Multi-wavelength Selected Compton-thick AGNs in Chandra Deep Field-South Survey

The starting point of the analysis is the X-ray sources in the CDFS survey. Straatman et al. (2016) provided photometric catalogs (from UV to IR; a total of 43 bands) for the CDFS survey. In order to obtain the photometric data of the X-ray sources in the CDFS survey, Guo et al. (2020) have cross-matched the X-ray catalog (Luo et al., 2017) and the photometric catalog (Straatman et al., 2016) with a matching radius of 1 arcsec. The matching results contained 839 X-ray sources. Based on the sample used in Guo et al. (2020), we further select our AGN sample using the following three criteria:

• 1)

  A source is classified as AGN by X-ray. This source can also be selected as AGN by its spectral energy distribution (SED, Pouliasis et al., 2020). That is to say, the confidence level of its AGN component, which is obtained by SED decomposition, is higher than 0.95.

• 2)

  The photometric data of the source needs more than 20 bands, and mid to far-IR band photometry is required (IRAC, Spitzer/MIPS 24 µm, Herschel/PACS 100 µm, and 160 µm), especially.

• 3)

  We exclude merger sources identified by their optical images.

The first criterion ensures that each source is an AGN. The second criterion guarantees that each source has enough photometric data to run the SED fitting model reliably. The point spread function in mid-IR and far-IR is large. The merger sources are near each other. Thus, the flux density of the merger sources may be overestimated at mid-IR and far-IR bands. To be accurate photometric data, we implement the third criterion. A final AGN sample (51 AGNs) is constructed, including eight known CT AGNs (XID 284, 332, 355, 405, 419, 587, 666, and 739), which are identified by the method of X-ray spectroscopy fitting (Liu et al., 2017; Li et al., 2019; Corral et al., 2019).

The X-ray emission of the radio-quiet AGN is believed to arise from the corona above the accretion disk, and it is predominantly produced by the inverse Compton scattering of photons from the accretion disk. Since the X-ray has strong pierced, the absorption-corrected X-ray luminosities with different correcting models for most AGNs are consistent. However, for CT AGN, there is a significant difference with differently correcting models (e.g., Zappacosta et al., 2018). In some studies (e.g., Luo et al., 2017), the corrected luminosities are used to represent the intrinsic X-ray luminosities of AGNs.

Table 1 shows our sample and the data used in this work. Column 5 of Table 1 lists the uncorrected 2–10 keV luminosities, which are converted by apparent 0.5–7 keV luminosities from Luo et al. (2017)²²2The 0.5–7 keV luminosities convert to 2–10 keV luminosities: ${\rm L}_{2\textendash 10}={\rm L}_{0.5\textendash 7}({\rm ln}10-{\rm ln}2)/({\rm ln}7-{\rm ln}0.5),\Gamma=2$; ${\rm L}_{2\textendash 10}={\rm L}_{0.5\textendash 7}(10^{-\Gamma+2}-2^{-\Gamma+2})/(7^{-\Gamma+2}-0.5^{-\Gamma+2}),\Gamma\neq 2$. Where $\Gamma$ is photon index., where they estimated absorption by assuming that the intrinsic power-law spectrum had a fixed photon index of 1.8. For sources with effective photon indices greater than 1.8, the absorption column densities were set to zero. Using the estimated column densities, Luo et al. (2017) computed the absorption-corrected 0.5–7.0 keV luminosities. The absorption-corrected 0.5–7.0 keV luminosities also convert to 2–10 keV luminosities listed in Column 6 of Table 1. Liu et al. (2017) performed an X-ray spectral analysis for the bright sources in the CDFS using the wabs * (zwabs*powerlaw + zgauss + powerlaw + zwabs*pexrav*constant) model, and provide the absorption-corrected 2–10 keV luminosities listed in Column 7 of Table 1. To search for a best fit, Li et al. (2019) adopted two models to fit the source spectra. Column 8 of Table 1 lists the absorption-corrected 2–10 keV luminosities provided by Li et al. (2019). Corral et al. (2019) selected 20 CT AGNs using the automated spectral analysis. Column 9 of Table 1 lists the absorption-corrected 2–10 keV luminosities of 3 CT AGNs, which are obtained by Corral et al. (2019).

Moreover, this work also uses the X-ray spectra of the eight CT AGN candidates. Their X-ray spectra are the merged spectra for which the 102 observations are matched to an identically astrometric frame. In this work, we use the X-ray spectra produced by Luo et al. (2017). For more detailed data reduction, please see Section 2.2 of Luo et al. (2017).

The photometric data from UV to IR can constitute a multi-band SED, which is composed of multiple thermal emission components. Thanks to SED fitting technique, we can perform a detailed multi-band SED decomposition, including the stellar radiation, the dust re-radiation, and AGN emission. SED fitting technique can also constrain some crucially physical quantities of both AGNs and host galaxies, such as AGN luminosity, AGN fraction, SFR, and stellar mass (M_∗). Guo et al. (2020) obtained some physical quantities (including AGN luminosity, AGN fraction, SFR, M_∗, dust luminosity, AGN type, and rest-frame 6 µm luminosity for AGN component) of all AGNs in our sample through SED fitting using photometric data from UV to IR. Therefore, some physical quantities used in this work are from Guo et al. (2020).

The mid-IR emission (3–30 µm) of an AGN is most distinct from that of its host galaxy. AGNs are usually bright in the mid-IR waveband because of the thermal emission from hot dust in the torus, which is heated by absorbing UV and optical photons from the accretion disk. The optical depth is low at the mid-IR waveband, and therefore the mid-IR emission of the AGN is not strongly suppressed. In this paper, we use the 6 µm emission to represent the mid-IR emission of the AGN (e.g., Lutz et al., 2004; Chen et al., 2017), which is obtained by SED decomposition. Column 10 of Table 1 lists the 6 µm luminosities of AGNs, provided by Guo et al. (2020).

The mid-IR and X-ray emission of radio-quiet AGNs are both excellent tracers of supermassive black hole accretion power. Previous studies have investigated the correlation between the mid-IR and X-ray emission (e.g., Lutz et al., 2004; Lanzuisi et al., 2009; Gandhi et al., 2009; Fiore et al., 2009; Chen et al., 2017). For instance, Stern (2015) derived a relation for the radio-quiet AGNs across a wide range of luminosity (from local Seyfert galaxies with L${}_{X}\sim 10^{42}$ erg s^-1 to the most luminous quasars with L${}_{X}\sim 10^{46}$ erg s^-1), which is

where L(2–10 keV) is in unit of erg s^-1 and $x\equiv{\rm log}(\nu{\rm L}_{\nu}(6~{}\micron)/10^{41}$ erg s^-1). In Stern (2015), the unobscured and Compton-thin AGNs are in good agreement with this relation. The relation should also be appropriate for CT AGNs, while the X-ray luminosities of CT AGNs mentioned by Stern (2015) are much lower than their mid-IR luminosities. The reason is that the absorption-corrected X-ray luminosities of CT AGNs are more model-dependent. Thus their intrinsic X-ray luminosities are underestimated. Indeed, the method of mid-IR excess is suitable to pre-select CT AGNs whose intrinsic X-ray luminosities are strongly underestimated.

The above mentioned are all radio-quiet AGNs. For the radio-loud AGNs, enhanced X-ray emission is usually observed in radio-loud AGNs (e.g., Miller et al., 2011; Ballo et al., 2012) due to the contribution of the jets to the total X-ray emission. While the mid-IR emission of radio-loud AGNs (excluding blazars) is similar to that of radio-quiet AGNs. The above method of mid-IR excess can be also used to select CT AGN candidates of radio-loud. There are nine radio-loud AGNs³³3Radio-loud AGNs: XID 119, 175, 342, 587, 623, 746, 760, 844, and 867. in our sample, including 2 CT AGNs that are identified by this work. The radio-loud AGNs will less impact on the result of mid-IR excess selection.

The left panel of Figure 1 shows that the mid-IR versus absorption-corrected X-ray luminosities. The absorption-corrected X-ray luminosities are estimated with a simple model (see Section 4.4 in Luo et al., 2017). The black line represents the relation of Equation 1. We obtain the 1$\sigma$ uncertainty of the Stern (2015) relation with using the data in Tables of Stern (2015). The gray shaded area is 1$\sigma$ dispersion for the Stern (2015) relation (also see Zou et al., 2020).

The solid grey circles represent the eight known CT AGNs. The dashed line is 3$\sigma$ lower limit. Most of the AGNs locate near the black line with 1$\sigma$ area. However, there are still 12 AGNs below the 3$\sigma$ lower limit, four of which are the known CT AGNs. The remaining 8 AGNs are also likely to be CT AGNs (hereafter referred to as CT AGN candidates).

The right panel of Figure 1 shows that the rest-frame 6 µm versus various absorption-corrected 2–10 keV luminosities of 8 known CT AGNs and 8 CT AGN candidates. Except for three known CT AGNs, the remaining five known CT AGNs corrected by Liu et al. (2017), Li et al. (2019), or Corral et al. (2019) are near the black line with 1$\sigma$ area. The absorption-corrected luminosities of the three known CT AGNs are above 3$\sigma$ lower limit, suggesting that their absorption correction should be reasonable. However, among 8 CT AGN candidates, only three candidates (XID 623, 802, and 844) are corrected by Li et al. (2019). Their absorption-corrected luminosities are still below the dashed line, suggesting that their column densities may be underestimated. There are still five candidates that are not corrected by previous works (Liu et al., 2017; Li et al., 2019; Corral et al., 2019). To determine whether these eight candidates are CT AGNs, we will perform X-ray diagnostics, optical spectral diagnostics, and mid-IR diagnostics on these eight candidates.

The X-ray diagnostics is the most reliable identification method for CT AGN (e.g., Lanzuisi, 2017). Li et al. (2019) have fitted the X-ray spectra of source XID 623, 802, and 884 using MYTorus, but did not confirm them as CT AGN. We believe that it is difficult to obtain reliable evidence through X-ray spectroscopy fitting to prove that these eight candidates are CT AGNs. Therefore, we only intend to use the characteristics of their X-ray spectra, in this section, to search for supporting evidence.

We fit the X-ray spectra of these eight candidates in Xspec v 12.11.0 (Arnaud, 1996) using the Cash statistic that is more appropriate with the low count regime (Cash, 1979). Due to the limited counts of these eight candidates, we do not bin the spectrum because it might lose some information. In order to obtain the characteristics of the X-ray spectrum at low-energy (2–10 keV) and high-energy ($>$ 10 keV), we use phabs * (powerlaw + zgauss) to fit the rest-frame 2–10 keV spectra, and use phabs * powerlaw to fit the rest-frame above 10 keV spectra. The phabs accounts for the Galactic absorption, which is fixed at a column density of $8.8\times 10^{-19}\ {\rm cm}^{-2}$ (Stark et al., 1992). The characteristics of the X-ray spectra of these eight candidates are derived by fitting their spectra. Their properties and the characteristics of their X-ray spectra are listed in Table 2.

Three candidates (XID 341, 623, and 844) have a flat spectrum with $\Gamma<1$ at energies below 10 keV. There are four candidates with X-ray spectra above 10 keV. However, only the X-ray spectrum of the source XID 623 is relatively reliable and shows a characteristic of the prominent iron K$\alpha$ emission line with the equivalent width of $0.234$ keV. Therefore, only the source XID 623 is more likely a CT AGN in terms of its X-ray spectrum characteristics.

The torus of an AGN does not obscure the high-ionization narrow optical emission lines from the narrow-line regions (NLRs). As long as the unified scheme holds, narrow emission lines have been considered to be a useful indicator of the AGN luminosity. Therefore, there should be a correlation between the narrow emission lines and X-ray luminosities, and the relationship has been investigated (e.g., Heckman et al., 2005). In this section, we use the high-ionization narrow optical emission lines to identify whether these eight candidates are CT AGNs. We try to collect the optical spectra of these eight candidates from the literature (e.g., Szokoly et al., 2004; Mignoli et al., 2005; Inami et al., 2017; Herenz et al., 2017), and finally we only find two spectra of XID 640 and 889. The spectrum of the source XID 640 originates from Herenz et al. (2017), and XID 889 from Figure 6 (Source 175b) in Szokoly et al. (2004). There is only [O III]$\lambda 5007$ emission line detected in the spectrum of source XID 640. We fit the [O III]$\lambda 5007$ emission line and obtain its flux, $f_{\rm[O~{}III]}=1.57\times 10^{-17}$ erg s^-1 cm^-2. In addition, we estimate the flux of the high-ionization optical emission lines ([O III]$\lambda 5007$ and [Ne V]$\lambda 3426$) in the spectrum of source XID 889, i.e. $f_{\rm[O~{}III]}=8.93\times 10^{-16}$ erg s^-1 cm^-2 and $f_{\rm[Ne~{}V]}=1.26\times 10^{-16}$ erg s^-1 cm^-2, respectively. In this section, the observed X-ray luminosities are from Section 4.1.

Maiolino et al. (1998) provided a diagnostic based on the ratio between the rest-frame 2–10 keV and the [O III] luminosities. The left panel of Figure 2 shows that the ${\rm L^{obs}_{X}/L_{[O~{}III]}}$ ratio as a function of ${\rm N_{H}}$. The cyan shaded regions correspond to $\pm 1\sigma$ around the $<$X/[O III]$>$ ratio (Maiolino et al., 1998). Based on the information in the left panel of Figure 2, we can derive the column densities of these two candidates. There is a 68.3% probability that the column density of the source XID 640 is larger than ${\rm 4.0\times 10^{23}\ cm^{-2}}$. Similarly, the column density of the source XID 889 is larger than ${\rm 10^{24}\ cm^{-2}}$. Besides, Gilli et al. (2010) provided another diagnostic based on the $<$X/[Ne V]$>$ ratio. The right panel of Figure 2 shows that the ${\rm L^{obs}_{X}/L_{[NeV]}}$ ratio as a function of ${\rm N_{H}}$. Based on the information in the right panel of Figure 2, we can derive that the column density of the source XID 889 is ${\rm N_{H}>4\times 10^{24}\ cm^{-2}}$ with a 90% probability. Through the above optical diagnosis, we can identify that the source XID 889 is a CT AGN.

The column densities of AGNs can be estimated by comparing the X-ray to mid-IR luminosities (e.g., Lanzuisi et al., 2015b; Lansbury et al., 2015, 2017).

Figure 3 shows the distribution of observed X-ray versus 6 µm luminosities for our sample. Except that the X-ray luminosities of eight candidates (solid red stars) are from Table 2, the X-ray luminosities of the remaining AGNs are from Table 1. The black line represents the relation from Fiore et al. (2009). Assuming that the intrinsic power-law spectrum has a photon index of 1.8 in the 2–10 keV energies range, the dashed line indicates the relationship after being absorbed by ${\rm N_{H}=1.5\times 10^{24}\ cm^{-2}}$ gas⁴⁴4Absorbed model : phabs., which means that the sources below the dashed line should be CT AGNs. We find that the 7 known CT AGNs are below the dashed line, indeed. Among eight CT AGN candidates, seven of which are below the dashed line. More interesting thing is that these seven candidates seem to have a higher column density than the eight known CT AGNs. One exception is XID 844 sitting above the dashed line, suggesting that this candidate may be not a CT AGN, but it is still a heavily obscured AGN. The column density of this candidate is ${\rm 1.8^{+0.8}_{-0.6}\times 10^{23}\ cm^{-2}}$ obtained by X-ray spectroscopy fitting (Li et al., 2019), and it may be underestimated compared to the result of Figure 3 (the column density of XID 844 is about ${\rm 10^{24}\ cm^{-2}}$). Besides, we find a source (the blue star in Figure 3, XID 133) that is not selected as a CT AGN candidate, but it is below the dashed line. The column density of this source is ${\rm 1.46_{-0.45}^{+0.84}\times 10^{24}\ cm^{-2}}$ (Li et al., 2019), its upper limit is larger than ${\rm 1.5\times 10^{24}\ cm^{-2}}$. The column density of this source, which is estimated by comparing its X-ray to mid-IR luminosity, should be larger than ${\rm 1.5\times 10^{24}\ cm^{-2}}$. Combining these two cases, we can infer that the source XID 133 should be a CT AGN.

Luckily, we identify hitherto unknown 8 CT AGNs by the multi-wavelength approaches. The summary of the multi-wavelength diagnostics is listed in Table 3.

For the galaxies that host AGNs, the majority of the X-ray emission originates from the center AGNs, and a small amount of X-ray emission can be contributed by the X-ray binaries (XRBs) in their host galaxies. If the central AGN is a CT AGN, then the X-ray emission from the central AGN may be completely absorbed. In order to diagnose whether the X-ray emission of 8 new CT AGNs originates from central AGNs or the XRB populations, we compare the observed X-ray and the X-ray luminosities expected from XRB populations.

The X-ray luminosity from XRB populations is composed of low-mass X-ray binaries (LMXBs) and high-mass X-ray binaries (HMXBs). Previous studies have shown that the total X-ray luminosity from the LMXB and HMXB population is proportional to M_∗ and SFR of the host galaxy, respectively (e.g., Gilfanov et al., 2004; Lehmer et al., 2016). Lehmer et al. (2016) provided the following empirical relation:

where log $\alpha=29.30$, log $\beta=39.40$, $\gamma=2.19$, and $\delta=1.02$ for 2–10 keV, respectively.

Figure 4 presents the observed X-ray over XRB X-ray luminosities as a function of observed X-ray luminosities. The gray shaded area shows the expected $1\sigma$ dispersion of the derived ${\rm L_{obs}/L_{XRB}}$ (0.17 dex, Lehmer et al., 2016). There are two CT AGNs (XID 341 and 433) that locate in the gray shaded area, indicating that their observed X-ray emission may be entirely contributed by the XRBs in their host galaxies. The dashed line is ${\rm L_{obs}/L_{XRB}}=2$, indicating that the X-ray emission from the center AGNs is equal to that from the XRBs. The CT AGNs XID 640 and 802 are near the dashed line. However, their lower limits are lower than or equal to those expected from XRB populations. We can’t exclude the possibility that their X-ray emission is from the XRB populations. There are still four remaining CT AGNs (XID 133, 342, 623, and 889) that are above the dashed line, suggesting that the majority of their X-ray emission still originates from the center AGNs. However, the lower limits of XID 342 and 889 are below the dashed line. We cannot rule out the possibility that the emission of the XRB populations dominates the X-ray emission of these two AGNs. Moreover, the X-ray spectrum analysis in Section 4.1 seems to assess whether the observed X-ray emssion is produced by the AGNs. The measured photon indices of four CT AGNs (XID 342, 433, 640, and 802) are $\Gamma>3$. However, the generally measured photon index for AGN is between 1.4 and 2.6, suggesting that their observed X-ray emission should not be dominated by (or come from) the AGNs.

Combining the above information, we can infer the X-ray origin of these eight CT AGNs, and the conclusions are as follows: (1) The observed X-ray emission of two CT AGNs (XID 133 and 623) should be dominated by the center AGNs. (2) The observed X-ray emission of two CT AGNs (XID 341 and 433) may come from the XRB populations. (3) The observed X-ray emission of three CT AGNs (XID 342, 640, and 802) should not be dominated by the AGNs. (4) There is still a CT AGN (XID 889) whose X-ray emission is dominated by central AGNs with a strong possibility. However, it cannot exclude the possibility that the emission of the XRB populations dominates its X-ray emission.

In Section 5.1, we have demonstrated that the X-ray emission of the two CT AGNs (XID 133 and 623) mainly originates from the central AGN. The column density of XID 133 obtained by the X-ray spectrum (Li et al., 2019) is reasonable within the error range (see Section 4.3 for detailed). While the XID 623 is a high-redshift source (z = 3.88), its rest-frame energies are mainly above 10 keV. One reason may be that the X-ray spectra above 10 keV are not significantly absorbed. Moreover, the XID 623 is a radio-loud AGN, so its jet will provide a part of the X-ray emission. Previous studies (Luo et al., 2017; Li et al., 2019) did not consider the contribution of the jet, thus underestimated its column density.

The X-ray emission of two CT AGN (XID 341 and 433) might originate from the XRBs in the host galaxies and suggested that their X-ray emission originated from AGNs was entirely absorbed. Therefore, the column densities of the center AGNs cannot be obtained by fitting their X-ray spectra. Similarly, the column density of the three AGNs (XID 342, 640, and 802) cannot be also obtained, because their X-ray emission is not dominated by AGNs. However, Luo et al. (2017) found that the X-ray spectra of those five CT AGNs presented weak absorbed or non-absorbed. Thus the column densities of these five CT AGNs are severely underestimated.

The X-ray emission of XID 889 is uncertain whether it is dominated by the AGN. If the AGN does not dominate the X-ray emission, then the reason for its underestimation of column density should be similar to the above five CT AGNs. If the AGN dominates, its X-ray emission may be the soft scattered component of the polar region, while the X-ray emission from the nucleus is hidden. However, Luo et al. (2017) misunderstood this source as a weakly absorbing AGN. Thus it column density is severely underestimated.

In the deep field survey, only about 10% of AGNs can be identified as CT AGN by X-ray spectroscopy fitting, showing the efficiency is low. In our sample, eight (15.6%) CT AGNs have been confirmed by X-ray spectroscopy fitting. The seven known CT AGNs are also confirmed in this work (see Section 4.3 for details). Besides, we identify newly 8 (15.6%) CT AGNs by the multi-wavelength approaches. The multi-wavelength approaches of selecting CT AGNs have a higher efficiency than the X-ray spectroscopy fitting. However, the multi-wavelength approaches require that a source has multi-wavelength data and that its AGN component decomposed by the SED is reliable, so our sample is only a small part of the AGNs in the CDFS survey. In other words, the multi-wavelength approach cannot be applied to all AGNs in the CDFS survey. In our sample, 16 sources can be identified as CT AGN, i.e. the fraction of CT AGNs is 31.2%, which agrees well with the theoretical expectation (e.g., Ricci et al., 2015; Lansbury et al., 2017).

CT AGNs usually have a higher accretion rate than non-CT AGNs (e.g., Draper & Ballantyne, 2010). Here, we compare the Eddington ratio ${\rm\lambda_{Edd}=L_{bol}/L_{Edd}}$ of CT AGNs and non-CT AGNs in our sample. Where ${\rm L_{bol}}$ is the bolometric luminosity, ${\rm L_{Edd}}$ is Eddington luminosity which is related to BH mass (i.e., ${\rm L_{Edd}=1.3\times 10^{38}\ M_{BH}/M_{\odot}}$ erg s^-1).

The bolometric luminosity is the total luminosity emitted at all wavelengths by the AGN (e.g., Krawczyk et al., 2013; Duras et al., 2020). The AGN luminosity obtained by Guo et al. (2020) is the integrated luminosity from UV to IR. The luminosity from UV to IR far exceeds the luminosity of other wavelengths. Therefore, for Type 1 AGNs, we use the AGN luminosities instead of their bolometric luminosities. Due to Type 2 AGN emission in the UV/optical bands is known to be obscured, we estimate their bolometric luminosities by integrating the AGN emission component in the range 1–1000 µm and simply rescaling the result by a factor of 1.7 (Pozzi et al., 2007; Zappacosta et al., 2018). There are 6 AGNs with broad optical emission lines in their spectra. Therefore, we estimated the BH masses of these six AGNs with the broad emission lines (H$\beta$ or Mg II, Netzer & Marziani, 2010; Trakhtenbrot & Netzer, 2012). The BH masses of other AGNs are estimated to use M_BH–M_∗ relation with a scaling of 0.003 (Suh et al., 2020).

The left panel of Figure 5 presents the Eddington ratio versus redshift for our sample. Except for 3 CT AGNs (XIDs 133, 623, and 802), the Eddington ratios of the remaining AGNs are lower than 1. Interestingly, these three CT AGNs of super-Eddington accretion are all newly identified in this work. The multi-wavelength approaches seem to prefer the selection of CT AGN with a high Eddington ratio. The right panel of Figure 5 shows the distributions of the Eddington ratio for CT AGNs and non-CT AGNs. We use the Kolmogorov–Smirnov test (KS-test) to examine their distributions of Eddington ratio: the p-value is 0.010, which suggests that their distributions are different. The Eddington ratios of CT AGNs are significantly higher than those of non-CT AGNs. Our results support that the BH masses of CT AGNs are more rapid growth than those of non-CT AGNs and that CT AGNs may be a rapid growth phase of BH masses in the evolutionary scenario of AGNs (Goulding et al., 2011).

Some studies suggested that the occurrence of CT AGNs is related to the properties of their host galaxies (e.g., Ricci et al., 2017). Goulding et al. (2012) pointed out that the host galaxies of CT AGNs in the nearby universe show a high level of star formation. Therefore, we compared the SFRs and stellar masses of the host galaxies of CT AGNs and non-CT AGNs in our sample.

Figure 6 presents the relation between SFRs and stellar masses for the host galaxies. We also use the KS-test to examine their stellar mass distributions, and the p-value is 0.87. We repeat the KS-test to examine their SFR distributions, and the p-value is 0.97. These suggest that the properties of CT AGN host galaxies are similar to those of non-CT AGN host galaxies in our sample. Our results are against the claim of Goulding et al. (2012), and the main reason may be that most of the sources in our sample are high redshift AGNs. The environment of AGN host galaxies at high-redshift may be different from the local universe. In addition, our sample contains fewer CT AGNs, we cannot rule out our results are biased in sample selection. In subsequent work, we will expand the CT AGN sample and further study whether there are differences in the host galaxies of both CT AGN and non-CT AGN.