709 0.470 0.037486 Unclass
2XMM J031659.6−441608 IRAS 03152−4427 0.0744 0.030 0.274 0.493 0.061670 Unclass
2XMMi J031821.9+410426 Abell 426:[BM99] 183 0.010 0.145 0.714 0.014358 Unclass
2XMMi J052555.5−661038 0.002 0.403 6.874 0.000221 Unclass
2XMMi J053123.9+120051 AKARI J0531228+120057 0.035 −0.069 1.830 0.019367 Unclass
2XMMi J053512.2−690009 0.013 −0.046 7.285 0.001829 Unclass
2XMM J054532.6−001131 CXO J054532.6−001129 0.005 −0.039 0.559 0.008733 Unclass
2XMMi J091836.4−121322 0.004 0.088 0.383 0.010511 Unclass
2XMM J095611.5+693805 CXO J095611.6+693803 0.002 0.167 6.359 0.000284 Unclass
2XMM J095653.2+694258 SDSS J095654.45+694257.6 0.006 0.013 4.899 0.001228 Unclass
2XMMi J095908.4+130343 CGCG 064−016 0.018 0.055 0.937 0.019106 Unclass
2XMMi J101815.0+731720 AKARI J1018190+731723 0.009 0.489 0.625 0.014765 Unclass
2XMMi J102955.2−031410 CGCG 009−061A 0.0377 0.002 0.286 0.507 0.004120 Unclass
2XMMi J104038.0−461855 ESO 264−G032 0.0228 0.011 0.537 0.474 0.022463 Unclass
2XMM J104219.6+061209 MCG +01−27−029 0.0276 0.017 0.296 1.015 0.016435 Unclass
2XMMi J115943.8−200657 2MASX J11594382−2006579 0.023 0.438 0.389 0.059469 Unclass
2XMM J123302.1+000016 IRAS 12304+0016 0.004 0.543 0.578 0.006639 Unclass
2XMMi J124721.7−411415 NGC 4696B 0.0104 0.072 0.026 0.906 0.079320 Unclass
2XMM J131459.3−163524 NGC 5037 0.00989 0.031 0.737 1.613 0.018920 Unclass
2XMM J132428.8+140533 NGC 5132 0.0244 0.027 −0.012 1.025 0.026722 Unclass
2XMMi J184540.6−630522 0.014 0.411 0.828 0.017253 Unclass
2XMM J222942.7−204607 0.025 0.015 1.375 0.018028 Unclass
2XMM J232008.9+080956 NGC 7617 0.0139 0.010 0.176 0.585 0.016701 Unclass 28
2XMM J232431.2+165204 UGC 12582 0.0356 0.007 0.130 0.616 0.011830 Unclass
2XMM J234044.2−115119 2MASX J23404437−1151178 0.004 0.812 0.496 0.007706 Unclass

Notes. Column 1: XMM name; Column 2: Alternative name; Column 3: Redshift; Column 4: EPIC-PN count rate in the 0.2–12 keV band; Column 4: Hardness ratio HR4; Column 5: Flux density at 90 μm in units of Jy; Column 6: EPIC-PN count rate in 0.2–12 keV to 90 μm flux density ratio; Column 7: Optical classification, Sy—Seyfert, L—LINER, T—transition object, H ii—H ii nucleus, normal—normal galaxy, Unclass—unclassified, 1—type 1, 2—type 2, and a fractional number denotes intermediate types; Column 8: Reference for optical classification; Column 9: Absorption column density in units of 1022 cm−2; Column 10: Classification of absorption measured with X-ray. Un—unabsorbed or slightly absorbed ( cm−2), Thin—Compton-thin ( cm−2), Thick—Compton-thick ( cm−2). If only the lower limit of cm−2 is obtained, such objects are regarded as Compton-thick. Column 11: Reference for absorption column density.

aCompton thick candidates judged from a large Fe–K line EW (>700 eV) and/or flat continuum. bCompton thick candidates in Severgnini et al. (2012). cX-ray spectrum shows no evidence for significant absorption.

References. Optical: (1) Veilleux et al. (1995), (2) Coziol et al. (1994), (3) Buchanan et al. (2006), (4) Parisi et al. (2012), (5) Osterbrock & Phillips (1977), (6) 6dF Galaxy Survey, (7) Fraquelli et al. (2000), (8) Lewis et al. (2011), (9) Véron-Cetty & Véron (2001), (10) Greene et al. (2010), (11) Ho et al. (1997), (12) Thornton et al. (2009), (13) SDSS, (14) LaMassa et al. (2009), (15) Brightman & Nandra (2011b), (16) Wang et al. (2013), (17) Corbett et al. (2003), (18) Masetti et al. (2008), (19) Veilleux et al. (1999), (20) Greenhill et al. (2002), (21) Updated Zwicky catalog (22) Coziol et al. (2004), (23) Ricci et al. (2014b), (24) Kewley et al. (2001), (25) Mauch & Sadler (2007), (26) Hunt & Malkan (2004), (27) Afonso et al. (2005), (28) Bettoni & Buson (1987). X-ray: (1) Brightman & Nandra (2011a), (2) Nandra & Iwasawa (2007), (3) Eguchi et al. (2011), (4) Severgnini et al. (2012), (5) Lewis et al. (2011), (6) Rasmussen et al. (2012), (7) Greenhill et al. (2008), (8) Cappi et al. (2006), (9) Thornton et al. (2009), (10) Sun & Murray (2002), (11) LaMassa et al. (2009), (12) Ruiz et al. (2007), (13) Teng & Veilleux (2010), (14) Hudaverdi et al. (2006), (15) Zhang et al. (2009), (16) Gilmour et al. (2009).

Download table as:  ASCIITypeset images: 1 2 3

Table 4.  XMM + AKARI 18 μm Sample

XMM Name Other Name Redshift CR HR4 F18 CR/F18 Class References X-ray References
      (s−1)   (Jy)       (1022 cm−2) Class  
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
2XMM J030030.5–112456 MCG -02-08-039 0.0299 0.200 0.517 0.371 10.785 Sy2 1 Thin 1
2XMMi J100149.4+284709 3C 234 0.185 0.231 0.402 0.274 14.602 Sy1 2 Thin 2
2XMM J113240.2+525701 HCG 056B/UGC 06527 NED03 0.0264 0.081 0.730 0.210 19.028 Sy2 3 Thin 3
2XMM J121857.5+471814 NGC 4258 0.00149 0.443 0.164 0.449 8.901 Sy1.9 4 Thin 4
2XMM J130527.3-492804 NGC 4945 0.00188 0.413 0.312 9.945 0.402 Sy2/L 5 Thick 4
2XMM J132527.6-430109 NGC 5128/Centaurus A 0.00183 11.563 0.238 13.148 0.304 Sy2 6 14.1 ± 0.7 Thin 4
2XMM J140700.4+282714 Mrk 668 0.0766 0.076 0.165 0.305 13.097 Sy1 1 >100a Thick 5
2XMM J224937.0-191627 MCG -03-58-007 0.0315 0.198 0.430 0.629 6.356 Sy2 1 Thin 1
2XMM J235122.6+200548 NGC 7770 0.0137 0.030 −0.071 0.255 15.669 H ii 8 ...b Un 6
2XMMi J050524.3-673435 2MASX J05052442-6734358 0.075 0.869 0.155 25.735 Unclass

Notes. Column 1: XMM name; Column 2: Alternative name; Column 3: Redshift Column 4: EPIC-PN count rate in the 0.2–12 keV band; Column 5: Hardness ratio HR4; Column 6: Flux density at 18 μm in units of Jy; Column 7: EPIC-PN count rate in 0.2–12 keV to 18 μm flux density ratio; Column 8: Optical classification. Sy—Seyfert, L—LINER, H ii—H ii nucleus, Unclass—unclassified, 1—type 1, 2—type 2, and a fractional number denotes intermediate types; Column 9: Reference for optical classification; Column 10: Absorption column density in units of 1022 cm−2; Column 11: Classification of absorption measured with X-ray. Un—unabsorbed or slightly absorbed ( cm−2), Thin—Compton-thin ( cm−2), Thick—Compton-thick ( cm−2). If only the lower limit of cm−2 is obtained, such objects are regarded as Compton-thick.; Column 12: Reference for absorption column density.

aCompton-thick candidates judged from a large Fe–K line EW (>700 eV) and/or flat continuum. bX-ray spectrum shows no evidence for significant absorption.

References. Optical classification: (1) de Grijp et al. (1992), (2) Buttiglione et al. (2009), (3) Coziol et al. (2004), (4) Ho et al. (1997), (5) Lipari et al. (1997), (6) Storchi-Bergmann et al. (1997), (7) Kewley et al. (2001), (8) Veilleux et al. (1995). X-ray:(1) Brightman & Nandra (2011a), (2) Piconcelli et al. (2008), (3) Guainazzi et al. (2005b), (4) Fukazawa et al. (2011), (5) Guainazzi et al. (2004), (6) Jenkins et al. (2005).

Download table as:  ASCIITypeset image

Table 5.  Log of XMM-Newton Observations

Name IRa Typeb Redshiftc Galactic ObsIDd Net Exposure
        (1022 cm−2)   (ks)
NGC 35 90 Sb 0.019894 0.0220 0150480501 9.2
2XMM J004330.4-180107 90 0.0180 0112880601 9.0
IRAS 00517+4556 90 0.107 0501230301 10.4
UGC 587 90 Sbc 0.049991 0.0343 0012440101 23.4
UGC 959 (Mrk 991) 90 Sa 0.035131 0.0525 0201090401 13.6
IRAS 01356-1307 18 + 90 S0-a 0.040211 0.0157 0502020201 6.8
2MASX J02253645-0500123 90 0.053967 0.0231 0111110301 17.7
2MASX J03084352+4051419 90 0.106 0112880701 44.7
IRAS 03152-4427 90 Sc 0.074421 0.0136 0105660101 16.7
A426[BM99]183 90 0.128 0405410201 8.1
NGC 1402/IRAS 03372-1841 90 SB0 0.014333 0.0536 0404750101 26.7
2MASX J05052442-6734358 18 0.186 0402000501 17.4
2XMMi J052555.5-661038 90 0.120 0505310101 60.2
AKARI J0531228+120057 90 0.226 0405210101 21.1
2XMMi J053512.2-690009 90 0.253 0506220101 69.0
2MASX J05391963-0726190 18 + 90 0.141 0503560101 35.4
2MASX J05430955-0829274 18 + 90 0.149 0503560501 37.2
CXO J054532.6-001129 90 0.191 0301600101 71.0
ESO 205-IG003 18 + 90 Sa 0.050468 0.0497 0554500301 34.6
NGC 2611 90 S0-a 0.017532 0.0277 0406740201 14.0
SDSS J085312.35+162619.4 90 0.063788 0.0251 0206340101 27.0
IC 614 18 + 90 Sab 0.034217 0.0376 0305800101 14.0
CGCG 009-061A 90 0.03770 0.036 0404840201 37.9
ESO 264-G032 90 SBa 0.022799 0.119 0401791201 10.7
MCG +01-27-029 90 Sc 0.027619 0.0258 0151390101 39.6
UGC 6046 90 SBc 0.022030 0.0283 0200530501 20.5
CGCG 213-027 90 0.028474 0.0191 0158970701 27.8
NGC 3953 90 SBa 0.003510 0.0186 0311791401 16.5
2MASX J11594382-2006579 90 0.0328 0555020201 20.1
NGC 4559A 90 Sa 0.025054 0.0137 0152170501 34.4
NGC 4696B 90 E-S0 0.010377 0.0872 0504360201 28.8
NGC 4713 90 SBcd 0.002183 0.0195 0503610101 52.2
IRAS 12596-1529 18 + 90 0.01592 0.036 0203391001 7.3
NGC 5037 90 Sa 0.006294 0.0517 0152360101 31.8
NGC 5132 90 SB0-a 0.024417 0.0171 0108860701 12.7
UGC 8621 90 Sa 0.020091 0.00761 0204651101 9.0
IRAS 13443+0802NE 90 0.135343 0.0193 0405950501 35.8
NGC 5350/Mrk 1485 90 SBbc 0.007742 0.00969 0041180401 16.3
2MASX J14341353+0209088 90 0.077367 0.0276 0553790301 14.7
NGC 5689 90 SB0-a 0.007205 0.0296 0110930401 2.7
2XMM J184540.6-630522 90 0.0688 0405380501 19.0
NGC 6926 90 SBc 0.019613 0.0772 0306050801 8.8
2XMM J222942.7-204607 90 0.0313 0125911001 10.5
IC 5264 90 Sab 0.006471 0.011 0135980201 23.9
NGC 7566 90 SBa 0.026548 0.0371 0501110101 29.3
NGC 7617 90 S0 0.013916 0.0466 0149240101 33.4
2MASX J23404437-1151178 90 0.0318 0152200101 40.6
NGC 7738 18 + 90 SBb 0.022556 0.0345 0211280101 26.9

Notes.

aInfrared sample. "18 + 90"—18 + 90 μm sample, 18—18 μm sample, 90—90 μm sample. bHubble type taken from the HyperLeda. cThe redshift with errors are determined from Fe–K line. dObservation identification number.

Download table as:  ASCIITypeset image

The X-ray and infrared luminosities for the 90 and 18 μm samples are summarized in Table 2 and Figures 5 and 6, in which both results taken from the literature and our own analysis presented in this section are shown. The K-corrections and references for X-ray data and distances are treated in the same manner for the sample presented in Section 2.2.

We retrieved the data for the sample from the XMM-Newton Science Archive and examined their spectra. The XMM-Newton Science Analysis Software (SAS) version 13.0.0 and the calibration files as of 2013 May were used in the data reduction and analysis. We first made light curves of a region that does not contain bright sources in 10–12 keV to examine background stability and time intervals with high background rates were excluded. We extracted source spectra from a circular region centered at the source position with a radius of 4''–60''. The extraction radii were determined to achieve good signal to noise ratio and to avoid nearby sources. Background spectra were made from an off-source region in the same CCD chip and subtracted from the source spectra. After data screening and background subtraction, net source counts in 0.2–12 keV for some objects were found to be lower than 60 counts because of the reduced exposure time. Such objects were excluded from the following analysis since their photon statistics are not sufficient to create spectra of reasonable quality. The final sample consisting of 48 objects for X-ray spectral analysis and the observation log are summarized in Table 6. The infrared sample from which these objects are taken and Hubble type taken from the HyperLeda database (Paturel et al. 2003) are also shown in Table 6.

Table 6.  Results of X-Ray Spectral Fits

Name Modela Γb kT C/dof X-ray
    (1022 cm−2)       (keV)         Class
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
UGC 959/Mrk 991 zpha(zpo) 0.0 0.95 1.39 136.0/134  
  apec+zpha(zpo) 7.90 0.697 110.0/132 16.7 0.584  
  apec+zpha(zoo)* 1.8f 6.25 0.675 110.1/133 17.3 0.584 Thin
IRAS 01356-1307 apec+zpha(zpo+zpha(zpo)) 0.0 1.8f 1.69 0.23 2.40 196.7/173  
    127 1.8f 7.26              
  2apec+zpha(zpo+zpha(zpo))* 0.0 () 1.8f 0.888 6.94 165.1/171 5.80 1.54 Thin
    105 () 1.8f 11.7(25.9c)   1.14        
2MASX J03084352+4051419 apec+pha(po) 0.0 −0.43 0.0840 0.23 0.944 221.8/214  
  apec+pha(po) 18 1.8f 7.77 0.25 0.985 268.1/215  
  apec+pha(po+pha(po))* 0 () 1.8f 0.464 1.04 201.7/213 12.7 Thin
    1.8f 1.94
IRAS 03152-4427 apec+zpha(zpo) 0.0 1.01 0.894 0.176 0.792 156.3/152
  apec+zpha(zpo) 0.20 1.8f 1.92 0.124 1.68 170.0/153
  apec+zpha(zpo+zpha(zpo))* 0.044 () 1.8f 1.28 0.14 ± 0.08 0.839 146.4/151 13.0 5.54 Thin
    1.8f 12.2
2MASX J05052442-6734358 zpha(zpo+zga) 47 0.99 30.7 149.5/150
  zpha(zpo+zga) 59 1.8f 177 149.9/151
  zpha(zpo+zpha(zpo+zga))* 1.8f 2.19 141.2/149 109 24.6 Thin
    1.8f 233
2MASX J05430955-0829274 zpha(zpo+zga) 1.61 84.6/93 11.8 2.44
  zpha(zpo+zga)* 1.8f 7.12 85.4/94 10.7 3.15 Thin
ESO 205-IG003 apec+zpha(zpo+zga) 23 1.51 84.4 0.84 0.986 264.1/174
  apec+zpha(zpo+zpha(zpo+zga))* 0 () 1.8f 1.17 0.480 155.0/174 180 31.6 Thin
    1.8f 163
NGC 2611 apec+zpha(zpo) 153 2.9 289 3.4 1.70 132.6/124
  apec+zpha(zpo) 137 1.8f 27.0 3.4 1.70 132.6/125
  apec+zpha(zpo+zpha(zpo))* 0 () 1.8f 0.446 0.0914 122.2/123 4.72 2.11 Thin
    1.8f 29.7(89.8c)
SDSS J085312.35+162619.4 apec+zpha(zpo+zga) 4.88 0.600 177.6/187 25.2 2.86
  apec+zpha(zpo+zga) 1.8f 9.66 0.604 177.8/188 23.6 3.03
  apec+zpha(zpo+zpha(zpo))* 0.084 (<1.0) 1.8f 0.538 0.386 171.2/187 24.9 3.50 Thin
    1.8f 11.0
CGCG 213-027 apec+zpha(zpo) 28 1.8f 12.2 0.85 1.54 228.9/220
  apec+zpha(zpo+zph(zpo))* 0.0 () 1.8f 1.07 0.820 201.2/218 14.1 0.764 Thin
    1.8f 11.3
NGC 5037 apec+zpha(zpo) 0.0 −1.3 0.0301 0.73 629 232.9/253
  apec+zpha(zpo) 27 1.8f 14.6 0.78 629 249.5/254
  apec+zpha(zpo+zpha(zpo))* 0 (<0.082) 1.8f 0.519 0.328 225.3/252 18.3 0.0746 Thin
    1.8f 24.1
UGC 8621 apec+zpha(zpo) 0.0 0.52 0.256 0.26 1.30 161.7/174
  apec+zpha(zpo) 35 1.8f 8.91 0.30 1.41 182.7/175
  apec+zpha(zpo+zpha(zpo))* 0 () 1.8f 0.576 0.960 145.2/173 8.92 0.376 Thin
    1.8f 11.7
IRAS 13443+0802NE apec+zpha(zpo) 0.0 1.36 0.334 0.554 0.229 172.4/200
  apec+zpha(zpo) 10 1.8f 0.494 0.50 0.186 176.9/201
  apec+zpha(zpo+zpha(zpo))* 0.0 () 1.8f 0.358 0.208 164.7/199 2.38 1.66 Thin
    1.8f 0.899
NGC 5689 zpha(zpo) 0.0 0.250 66.3/55
  zpha(zpo) 23 1.8f 18.4 73.6/56
  zpha(zpo+zpha(zpo))* 0 () 1.8f 0.377 52.9/54 23.0 0.0791 Thin
    1.8f 19.6
IC 5264 pha(zpo) 0 1.38 0.240 60.4/62
  apec+zpha(zpo)* 1.8f 0.807 0.201 48.4/61 2.24 0.00259 Thin
NGC 7566 apec+zpha(zpo+zga) 0.0 −0.50 0.123 0.28 0.936 100.5/98
  apec+zpha(zpo+zpha(zpo+zga))* 0.0 () 1.8f 0.669 0.461 89.2/97 25.0 2.12 Thin
    1.8f 37.9
IRAS 00517+4556 pha(po) 0.0 0.65 0.379 84.5/72  
  pha(ref) 0.0 1.8f 70.1 105.2/73  
  pha(po+pha(po)) 0.0 () 1.8f 0.434 78.9/71 8.34  
    1.8f 3.43
  pha(po+ref)* 0.0 () 1.8f 0.382 424 79.6/72 9.65 Thick?
NGC 1402 apec+zpha(zpo+zga) 0.12 () 0.724 0.647 141.5/163 3.61 0.0168
  apec+zpha(zpo+zga)* 1.8f 0.704 0.638 141.6/164 3.57 0.0166 Thick
  apec+zpha(ref+zga) 0..0 1.8f 23.8 0.767 0.953 172.7/164
  apec+zpha(zpo+ref+zga) 1.8f 0.728 <8.3 0.648 141.5/164 3.63 0.0169
2XMMi J052555.5-661038 pha(po) 0.28 (<1.9) 0.083 123.2/117 1.46
  pha(ref) 0.0 1.8f 9.40 131.0/118  
  pha(po+pha(po)) 1.8f 0.209 120.5/116 1.60  
    1.8f 1.26
  pha(po+ref)* 1.8f 0.174 4.50 122.0/117 1.43 Thick?
2MASX J05391963-0726190 pha(po) 0.95 (<5.3) 0.542 95.8/90 4.62
  pha(po) 1.8f 1.16 99.0/91 2.80  
  pha(ref)* 0.19 () 1.8f 19.8 96.0/91 3.86 Thick?
  pha(po+pha(po)) 1.8f 0.643 94.2/89 4.18
    1.8f 12.0 (27.7c)
IC 614 apec+zpha(zpo+zga) 0.0 () 0.142 0.964 89.4/101 25.7 0.651
  apec+zpha(zpo+zga) 8.9 1.8f 6.84 0.284 1.03 96.0/102
  apec+zpha(zpo+zpha(zpo+zga)) 0.017 () 1.8f 0.457 0.707 89.1/100 18.2 0.923
    1.8f 8.49
  apec+zpha(ref+zga)* 0 () 1.8f 85.9 0.951 90.6/102 21.2 0.548 Thick
2MASX J11594382-2006579 apec+pha(po) 0.0 () 0.556 0.733 83.4/66 6.34
  apec+pha(po) 1.8f 1.49 0.632 84.8/67 3.83
  apec+pha(ref)* 0 () 1.8f 27.8 0.657 83.7/67 5.47 Thick?
2XMMi J184540.6-630522 zpha(zpo+zga) 0.070 (<0.89) 0.554 61.3/55 7.39 6.51
  zpha(zpo+ref+zga)* 0.22 () 1.8f 0.833 17.0 60.8/55 8.10 7.43 Thick
NGC 6926 apec+zpha(zpo+zga) 0.0 () 3.21 0.700 93.0/89 4.41 0.038
  apec+zpha(zpo+zga) 0.0 () 1.8f 5.13 0.517 94.0/90 3.70 0.032
  apec+zpha(ref+zga)* 0 (<2.5) 1.8f 15.7 1.24 94.4/90 5.16 0.0445 Thick
  apec+zpha(zpo+ref+zga) 0 (<2.5) 1.8f 0.321 9.40 0.739 91.5/89 4.87 0.042
2MASX J23404437-1151178 apec+zpha(zpo+zga) 3.2 1.8f 1.88 0.74 0.115 64.8/80
  apec+zpha(zpo+zpha(zpo+zga)) 0.16 (<1.7) 1.8f 0.094 0.0851 59.0/79 1.18 1.30
    74.4fd 1.8f 1.58  
  apec+zpha(ref+zga) 0.0 1.8f 3.07 0.71 0.113 63.3/80 0.959 0.204
  apec+zpha(zpo+ref+zga)* 0.19 (<1.90) 1.8f 0.0938 1.41 0.0860 59.8/79 0.861 0.194 Thick
NGC 7738 apec+zpha(zpo+zga) 0.062 () 0.399 0.821 64.8/96 3.62 0.0419
  apec+zpha(zpo+zga) 0.16 () 1.8f 0.528 0.798 65.1/97 3.11 0.0365
  apec+zpha(ref+zga) 0 () 1.8f 14.3 1.26 69.7/97 4.28 0.0485
  apec+zpha(zpo+ref+zga)* 0 () 1.8f 0.439 4.87 0.824 64.8/96 3.75 0.0434 Thick
NGC 35 zpha(zpo) 0.23 4.1 1.50 64.7/54
  apec 0.41 0.591 69.5/55
  apec+zpha(zpo)* 0 (<0.83) 1.8f 0.312 0.380 58.3/53 1.06 0.0094
2XMM J004330.4-180107 pha(po)* 1.31 101.3/111 8.32
  apec 64e 5.60 107.6/112  
UGC 587 zpha(zpo) 0.0 (<0.11) 1.8f 0.619 68.8/78 1.98f 0.116f
  apec* 0.496 65.7/78 0.070f 0.0050f
2MASX J02253645-0500123 zpha(zpo)* 0 (<0.037) 1.39 98.0/123 3.34 0.232
  apec 4.7 4.35 115.8/124
A426[BM99]183 pha(po)* 1.00 63.9/92 1.52
  apec 1.96 62.0/93 1.18
AKARI J0531228+120057 pha(po)* 0 (<0.13) 2.70 113.4/143 8.84
  apec 8.97 110.6/143 8.59
2XMMi J053512.2-690009 pha(po)* 0 () 0.862 158.7/180 5.47
  apec 43 () 4.43 159.3/181 6.05
CXO J054532.6-001129 pha(po)* 0.563 146.4/151 2.74
  apec 64e 1.59 159.3/152
CGCG 009-061A zpha(zpo) 0.0 () 1.8f 0.169 74.1/86 0.552f 0.0182f
  apec* 0.101 71.8/86 0.0022f 0.000081f
ESO 264-G032 zpha(zpo) 0.11() 0.507 35.4/42 0.352 0.00438
  apec 0.854 35.0/43 0.520 0.00646
  apec+zpha(zpo)* 0.18(<2.2) 1.8f 0.228 0.257 33.2/41 0.746 0.00902
MCG +01-27-029 zpha(zpo)* 2.04 104.5/109 2.32 0.0429
  apec 15 () 3.04 110.7/110 4.27 0.0741
UGC 6046 zpha(zpo)* 1.29 77.9/85 0.854 0.00998
  apec 1.64 79.7/86 1.29 0.0145
NGC 3953 zpha(zpo)* 0.045 (<0.17) 0.734 138.7/153 2.61 0.000713
  apec 2.32 140.2/154 3.24 0.000883
NGC 4559A zpha(zpo) 0.0 1.8f 0.373 44.9/58
  apec 0.373 42.3/58 0.0012 0.000020
  apec+zpha(zpo)* 0 () 1.8f 0.162 0.244 38.7/56 0.542 0.00773
NGC 4696B apec+zpha(zpo) 0.008 1.8f 0.721 0.46 2.27 353.0/355
  2apec+zpha(zpo)* 1.8f 0.844 2.36 331.2/353 2.71 0.00748
            1.01
NGC 4713 zpha(zpo+zga) 0.0 2.1 0.962 368.0/349
  apec 4.4 3.07 491.5/351
  apec+zpha(zpo+zga)* 0.0 (<0.071) 0.581 0.499 318.1/347 3.08 0.000324  
IRAS 12596-1529 zpha(zpo) 2.73 131.5/144 3.50 0.0207
  apec 4.35 132.0/145 4.17 0.0242
  apec+zpha(zpo+zga)* 1.8f 1.16 0.731 119.4/142 4.88 0.0283
NGC 5132 zpha(zpo) 3.12 75.5/102 1.20 0.0177
  apec 3.4 3.54 87.3/103
  apec+zpha(zpo)* 0.069 (<0.30) 1.8f 0.931 0.329 75.6/101 3.10 0.0423
NGC 5350 zpha(zpo) 0.23 3.8 2.61 144.2/142
  apec 2.6 3.24 184.8/143
  apec+zpha(zpo)* 0 () 1.8f 0.577 0.484 132.2/141 2.02 0.00269
2MASX J14341353+0209088 zpha(zpo) 0.19 () 0.561 53.9/50 0.017 0.0280
  apec 2.2 0.763 58.4/51
  apec+zpha(zpo)* 1.8f 0.204 0.164 53.9/49 0.617 0.0894
2XMM J222942.7-204607 pha(po)* 0 () 0.261 11.0/141 1.01
  apec 5.4 () 0.875 11.1/142 1.04
NGC 7617 zpha(zpo) 0.036 () 0.562 210.0/220 0.473 0.00210
  apec 0.91 0.403 259.1/221
  apec+zpha(zpo)* 0 (<0.062) 1.8f 0.281 0.266 209.9/219 0.96 0.00416

Notes. Column 1: Galaxy Name; Column 2: Spectral Model; Column 3: Absorption column density; Column 4: Photon index of the power-law component; Column 5: Normalization of the power-law component in units of photons cm−2 s−1 at 1 keV; Column 6: Normalization of the reflection component represented by the pexrav model in units of photons cm−2 s−1 at 1 keV; Column 7: Temperature of the apec plasma model; Column 8: Normalization of the apec thermal plasma model in units of where is the angular size distance to the source (cm), is the electron density (cm−3), and is the hydrogen density (cm−3); Column 9: C statistic/degree of freedom; Column 10: Observed flux in 2–10 keV in units of erg cm−2 s−1; Column 11: Absorption-corrected luminosity in 2–10 keV in units of 1042 erg s−1; Column 12: Classification of absorption for adopted spectral model. Thin: Compton-thin ( cm−2). Thick: Compton-thick ( cm−2).

apha: photoelectric absorption; po: power law; apec: thermal plasma emission model; ref: reflected emission from cold matter; ga: Gaussian. The first letter of z stands for emission/absorption at the source redshift. * denotes the adopted model in Figures 79. b"f" denotes fixed parameter. cThe effect of Compton scattering is approximately taken into account by using the cabs model in XSPEC. The column density is assumed to be same as that for photoelectric absorption. dThe absorption column density was not constrained and fixed at the best-fit value. ePegged at the upper bound allowed in the fit. fFlux and luminosity are derived by extrapolating the result of fit using energies below 2 keV. gCenter energy of the Gaussian component is fixed at 6.97 keV.

Download table as:  ASCIITypeset images: 1 2 3 4

The response matrix file and ancillary response file were made by using the SAS. The spectra were binned so that each bin contains at least one count. More channels are binned in the figures shown below for presentation purposes. A maximum-likelihood method using the modified version of C statistic (Cash 1979) was employed to fit background-subtracted spectra. Spectral fits were performed with XSPEC version 12.8.0. The errors represent the 90% confidence level for one parameter of interest. Errors are not shown for the cases that the value of C statistic is much worse than that for the best-fit model.

The Galactic absorption was applied to all the models examined below. The Galactic absorption column densities (Kalberla et al. 2005) were obtained by the FTOOL nh and shown in Table 6. phabs or zphabs in XSPEC were used as a photoelectric absorption model. We examined the presence of an Fe–K emission line at around 6.4 keV by adding a Gaussian component. The line center energy was left free if the photon statistics were sufficient to constrain the energy, otherwise 6.4 keV was assumed. The line width was fixed at a Gaussian σ of 10 eV. For objects with a known redshift, all the model components except for the Galactic absorption were assumed to be emitted or absorbed at the source redshift. If a redshift is previously unknown and an Fe–K emission line is visible in our X-ray spectra, the source redshift was treated as a free parameter and was determined from the line, whose central energy was assumed to be 6.4 keV in the source rest frame. The redshifts for the X-ray analysis sample range from 0.00218 to 0.188 with a median of 0.020, where redshifts determined by an Fe–K line are included. We assumed a redshift of z = 0 for all the other cases. Fe line parameters are not shown if the photon statistics around 6.4 keV are not sufficient and if no meaningful constraints on the line is obtained. The results of the spectral fits described below are summarized in Tables 6 and 7. The adopted model is marked with an asterisk in Table 6. In these tables, spectral parameters for the best-fit models and models with fewer model components are shown for comparison. The observed fluxes and luminosities corrected for absorption (for objects with a known redshift) in the 2–10 keV band were derived for the best-fit spectral model and models giving a similar quality of fits to the best fit.

4.2. Results

We first summarize measurements of absorption column densities taken from the literature, and then provide detailed results of our own spectral analysis for the 48 objects, for which no X-ray results have been published so far. The 48 objects analyzed are devided into 3 groups: (1) objects with absorbed spectrum, (2) objects showing very flat continuum and/or strong Fe–K fluorescent line, and (3) objects showing unabsorbed spectrum, and explained in turn.

4.2.1. X-Ray Results Taken from the Literature

We compiled the results of X-ray spectral fits for our samples from the literature. Since our aim is to find obscured AGNs by combining X-ray and infrared data, we tabulated absorption column densities () derived from X-ray spectra in Tables 1, 3, and 4 for the 18 + 90 μm, 90, and 18 μm samples, respectively. The values are classified into three classes "Unabsorbed," "Compton-thin," and "Compton-thick" for objects with cm−2, cm−2, and cm−2, respectively. If only a lower limit on of 1 × 1024 cm−2 is presented, such sources are regarded as Compton-thick. These classifications are shown in the "X-ray Class" column in Tables 1, 3, and 4. The classifications of X-ray absorption are available for 72, 22, and 9 objects for the 18 + 90 μm, 90, and 18 μm samples, respectively.

4.2.2. Absorbed Spectrum

X-ray spectra of 16 of the 48 objects we analyzed show convex shape, which is a signature of absorbed emission, implying the presence of an obscured AGN, at energies above ∼2 keV. Their spectra are shown in Figure 15. We fitted their spectra by a power-law model absorbed by neutral matter. First the photon index was treated as a free parameter, and then a model with a photon index fixed at 1.8 was examined. The result for a free photon index is shown if a meaningful constraint on the index is obtained.

The spectrum of 2MASX J05430955−0829274 is well fitted with this absorbed power-law model accompanied by an emission line at 6.4 keV. The rest of the objects show additional emission at energies below a few keV. We tried to model this emission by power law, APEC thermal plasma (Smith et al. 2001), or a combination of both. For the models with one power-law component, the photon index was left free or fixed at 1.8. The photon indices of the heavily absorbed power law, which dominates hard emission, and the additional power law representing the soft part of the spectra were fixed at 1.8 for models containing two power-law components. A common photoelectric absorption model, which represent the absorption in the host galaxies, was applied to both of the power-law components. The APEC component was assumed to be absorbed only by the Galactic column. The spectra of the two objects 2MASX J05052442−6734358 and NGC 5689 are represented by the two power-law model. Other objects show excess emission around 0.6–1.0 keV, which implies the presence of Fe–L emission lines from hot plasma, and the APEC model was used to express this feature. An APEC + absorbed power-law model describes the spectra of UGC 959 and IC 5264. The spectra of the rest of the objects were fitted with a combination of two power-law and APEC components. IRAS 03156−1307 required two temperature APEC components in addition to the two power-law components.

The best-fit column densities for the heavily absorbed power-law component are ≈ 2×1022 − 1.4×1024 cm−2, which is a range expected for obscured AGNs. The column densities of the two objects IRAS 01356−1307 and NGC 2611 exceed 1 × 1024 cm−2 and the effect of Compton scattering cannot be neglected. Therefore, we multiplied the cabs model in XSPEC (), where is the Thomson scattering cross section. Although the energy dependence of the cross section is not taken into account, this model approximates the shape of the continuum transmitted through Compton-thick matter (Ikeda et al. 2009). This model affects only the normalization of the heavily absorbed power-law component.

The results of the spectral fits are summarized in Table 6, and the adopted models and data/model ratios are shown in Figure 15. If two or more models provide similar C statistics, we adopt models with the best-fit photon index in the range of 1.5–2.1, which is typical for X-ray spectra of AGNs. The results of spectral fits to the Fe–K line are shown in Table 7. The improvement of the C statistic by adding a Gaussian line is also shown in Table 7. An Fe–K emission line is seen in the spectra of 2MASX J05052442−6734358, 2MASX J05430955−0829274, and ESO 205−IG003. A hint of Fe–K emission is seen in SDSS J085312.35+162619.4. The photon statistics are poor around 6.4 keV in IRAS 03136−1307 and Fe line parameters are not shown for this object. The improvement of the C statistic for other objects is small for one additional parameter (normalization of a Gaussian).

Table 7.  Parameters of the Fe–K Emission Line

Name Modela Line Centerb EW
    (keV)   (eV)  
(1) (2) (3) (4) (5) (6)
UGC 959/Mrk 991 apec+zpha(zoo) 6.4f 0.0() 0 () 0.0
IRAS 01356-1307
2MASX J03084352+4051419 apec+pha(po+pha(po))c 6.4f 0.025 () 50 () 0.1
IRAS 03152-4427 apec+zpha(zpo+zpha(zpo)) 6.4f 0.026 () 47(<360) 0.1
UGC 2730 apec+zpha(zpo) 6.4f 0.052 () 450 () 1.2
2MASX J05052442-6734358 zpha(zpo+zpha(zpo+zga)) 6.4f 8.8
2MASX J05430955-0829274 zpha(zpo+zga) 6.4f 7.2
ESO 205-IG003 apec+zpha(zpo+zpha(zpo+zga)) 6.38 ± 0.05 0.49 ± 0.03 12.8
NGC 2611 apec+zpha(zpo+zpha(zpo)) 6.4f 0.853 () 200 () 0.4
SDSS J085312.35+162619.4 apec+zpha(zpo+zpha(zpo)) 6.4f 3.2
CGCG 213-027 apec+zpha(zpo+zph(zpo)) 6.4f 0.0 () 0 () 0.0
NGC 5037 apec+zpha(zpo+zpha(zpo)) 6.4f 0.0 (<0.12) 0 () 0.0
UGC 8621 apec+zpha(zpo+zpha(zpo)) 6.4f 0.10 () 260 () 1.1
IRAS 13443+0802NE apec+zpha(zpo+zpha(zpo)) 6.4f 0.020 () 440 () 2.3
NGC 5689 zpha(zpo+zpha(zpo)) 6.4f 0.0580 () 84 () 0.1
IC 5264 apec+zpha(zpo) 6.4f <0.062 0.3
NGC 7566 apec+zpha(zpo+zpha(zpo+zga)) 6.4f 0.378 () 260 () 1.2
IRAS 00517+4556 pha(po+pha(po)) 6.4f 0.0() 0 () 0.0
  pha(po+ref) 6.4f 0.0 () 0 () 0.0
   
NGC 1402 apec+zpha(zpo+zga) 18.7
2XMMi J052555.5-661038
2MASX J05391963-0726190 pha(po) 6.4f 4.8
  pha(ref) 6.4f 0.029 () 460 () 2.1
IC 614 apec+zpha(zpo+zpha(zpo+zga)) 6.31 ± 0.053 12.5
  apec+zpha(ref+zga) 11.0
2MASX J11594382-2006579 apec+pha(ref) 6.4f 0 () 0 (<500) 0.0
  apec+zpha(ref) 6.4f 3.0d
2XMM J184540.6-630522 zpha(zpo+zga) 6.4f 6.6
  zpha(zpo+ref+zga) 6.4f 5.3
NGC 6926 apec+zpha(ref+zga) 0.168 19.5
  apec+zpha(zpo+ref+zga) 21.3
2MASX J23404437-1151178 apec+zpha(zpo+zpha(zpo+zga)) 6.4fd 4.7
  apec+zpha(zpo+ref+zga) 6.4fd 6.9
NGC 7738 apec+zpha(zpo+ref+zga) 6.434 ± 0.067 9.4
NGC 35 apec+zpha(zpo) 6.4f 0.0234 (<0.111) 2110 (<10150) 0.9
2XMM J004330.4-180107 pha(po) 6.4f 0.0 () 0 () 0.0
UGC 587
2MASX J02253645-0500123 zpha(zpo) 6.4f 0.0014 () 37 (<1610) 0.0
A426[BM99]183
AKARI J0531228+120057
CXO J053513.5-690013
CXO J054532.6-001129 pha(po) 6.4f 0 () 0 () 0.0
CGCG 009-061A
ESO 264-G032
MCG +01-27-029
UGC 06046
NGC 3953
NGC 4559A
NGC 4696B
NGC 4713 apec+zpha(zpo+zga) 6.4f 5.3
IRAS 12596-1529 apez+zpha(zpo) 6.4f 0.0359 () 810 () 1.6
  apez+zpha(zpo+zga) 6.97f 5.5
NGC 5132
NGC 5350
2MASX J14341353+0209088
2XMM J222942.7-204607 pha(po) 6.4f <0.053 0
NGC 7617 apec+zpha(zpo) 6.4f 0.9

Notes. Column 1: Galaxy Name; Column 2: Spectral Model used to derive the Fe–K line parameters; Column 3: Center energy of the Gaussian line; Column 4: Normalization of the Gaussian in units of photons cm−2 s−1 in the line; Column 5: Equivalent width; Column 6: The improvement in fit compared to the model without a Gaussian line.

aThe description of model components is same as in Table 5. A photon index of 1.8 was assumed for the power-law component. "..." menas that X-ray flux around Fe–K line is low and that no meaningful constraint on Fe–K line is obtained. b"f" denotes fixed parameter. cIf a redshifted model is used and if the line center energy is assumed to be 6.4 keV, redshift is weakly constrained as 0.046 ± 0.066 and the improvement of the fit is dThe peak at 5.8 keV seen in the spectrum was assumed to be Fe–K line at 6.4 keV in the source rest frame and the best-fit redshift of was obtained. was calculated by assuming the redshift determined from the model containing a Gaussian line.

Download table as:  ASCIITypeset images: 1 2

4.2.3. Flat Continuum and Strong Fe–K Emission Line

Ten objects show a flat continuum at energies above a few keV and/or a strong Fe–K emission line at around 6.4 keV. The spectral shape is much flatter than that typically observed in AGNs and implies that the spectrum is a combination of heavily absorbed and less absorbed power laws with typical photon indices for AGNs (two power-law model) or that the spectrum is reflection-dominated. Therefore, we examined continuum models for these two cases. We assumed a common photon index of 1.8 for the two power-law model. The pexrav model in XSPEC was used to represent a continuum reflected from cold matter (Magdziarz & Zdziarski 1995). The incident spectrum is assumed to be a power law with a photon index of 1.8 and an exponential cutoff at 300 keV. The reflection scaling factor (rel_refl) was set to to represent reflected emission alone. The inclination angle of the reflector was assumed to be where the inclination of 0° corresponds to face on.

The spectrum of 2MASX J05391963−0726190 is fitted either by a pure reflection model or a two power-law model. A pure reflection model does not describe the spectra of IRAS 00517+4556, 2XMM J052555.5−661038, and 2XMM J184540.6−630522. The continuum of 2XMMi J184540.6−630522 is fitted by a flat power-law model or a combination of reflection and slightly absorbed power law. A combination of reflection and virtually unabsorbed power law or a two power-law model represent the spectra of IRAS 00517+4556 and 2XMMi J052555.5−661038.

The six objects, NGC 1402, IC 614, 2MASX J11594382−2006579, NGC 6926, 2MASX J23404437−1151178, and NGC 7738, show excess emission around 0.6–1.0 keV, suggesting the presence of emission from optically thin plasma. We used an APEC plasma model to represent this feature. An APEC + reflection model describes the continua of IC 614, 2MASX J11594382−2006579, and NGC 6926. An APEC + absorbed power-law model also provided a similar quality of fit to 2MASX J11594382−2006579. 2MASX J23404437−1151178 and NGC 7738 require an additional power-law component, where we assumed a common for the reflection and power-law components since the quality of the data is not sufficient to constrain values for these components separately. The spectrum of NGC 1402 is not very flat, and a combination of APEC and lightly absorbed power law represents the shape of the continuum.

The six objects, NGC 1402, IC 614, 2XMMi J184540.6−630522, NGC 6926, 2MASX J23404437 −1151178, and NGC 7738, show an emission line at around 6.4 keV in the source rest frame with an EW in the range of 1.1–4.6 keV. An Fe–K line at 6.4 keV is not detected in other objects except for a weak hint of a line in NGC 4713. An emission-line-like excess at around 7.0 keV is seen in the spectrum of IRAS 12596−1529. If this line is assumed to be from H-like Fe at 6.97 keV, the C statistic is improved by ΔC = 5.5 for one additional parameter (normalization of a Gaussian). 2MASX J11594382−2006579 shows a line-like emission at If this emission is an Fe–K line at 6.4 keV in the source rest frame, the redshift is estimated to be though the improvement of the fit is only ΔC = 3.0 for two additional parameters (normalization of a Gaussian and source redshift). Limits on the EW of Fe–K line at 6.4 keV were derived for objects with sufficient counts around 6.4 keV.

The results of the spectral fits are summarized in Tables 6 and 7. The observed spectra, adopted models, and data/model ratios are shown in Figure 16. If two or more models give fits of similar quality, we adopt one or two models as the most appropriate ones satisfying the following conditions: (1) the best-fit photon index is in the range of 1.5–2.1, and (2) the constraint on the EW on an Fe–K fluorescent line at 6.4 keV is consistent with the best-fit as observed in obscured AGNs (Guainazzi et al. 2005a; Fukazawa et al. 2011).

4.2.4. Unabsorbed Spectrum

The rest of the objects do not show a signature of heavy absorption. We applied an absorbed power-law model to the spectra. The photon index Γ was first treated as a free parameter. If the photon index was not well constrained, Γ = 1.8 was assumed. Good fits were obtained for 12 objects (2XMM J004330.4−180107, 2MASX J02253645−0500123, A426[BM99]183, AKARI J0531228+120057, 2XMMi J053512.2−690009, CXO J054532.6−001129, MCG +01−27−029, UGC 6046, NGC 3953, NGC 5132, 2XMM J222942.7−204607, and NGC 7617). An APEC plasma model was also examined instead of power law, and similar quality of fits were obtained for A426[BM99]183, AKARI J0531228+120057, 2XMMi J053512.2−690009, UGC 6046, NGC 3953, and 2XMM J222942.7−204607. The spectra of UGC 587 and CGCG 009−061A appear very soft, and an APEC model with a temperature of 0.3–1 keV provided a good description of the data.

The single-component models do not fit the spectra of eight objects (NGC 35, ESO 264−G032, NGC 4559A, NGC 4696B, NGC 4713, IRAS 12596−1529, NGC 5350, and 2MASX J14341353+0209088). Although the spectra of NGC 5132 and NGC 7617 are fitted by a power-law model, the resulting photon indices are very steep (3.3 and 2.8, respectively) and may indicate the presence of a soft component. We therefore examined a two-component model consisting of APEC plasma and absorbed power law with a fixed photon index of 1.8. This model describes the spectra of all but NGC 4696B. The addition of a second APEC component well fits the spectrum of NGC 4696B.

Two objects (NGC 4713 and IRAS 12596−1529) show a hint of an emission line at 6–7 keV. The improvements of the C statistic are 5.3 and 5.5 for one additional parameter (normalization of a Gaussian) if the line center energies of 6.4 and 6.97 keV were assumed for NGC 4713 and IRAS 12596−1529, respectively. Other objects do not show an Fe–K emission line feature. The upper limits on the EW of a Gaussian line at 6.4 keV were derived for objects with sufficient photon statistics around an Fe–K line and summarized in Table 7. The results of the fits are summarized in Table 6. The observed spectra and the adopted models are shown in Figure 17.

5. DISCUSSION

5.1. Selection of Obscured AGNs

We made diagnostic diagrams using X-ray CRs, X-ray hardness, and infrared fluxes, which can be used to select candidates for obscured AGNs. By accumulating the published results of X-ray spectral analysis, we found that the regions satisfying ( or ) and HR4 > −0.1 are the loci for obscured AGNs. We analyzed the X-ray spectra of 48 objects with X-ray counts greater than 60, for which no X-ray spectra are published. Of these, 26 show a signature of absorbed X-ray spectra. Their classifications of X-ray absorption as Compton-thin or Compton-thick are summarized in Table 6. The 16 objects analyzed in Section 4.2.2 show a spectral curvature indicative of the continuum absorbed by ≈ 1022–24 cm−2. Their best-fit values are in the range of cm−2, which are typically observed in Seyfert 2 galaxies. The two objects showing the largest best-fit ( cm−2 for IRAS 01356−1307 and 1.4 × 1024 cm−2 for NGC 2611) are regarded as Compton-thin in the discussion below since these values are slightly below the boundary of Compton-thin/thick column densities (1.5 × 1024 cm−2; Comastri et al. 2004). Note, however, that the boundary value of between Compton-thin/thick depends on the assumed abundance of the absorber (Yaqoob et al. 2010). The 10 objects analyzed in Section 4.2.3 show a flat continuum and/or a strong Fe–K emission line. Six of these objects are most likely to be Compton-thick AGNs judging from their flat continuum and strong Fe–K emission line. The Fe–K emission line is not significant in 4 (IRAS 00517+4556, 2MASX J05255807−6610523, 2MASX J05391963−0726190, and 2MASX J11594382−2006579) of the 10 objects, and their flat continuum could be interpreted as either a reflection-dominated spectrum or a combination of mildly absorbed (4 × 1022–4×1023 cm−2) and less absorbed components. These objects are tentatively regarded as Compton-thick in the following discussion and denoted as "Thick?" in the X-ray class column of Table 6. Thus, 26 objects in total are most likely obscured AGNs, for which X-ray signatures of the presence of AGNs are reported for the first time in this paper. The 22 objects analyzed in Section 4.2.4, on the other hand, show no clear evidence for the presence of obscured AGNs.

By combining the X-ray results taken from the literature and our own analysis, Table 8 summarizes the number of objects in the 18 + 90 μm, 90, and 18 μm samples, objects with X-ray measurements of absorption column densities, and unabosrbed/Compton-thin/Compton-thick objects. Usable X-ray data are available for 84% of the combined sample. Among the 151 objects with X-ray measurements, 113 (75%) are absorbed by Compton-thin or Compton-thick matter. If only sources securely detected at 18 μm are used, 82 of 89 objects (92%) are absorbed.

Table 8.  Summary of the Number of Objects

Sample No. X-ray Unabsorbed Thin Thick Absorbed
            Fraction
(1) (2) (3) (4) (5) (6) (7)
18 + 90 85 79 (93%) 6 44 29 92%
90 84 62 (74%) 31 21 10 50%
18 10 10 (100%) 1 7 2 90%
Total 179 151 (84%) 38 72 41 75%

Note. Column 1: Infrared sample. 18 + 90— the 18 + 90 μm sample, 18— the 18 μm sample in Table 4; 90— the 90 μm sample in Table 3. Column 2: Number of objects in the sample. Column 3: Number of objects with X-ray measurement of absorption. The fraction of objects with available X-ray absorption data is shown in parenthesis. Column 4: Number of unabsorbed objects with cm−2. Column 5: Number of objects absorbed by Compton-thin matter. Column 6: Number of objects absorbed by Compton-thick matter. Column 7: Fraction of objects absorbed by Compton-thin or Compton-thick matter in the sample.

Download table as:  ASCIITypeset image

The MIR band is better for detecting emission from warm dust heated by an AGN compared to the far-infrared band. Therefore, we first discuss the success rate to find obscured AGNs selected by the criteria using 18 μm data. Among our XMM+AKARI sample, 79 objects satisfy the conditions and HR4 > −0.1. The absorption column densities of 44 and 29 objects among them are in cm−2 and larger than 1.5 × 1024 cm−2, respectively. The detection rate of AGNs absorbed by > 1022 cm−2 is 73/79 ≈ 92%, which demonstrates the efficiency of our criteria finding obscured AGNs.

Severgnini et al. (2012) applied a similar selection technique to the sample derived from the IRAS point-source and 2XMM catalogs. They used the conditions of and HR4 > −0.2 and selected 43 candidates for Compton-thick AGNs. Of the 43 AGNs, 40 are in common with our sample of obscured AGNs. Two objects (IRAS 04507+0358 and 3C321) are not selected in our sample since EPIC-PN data are not available for them. One object (NGC 5194) has an HR4 value of that is slightly smaller than our adopted boundary (HR4 > −0.1) but satisfies Severgnini et al.'s criterion (HR4 > −0.2).

In Severgnini et al.'s sample, 32 of 43 are confirmed to be Compton-thick AGNs. The classification of Compton thickness for four objects are model-dependent, and seven are Compton-thin. If the conversion between IRAS 25 μm and AKARI 18 μm flux densities of Equation (3) in Ichikawa et al. (2012) is assumed, their criterion translates into 0.24 and 0.64 for assumed of 5 × 1023 cm−2 and 1 × 1022 cm−2, respectively, where the same spectral shape as in Section 3.1 was used. Thus, our criteria () cover X-ray brighter objects relative to MIR fluxes, and this might result in the smaller Compton-thick fraction (44/79) derived for our sample.

Another difference between Severgnini et al.'s and our selection criteria is the limited X-ray and infrared flux levels. Severgnini et al. selected objects with an X-ray flux in 4.5–12 keV larger than 1 × 10−13 erg s−1 cm−2. This flux corresponds to 7 × 10−14 erg s−1 cm−2 in 2–10 keV, if = 5 × 1023 cm−2 and the spectral shape used in Section 3.1 are assumed. The fluxes in 2–10 keV of all but one in our sample are larger than this flux. The flux of one object (2MASX J05391963−0726190) is ∼4 × 10−14 erg s−1 cm−2 in 2–10 keV, which is only slightly below the flux limit. We also compared the distributions of X-ray CRs in 0.2–12 keV. The solid histogram in Figure 13 is the distribution for 68 obscured AGN candidates satisfying the 18 or 90 μm criteria selected from Table 1. The distribution of the CRs for the Compton-thick candidates in Severgnini et al. (2012) is shown as a dashed histogram in the same figure. Our sample contains more X-ray fainter objects compared to Severgnini et al.'s sample. Therefore, the X-ray flux limit likely explains the difference in the Compton-thick fraction in part. The distributions of 18 μm fluxes for the 70 objects satisfying the condition in the 18 + 90 μm sample and Severgnini et al's Compton-thick candidates are also compared in Figure 14. We compiled 18 μm fluxes measured with AKARI IRC for the latter. Six sources do not have 18 μm data, and their IRAS 25 μm fluxes are converted to 18 μm fluxes by using the Equation (3) in Ichikawa et al. (2012). Although the lower bounds of the distribution are similar, our sample contains a somewhat larger number of objects at fainter infrared fluxes. In summary, the combination of the larger numbers of X-ray faint or infrared faint objects compared to Severgnini et al.'s is likely to be a reason why we were able to find new obscured AGNs not included in Severgnini et al.'s sample.

Figure 13. Refer to the following caption and surrounding text.

Figure 13. Distribution of X-ray count rates in the 0.2–12 keV band for 18 + 90 μm sample (solid histogram) and Severgnini et al.'s sample (dashed histogram).

Standard image High-resolution image
Figure 14. Refer to the following caption and surrounding text.

Figure 14. Distribution of flux densities at 18 μm for objects satisfying the condition in 18 + 90 μm sample (solid histogram) and Severgnini et al.'s sample (dashed histogram).

Standard image High-resolution image
Figure 15. Refer to the following caption and surrounding text.
Standard image High-resolution image
Figure 15. Refer to the following caption and surrounding text.
Standard image High-resolution image
Figure 15. Refer to the following caption and surrounding text.

Figure 15. EPIC-PN spectra of objects showing absorbed continuum. (Upper panel) Data (crosses) and adopted best-fit model (solid histogram). Spectral components are shown as dashed, dotted, dotted–dotted–dotted–dashed lines. (Lower panel) Data/Model ratio.

Standard image High-resolution image
Figure 16. Refer to the following caption and surrounding text.
Standard image High-resolution image
Figure 16. Refer to the following caption and surrounding text.

Figure 16. EPIC-PN spectra of objects showing flat continuum and/or strong Fe–K emission.

Standard image High-resolution image
Figure 17. Refer to the following caption and surrounding text.
Standard image High-resolution image
Figure 17. Refer to the following caption and surrounding text.
Standard image High-resolution image
Figure 17. Refer to the following caption and surrounding text.
Standard image High-resolution image
Figure 17. Refer to the following caption and surrounding text.

Figure 17. EPIC-PN Spectra of unabsorbed objects.

Standard image High-resolution image

We also used criteria using far-infrared fluxes at 90 μm, and HR4 > −0.1. The contribution from cold dust heated by stellar processes in the host galaxy is likely to be significant in the far-infrared band unless an AGN overwhelms the emission from the host galaxy. Therefore, a low ratio does not necessarily mean that X-ray emission is weak relative to the infrared because of the heavy absorption of an AGN. We thus expect the criteria using far-infrared are less efficient to select obscured AGNs compared to the MIR selection. On the other hand, the combination of X-ray hardness and a low X-ray/IR ratio provides us with a chance to find AGNs buried in starburst activity. There are 83 objects satisfying the conditions HR4 > −0.1, and not securely detected in the 18 μm band. The presence of AGNs absorbed by a column density greater than 1022 cm−2 is reported for 31 objects in the literature or in this paper. The values for 21 and 10 objects are in the range of cm−2 and larger than 1.5 × 1024 cm−2, respectively, where 3 objects classified as "Thick?" are regarded as Compton-thick. There is no indication of absorption in excess of 1022 cm−2 in 31 objects. No measurements are available for the rest of the objects (19 objects). Therefore, the detection rate of obscured AGNs using the far-infrared criteria is 31/62 = 50% for the sample with measurements.

The 22 objects analyzed in Section 4.2.4 show no clear evidence for the presence of an obscured AGN. This result apparently contradicts our selection using hard X-ray spectra measured by the hardness ratio of HR4 A possible reason for this contradiction is the faintness of the sources in the hard X-ray band. The net counts in the 4.5–12 keV for 15 among the 22 objects are less than 42 according to the 2XMMi catalog. The uncertainties of the hardness ratios are very large for such faint objects and sources with an unobscured spectrum could be chosen by our criteria. In one case (UGC 587), almost no photons are seen above 2 keV in our X-ray spectrum, but the 4.5–12 keV count in the 2XMMi catalog is 110 counts. This case is likely to be due to a combination of very low real X-ray counts and uncertainties in background estimation. The 4.5–12 keV counts for rest of the objects are in the range of 120–250 and their HR4 values are around 0.0, which is relatively soft among the objects selected by our criteria. The inspection of their spectra indicates that the HR4 values are reliable and consistent with the observed unabsorbed spectra within the errors.

5.2. Obscured AGNs Outside the Selection Criteria

While our selection criteria efficiently select obscured AGNs as discussed in the previous subsection, there are some obscured AGNs located outside our criteria in Figure 9. We examine the nature of these outliers using the 18 + 90 μm sample. None of the 29 objects absorbed by cm−2 are located in the region F18 < 1. Of the 26 objects in the range of cm−2, 9 have This ratio scatters from object to object, and the number of objects with F18 > 1 depends on the choice of the boundary. We set the criteria to efficiently select more absorbed objects (>1024 cm−2), which resulted in missing some moderately absorbed objects ( cm−2) in our criteria.

Seven out of 29 sources absorbed by cm−2 and 12 out of 26 absorbed by cm−2 are outside our criteria using 90 μm data. Inspection of Figure 9 (left and right) clearly shows that scatters in X-ray to infrared ratios are much larger if 90 μm data are used. A possible reason is that there is a wide range of the contribution of infrared emission from relatively cold dust heated by sources other than AGNs depending on the nature of host galaxies. While the scatters naturally worsen the success rate to find obscured AGNs, there is a higher probability of finding AGN activity hidden behind starbursts.

Among 55 sources with cm−2, 3 objects do not satisfy the condition HR4 In these objects (NGC 3690, HR4 = Mrk 1, HR4 = NGC 2623, HR4 = ), there is a considerable contribution from soft thermal emission to the band pass used to calculate HR4, where the 2.0–4.5 keV band is used as the soft band. Since their absorption column densities are above 1024 cm−2, their X-ray spectra of AGN component below 10 keV are reflection-dominated and the low-energy cutoff due to photoelectric absorption is not clearly seen. The combination of the significant contribution of soft thermal emission and the absence of absorbed direct emission results in the relatively small HR4 values. If starburst activity coexists with obscured AGNs, soft thermal emission from starburst contributes to X-rays. In the infrared band, emission from cool dust associated with star formation activity results in smaller X-ray to 90 μm ratios. Such cases tend to be missed if our criteria are applied.

5.3. Optically Elusive AGNs

There are several objects classified as an H ii nucleus showing a relatively large hardness ratio. In the 18 + 90 μm sample, the hardness ratios HR4 of nine objects (IRAS 01173+1405, NGC 695, NGC 3877, IRAS 12550−2929, IRAS 12596−1529, NGC 5253, IRAS 20551−4250, IRAS 23128−5919, and NGC 7738) are greater than The presence of an AGN is known in NGC 695, IRAS 12550−2929, IRAS 20551−4250, and IRAS 23128−5919 (Brightman & Nandra 2011a; Severgnini et al. 2012) and the latter three show significant absorption. A signature of the presence of a Compton-thick AGN is clearly seen in our spectrum of NGC 7738. All of these AGNs satisfy the conditions and

Fifteen objects classified as H ii nuclei in the 90 μm sample have HR4 Obscured AGNs are found in NGC 1402, SDSS J085312.35+162619.4, IRAS 13443+0802NE, and IC 5264. The former object is Compton-thick and the latter three are Compton-thin. Seven objects show no clear indication of the presence of an obscured AGN (Section 4.2.4; IRAS 10190+1322, Teng & Veilleux 2010; NGC 3314, Hudaverdi et al. 2006) and four are excluded from our analysis sample (Section 4.1). These AGNs with an optical classification of an H ii nucleus are a class of "optically elusive" AGNs. Examples of optically elusive AGNs have been reported from infrared-selected samples. Classical examples are the discovery of Compton-thick AGN in infrared bright starburst galaxies (NGC 4945, Iwasawa et al. 1993; NGC 6240, Iwasawa & Comastri 1998; Arp 299, Della Ceca et al. 2002). Maiolino et al. (2003) compiled a sample of non-Seyfert galaxies selected by infrared luminosity, infrared color, and the presence of a compact radio core. AGNs are found in at least 6 of 13 objects in their sample and AGN fraction becomes higher for more infrared luminous samples. Most of such elusive AGNs are found to be Compton-thick. Brightman & Nandra (2011a) found clear evidence for AGNs in four H ii-like objects in the 12 μm selected sample. Two of them (Arp 299 and ESO 148−IG002) are Compton-thick. Our results, together with these previous findings, demonstrate the efficacy of a combination of infrared and X-ray selection to find hidden AGNs in galaxies with significant star formation activity.

We examined infrared color () for the elusive AGNs in the 18 and 90 μm sample. The ratios are in the range from 0.038 to 0.067. This color is typical for H ii nuclei and much colder than usual Seyferts (Figure 11). The cold infrared color also supports that these AGNs are hidden behind star formation activity. Reliable intrinsic X-ray luminosities are available for Compton-thin objects. The logarithm of intrinsic luminosities in 2–10 keV are 41.6 and 42.3 for NGC 695 and IRAS 20551−4250, respectively. These luminosities are relatively low for Seyfert nuclei, and could be easily overwhelmed by star formation activity in optical and infrared.

6. SUMMARY

We cross-correlated 18 and 90 μm sources in the AKARI PSC and X-ray sources in the 2XMMi-DR3 catalog, and made a sample of infrared/X-ray-selected galaxies. As the parent sample of X-ray sources, we used objects located at and with EPIC-PN counts in 0.2–12 keV greater than 60 counts (60851 unique sources). Infrared sources at with reliable flux measurements at 18 μm and/or 90 μm (FQUAL = 3) are used. There are 43865 and 62326 18 and 90 μm sources, respectively, satisfying these criteria. The matched sample combining 18, 90 μm, and X-ray sources consist of 173 objects. Most of them are at a low redshift; the highest redshifts is 0.31 and 90% of objects are at a redshift smaller than 0.05.

The sample was divided into various activity classes and groups of various absorption column densities derived from X-ray spectra. Diagnostic diagrams using X-ray hardness (HR4) and X-ray CR to infrared flux density ratios were made using 173 objects in the matched sample of 18 μm, 90 μm, and X-ray sources. AGNs obscured by a column density greater than 1023 cm−2 are located in the lower right part of the diagrams HR4 versus F18 and HR4 versus F90. We selected objects in the region HR4 and without published X-ray results to search for obscured AGNs. An object not detected in the 90 μm band and having HR4 and was also selected as a candidate obscured AGN. We analyzed X-ray spectra of 48 objects in total after excluding Galactic sources, sources in complex fields, or sources with very low X-ray counts after data screening.

X-ray spectra of 26 among the 48 objects (54%) show clear evidence for the presence of absorbed AGNs. Sixteen objects (33%) show a continuum absorbed by a column density ranging from 3 × 1022 cm−2 to cm−2. Six objects (13%) show a strong Fe–K emission line and a flat continuum, indicating the presence of a Compton-thick AGN. The spectra of four objects are explained by either Compton-thin or Compton-thick AGNs. Twenty-two objects (46%) show no clear evidence for the presence of an obscured AGN. These objects are either very faint in hard X-rays or hardness ratio is modest (HR4 ≈ 0).

Reliable constraints on X-ray absorption are available for 151 among 179 objects, satisfying the conditions HR4 > −0.1 and ( or ). One hundred and thirteen objects show clear evidence for the presence of absorbed AGNs, resulting in the success rate of 75%. If only objects satisfying the 18 μm condition are used, the detection rate of absorbed AGN becomes 92%.

At least seven objects with an optical classification of an H ii nucleus show evidence for the presence of obscured AGNs, four of which are reported for the first time in this paper. These "optically elusive" AGNs have cold infrared color () typical for H ii nuclei. Their optical classifications and infrared colors are consistent with the idea that the star formation activity overwhelms their AGN in the optical and infrared wavelengths.

The authors thank an anonymous referee for constructive comments that improved the clarity of the paper. This research is based on observations obtained with XMM-Newton, an ESA science mission with instruments and contributions directly funded by ESA Member States and NASA, and AKARI, a JAXA project with the participation of ESA. This research made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration, and the HyperLeda database (http://leda.univ-lyon1.fr).

Facilities: XMM-Newton, AKARI

10.1088/0004-637X/814/1/11
undefined