BASS. XXXIV. A Catalog of the Nuclear Millimeter-wave Continuum Emission Properties of AGNs Constrained on Scales ≤ 100–200 pc

We reconstructed mm-wave continuum emission images from the ALMA data using the Common Astronomy Software Applications package (CASA; McMullin et al. 2007). The raw data were reduced and calibrated by running scripts provided by the ARC in CASA with the versions adopted to create the scripts. For the rest of the analysis, we used CASA v.6.1.0.118. By considering only channels devoid of any strong emission lines, continuum emission was imaged not only for each spectral window, but also by combining all available spectral windows. In these imaging processes, we utilized the task tclean along with the deconvolver clark and the Briggs weighting method employing robust = 0.5. clark was adopted rather than the often-used alternative one hogbom, because the former properly subtracts clean components in the visibility space and also is faster, having enabled us to reconstruct images efficiently. The robust value does not offer the highest resolution nor the best sensitivity, but we opted for it. This is because we were able to preserve nearly the highest resolution while retaining a sensitivity close to the case of robust = 2.0, which provides the lowest noise level. However, if no significant emission was detected from a target, we reanalyzed its data adopting robust = 2.0. The details are described later. If data were obtained in the mosaic observation mode, we set gridder to mosaic. The cell size was set so that cell $\lt 1/3\times {\theta }_{\mathrm{beam}}^{\min }$, where ${\theta }_{\mathrm{beam}}^{\min }$ is the FWHM of the beam size along the minor axis. The threshold parameter was also set so that 3σ_mm < threshold < 4σ_mm, where σ_mm is the noise level derived from a region free from mm-wave emission. Finally, a primary-beam correction was applied to all reconstructed images. Here, we mention that we exceptionally performed a self-calibration process for NGC 1068 and Circinus galaxy, as fringe-like features were clearly left in the images without this calibration. According to the ALMA technical handbook, ⁴⁶ flux densities measured in Band 6 can have a systematic uncertainty of 10%, and this is taken into consideration throughout this paper.

To identify mm-wave emission possibly associated with an AGN in each constructed ALMA image, we first created a list of AGN positions (R.A. and decl.) using data and catalogs at different wavelengths. As X-ray emission is one of the best probes for AGN activity, we primarily relied on 3–7 keV images created by using Chandra data, which provide the highest resolution in the X-ray band. For 56 objects, we found on-axis Chandra data where the separation angle between the target and the focal plane is less than 1'. According to the Chandra X-Ray Center, ⁴⁷ the positional accuracy is 14 at the 99% level for such on-axis observations. Therefore, for each target, we searched for a mm-wave peak within the 14 radius around a Chandra-based AGN position. As thresholds of 3–4σ_mm were adopted for the cleaning processes, only emission whose flux density per beam was greater than 5σ_mm was considered a detection.

For the 42 objects that did not have Chandra data, we considered the IR (≈3–22 μm) ALLWISE catalog (Wright et al. 2010; Mainzer et al. 2011), which includes all our targets. In the IR band, the AGN position can be inferred from the observation of dust emission due to an AGN (e.g., Gandhi et al. 2009; Wright et al. 2010). The positional accuracy of the catalog was estimated to be ≈03 by crossmatching with the Two Micron All Sky Survey (2MASS) catalog, ⁴⁸ but we adopted 15 as a search radius for Wide-field Infrared Survey Explorer (WISE) positions. This is because the 2MASS accuracy for a peak, or the nucleus, of a spatially resolved galaxy at the 99% limit was estimated to be ∼15 (She et al. 2017).

We adopted the mm-wave peaks identified by Chandra and WISE, basically in this order of priority, as candidates that are associated with AGNs. Further details can be found in Paper I. As a result, we identified significant (>5σ_mm) nuclear peaks for 89 AGNs, corresponding to a high detection rate of ≈91%.

For the nine objects without significant mm-wave peak detections, we performed the cleaning process with robust = 2. As a result, only for IC 4709, significant emission with ${S}_{\nu ,\mathrm{mm}}^{\mathrm{peak}}/{\sigma }_{\mathrm{mm}}=7.2$ was identified (${S}_{\nu ,\mathrm{mm}}^{\mathrm{peak}}/{\sigma }_{\mathrm{mm}}=4.9$ for robust = 0.5). By considering this detection, mm-wave emission was detected from 90 AGNs (i.e., the detection fraction is 92%). As the last attempt to detect emission from the eight remaining objects, we proceeded to the analysis of alternative subarcsecond-resolution data. Eventually, even with the newly adopted data, no significant signal was found for the eight objects.

In Table 1, we list the positions of the identified mm-wave peaks as AGN positions. For the objects without significant mm-wave detections, we list AGN positions at different wavelengths. Figure 5 of Appendix B summarizes the 98 images centered at these AGN positions.

The high-resolution ALMA observations (≲06, or ≲100–200 pc) can reveal mm-wave emission that cannot be explained by a central unresolved source. The capability can be seen in Figure 1, where NGC 526A appears to show only an unresolved component, while NGC 3281, NGC 3393, and IC 5063 have other components. In the following, we refer to such resolved emission as nonnuclear emission, and in the next subsection, we describe our analyses for identifying nonnuclear components.

Zoom In Zoom Out Reset image size

3.3.1. Quantitative and Systematic Identification of Nonnuclear Millimeter-wave Components

We quantitatively identified nonnuclear components that extended from central unresolved sources by deriving radial surface brightness profiles. Specifically, we defined concentric annuli centered on the nuclear emission peak. The radial width of each annulus was set to half of the average beam size of ${\theta }_{\mathrm{beam}}^{\mathrm{ave}}$ = ${({\theta }_{\mathrm{beam}}^{\min }\times {\theta }_{\mathrm{beam}}^{\mathrm{maj}})}^{1/2}$, and the outer radius was set to $15\times {\theta }_{\mathrm{beam}}^{\mathrm{ave}}$. We then derived the averaged surface brightnesses (${S}_{\nu ,\mathrm{mm}}^{\mathrm{ave}}$ in units of Jy beam^–1) for all annuli using the imstat task. To confirm whether mm-wave emission is an excess over a central unresolved source that extends by the point-spread function (PSF), we also calculated the surface brightnesses of the achieved synthesized beam for the same annuli. Then, for comparison, the brightness in the innermost circle was adjusted to the observed one. As examples, Figure 2 shows observed radial profiles and adjusted synthesized-beam ones for the four objects mentioned previously (i.e., NGC 526A, NGC 3281, NGC 3393, and IC 5063). The profiles for all 98 AGNs are displayed in Figure 6 in Appendix B. For NGC 526A, no significant excess is found, consistent with the visual inspection; nonnuclear components do not exist. In contrast, the profile for NGC 3281 infers the presence of significant extended emission with respect to the unresolved component above three times the standard error (SE). We used the obtained radial profiles for identifying such emission extending from the central unresolved component, and labeled AGNs for which significant extended components were confirmed with an "e." Besides the extended emission, blob-like emission, which we define as emission separated from the central emission, can be seen in some objects. Clear examples are NGC 3393 and IC 5063, as seen in Figure 1. The blob-like features of IC 5063 are significantly detected also in its radial profile around the range ${\theta }_{\mathrm{beam}}^{\mathrm{ave}}\sim 4$–6 (Figure 2), while the blob-like component of NGC 3393 becomes ambiguous in the radial profile around ${\theta }_{\mathrm{beam}}^{\mathrm{ave}}=2$–4. This is likely due to noise in the same annulus, and indicates that, in the radial profile, we can easily miss significant blob-like components. We thus identified blob-like components in an alternative simple way: an image with 5σ contours was created, and then any components that were isolated from the central emission enclosed by the 5σ_mm contour and were significant above 5σ_mm were systematically regarded as blob-like components. We labeled AGNs for which blob-like components were found with a "b."

Zoom In Zoom Out Reset image size

With the two flags, e and b, NGC 3281, NGC 3393, and IC 5063, shown in Figure 1, were labeled with e, b, and b, while no flag was raised for NGC 526A. Of the 90 AGNs, 37 AGNs were quantitatively and systematically found to have nonnuclear emission, corresponding to ≈41%. The flags are tabulated in Table 3.

We warn readers that structures larger than an MRS are resolved out and therefore there should be nonnuclear emission that we missed. Indeed, we confirmed that the size of each identified nonnuclear component is less than the MRS of the data used for the identification.

While the identification of the nonnuclear components depends on whether there are actually nonnuclear components and also on the MRS as described above, some other observational properties can be factors that affect the identifiability. As possible properties, we investigated two parameters: the detection significance for peak emission and the physical resolution. The former was considered because we have defined peak emission as the nuclear component, and therefore nonnuclear emission around that (if any) is basically fainter and would not be detected for sources detected at low significance. The latter was based on the simple fact that a coarse resolution makes it difficult to resolve emission components. To assess these hypotheses, we created two samples: one consists of the 53 AGNs without any morphological flags and the other includes the other 37 AGNs with significant nonnuclear components. We refer to these samples as Sample-n and Sample-eb, respectively. Using the Kolmogorov–Smirnov test (K-S test), we then compared their detection significance and beam size distributions, as shown in Figure 3. The K-S test rejects their similarities with p-values of P_KS ≈ 7.9 × 10⁻⁴ and P_KS ≈ 2.3 × 10⁻⁷, respectively. Also, judging from the distributions in Figure 3, we confirmed that the Sample-n objects were detected at lower significances and were observed at coarser physical resolutions, as expected.

Zoom In Zoom Out Reset image size

3.4.1. Caution on Spectral Indices

3.4.2. Actual Derivation

The spectral index is necessary to derive the fundamental parameters of luminosity and flux. The k-corrected luminosity can be derived using the equation below (e.g., Novak et al. 2017):

$$\begin{eqnarray}&&\nu {L}_{\nu ,\mathrm{mm}}^{\mathrm{peak}}=\nu \displaystyle \frac{4\pi {D}^{2}}{{(1+z)}^{1-{\alpha }_{\mathrm{mm}}}}{\left(\displaystyle \frac{\nu }{\nu ^{\prime} }\right)}^{-{\alpha }_{\mathrm{mm}}}{S}_{\nu ^{\prime} ,\mathrm{mm}}^{\mathrm{peak}},\end{eqnarray} \tag{ 1 }$$

where ν is set to 230 GHz, i.e., the representative frequency considered in this study, and ${S}_{\nu ^{\prime} ,\mathrm{mm}}^{\mathrm{peak}}$ is the peak flux density per beam at the observed frequency of $\nu ^{\prime}$. Each flux density per beam was derived from an image consisting of all available spectral windows. By following the definition of the luminosity, the flux is defined as follows:

$$\begin{eqnarray}&&\nu {F}_{\nu ,\mathrm{mm}}^{\mathrm{peak}}=\nu \displaystyle \frac{1}{{(1+z)}^{1-{\alpha }_{\mathrm{mm}}}}{\left(\displaystyle \frac{\nu }{\nu ^{\prime} }\right)}^{-{\alpha }_{\mathrm{mm}}}{S}_{\nu ^{\prime} ,\mathrm{mm}}^{\mathrm{peak}}.\end{eqnarray} \tag{ 2 }$$

While the effect of the k-correction is ≈0.8% as the median value for the sample, it can be ≈5%–10% for six distant objects. This is comparable to the systematic uncertainty of ∼10%, and therefore the k-correction was applied uniformly.

To obtain the spectral index of α_mm and its error of Δα_mm necessary to calculate the luminosity and flux and their errors, we fitted a power-law function to the flux measurements from individual spectral windows based on the chi-square method. For 70 objects that were observed with more than two spectral windows and for which peak emission was detected in all windows, the spectral index and statistical error were estimated. To the errors, we added a systematic uncertainty of 0.2, which takes into account a possible flux calibration uncertainty between spectral windows at ∼230 GHz (Francis et al. 2020). The mean and standard dispersion of the indices of 68 sources whose indices were individually calculated are α_mm = 0.5 ± 1.2. Two objects, NGC 1068 and NGC 1194, were excluded from this calculation because their constraints could be unreliable, which is described in the paragraph after the next two; the actual indices assigned to the two objects to calculate ${L}_{\nu ,\mathrm{mm}}^{\mathrm{peak}}$ and ${F}_{\nu ,\mathrm{mm}}^{\mathrm{peak}}$ are also detailed therein. For the remaining 28 objects, the representative value of α_mm = 0.5 ± 1.2 was adopted to calculate their luminosities and fluxes.

We emphasize here that many objects, especially those with a low detection significance, have large uncertainties of ≥1 in their indices; the corresponding fraction is ∼40%. This is seen in Figure 4, where the indices are plotted as a function of signal-to-noise ratio of the nuclear mm-wave emission. Therefore, we caution that many of the estimates are not so reliable, and that additional data are important for more reliable estimates.

Zoom In Zoom Out Reset image size

In addition to showing the above fact, Figure 4 suggests an interesting result that the indices are well distributed around a mean value of ≈0.5. This suggests that the dust emission, for which a steep slope is expected (e.g., ∼−3 to ∼−4), is unlikely to be the dominant source, and flatter components, such as synchrotron and free–free emission, are likely to be more important.

Returning to NGC 1068, we find that its spectral index (α_mm = 1.5 ± 0.3 around 270 GHz) is largely different from an index of α_mm ∼ −1.3 expected from a mm-wave band SED constructed by Inoue et al. (2020), and also is a few sigma higher than the index more carefully derived by Michiyama et al. (2023) (i.e., α_mm ∼ 0.5–0.8). The latter study constructed an SED denser than that of Inoue et al. (2020) by using high-resolution (≈005) ALMA data. First, as a possible cause, we considered that this discrepancy might be ascribed to our data having a narrower spacing between the lowest and highest frequency spectral channels than those for the others (i.e., ≈5.4 GHz with respect to ∼14–20 GHz), except for NGC 1194. The object was also observed with a similar narrow spacing (≈4.9 GHz). To estimate the index better, we analyzed the second-highest-resolution data with ${\theta }_{\mathrm{beam}}^{\mathrm{ave}}\approx 0\buildrel{\prime\prime}\over{.} 018$ and a wider spacing of ≈19 GHz. Then, a spectral index was constrained to be α_mm = 0.4 ± 0.2 around 260 GHz. We additionally analyzed Band 7 data with comparable resolution (≈0018) at 331 GHz (group OUS ID = A001/X1465/X212e in 2019.1.00854.S). We then found that the interband spectral index between the Band 6 and Band 7 frequencies (256 GHz and 331 GHz) also inferred α_mm ≈ 0.4. The two measurements disfavor the result of Inoue et al. (2020) but are virtually consistent with the updated one of Michiyama et al. (2023). Thus, the true spectral index is possibly ∼0.4, and it is suggested that the spectral index derived from the narrow spacing has a large systematic uncertainty. Eventually, for NGC 1068, we conservatively adopted the representative value of α_mm = 0.5 ± 1.2, consistent with the possible value of 0.4. For the remaining object NGC 1194, as this did not have other subarcsecond ALMA data, we also adopted the same representative value.

In addition to the narrow spectral coverage suspected as a cause of the unreliable constraint on the spectral index, we checked whether there were objects whose estimated indices could be unreliable due to observations with smaller aggregate bandwidths of <3.75 GHz. As a result, it was confirmed that no objects were observed at such small aggregate bandwidths and have estimated indices.

Lastly, we attempted to improve the constraints on the spectral indices with large uncertainties of Δα_mm ≥ 1. For the 27 objects with Δα_mm ≥ 1, we searched for different subarcsecond-resolution Band 6 data, and found that only NGC 4388 had available data. With the data, the spectral index was constrained to be α_mm = −0.5 ± 0.8 around 220 GHz, while that obtained with the highest-resolution data was α_mm = 1.0 ± 1.0 at ∼260 GHz. This apparently large difference could be ascribed, in part, to a synchrotron self-absorption component having a peak around these frequencies, as suggested for nearby AGNs (e.g., Inoue & Doi 2018; Inoue et al. 2020). Anyway, the derived luminosities and fluxes from the two different data sets differ by only ≈0.1 dex and are consistent within 1σ.

With the indices estimated from the highest-resolution data, we calculated the luminosities and fluxes. To their errors, we added a systematic uncertainty of 10%, as recommended by the ALMA technical handbook. Table 3 in Appendix A provides the mm-wave properties obtained for all objects in our sample (e.g., spectral indices, luminosities, and fluxes).

The ALMA image of this object was taken at a resolution of 005, revealing a patchy structure around the bright nuclear component within a radius of 02. Based on a Very Large Array (VLA) observation at 5 GHz, Ekers et al. (1978) reported an unresolved component whose diameter size was constrained to be < 03. This resolution would be almost the highest ever achieved for this object, except for that of the ALMA observation, and thus there seem to be no data that help us to understand the patchy structure better.

Our 266 GHz image at a resolution of ∼002 shows diffuse emission around the nucleus (a in the corresponding panel of Figure 7; hereafter italic characters are used to indicate such components in the same way) and also three other diffuse components ∼03 away at a position angle (PA) ⁴⁸ of ∼10° (b), ∼06 away at a PA of ∼20° (c), and ∼02 away at a PA of ∼180° (d). A high-resolution continuum image obtained at almost the same resolution and frequency (i.e., ≈002 and 257 GHz) was presented by Impellizzeri et al. (2019). Their image revealed an X-shaped structure with a PA of ∼20° (blue dotted arrow) around the nucleus, and they interpreted this as being due to edge brightening of a bicone-like ionized region. The X-shaped structure is clearly seen also in our image. Impellizzeri et al. (2019) measured the mm-wave and radio fluxes within a 013 square area with their ALMA data and VLA 43 GHz data, and then derived a spectral index of ∼0.1 for the central emission. This is consistent with free–free emission (α_mm= 0.1). However, based on an SED covering a wider range of ∼1–1000 GHz, Michiyama et al. (2023) presented two other possible scenarios: self-absorbed synchrotron emission from an area with a radius of ≈30 Schwarzschild radii and synchrotron emission from a jet.

The other diffuse component of b appears to be closely located in an area with bright emission found in the radio band (∼1–30 GHz). For example, a 1.4 GHz component, labeled C in Gallimore et al. (2004), was found at a resolution ∼001 (see also Gallimore et al. 1996; Muxlow et al. 1996). However, it is seen that the diffuse mm-wave emission (b) may be closer to the center by ∼004 (i.e., 2 pc) than the radio emission. Further investigation is crucial to confirm this and, if true, to understand this offset. Similarly, the components c and d virtually appear to coincide with the radio-bright northeast area and S2, reported in Gallimore et al. (2004). In their paper, it was suggested that the northeast emission might originate in an internal shock or denser jet plasma, and the S2 emission may be the result of a jet-driven shock.

In addition to the radio emission, IR emission at 12.5 μm, for example, was mapped at a resolution of 05 within ∼1'' by Bock et al. (2000). They revealed a structure that extends from the nucleus in the north and south directions. Apparently, no peculiar MIR structures are associated with the structures b, c, and d.

Finally, we mention the secondary AGN recently suggested by Shin et al. (2021). The proposed area is ∼3'' away from the nucleus at a PA of ∼44° (i.e., it thus lies outside the cutout of the image shown). No counterpart in the mm-wave band is however found. Given the very low X-ray luminosity of ∼10³⁹ erg s⁻¹ estimated by Shin et al. (2021), the nondetection may be natural. Our ALMA data have a sensitivity down to ∼10³⁷ erg s⁻¹ in the mm-wave band, which can be converted to a 2–10 keV luminosity of ∼10⁴¹ erg s⁻¹ according to a relation presented in Paper I. Thus, the current data would be just insufficient to assess the presence of the suggested secondary AGN.

This source was observed at a resolution of ≈01 using ALMA and the image shows central core emission (a) and two nonnuclear components to the southwest and the southeast, labeled b and c, respectively. A moderate resolution (≈03) image at 8.4 GHz was obtained with the VLA by Thean et al. (2000) and revealed a linear structure, which extends at a PA of 120° and has two peaks in the southeast and northwest regions, looking like radio lobes. The brightest radio component in the southeast direction appears to coincide with the mm-wave blob of c. No mm-wave counterpart is however found for the fainter northwest radio component. This could be because mm-wave emission is also fainter and thus cannot be detected at the achieved sensitivity. Regarding the component of b, no radio counterpart is found, perhaps due to the lower resolution of the radio image. Interestingly, the relative strength between the mm-wave and radio emission is highest in the center. This could suggest that the central and other lobe-like structures (c and maybe b as well) have different physical origins.

The image at a resolution of ≈02 shows seven significant components, labeled b, c, d, e, f, g, and h around the central component labeled a. Radio maps at different frequencies were presented in past studies (5 GHz–15 GHz; Sandqvist et al. 1982; Stevens et al. 1999; Thean et al. 2000; Sakamoto et al. 2007). At a, there appears to be a radio component according to a ∼1'' resolution map created by Stevens et al. (1999). Furthermore, the five mm-wave components of b, c, d, e, and f appear to have radio counterparts that have been labeled D, E, G, H, and F, respectively, in previous studies (e.g., Sandqvist et al. 1982). Sakamoto et al. (2007) suggested that each of the D, E, G, and H regions possibly hosts a supercluster. Stevens et al. (1999) suggested that the components have moderate to steep spectral indices and are consistent with nonthermal emission. Regarding region F, the interpretation is not so clear, as discussed in Stevens et al. (1999). It is also currently unclear why the g and h components have no radio counterparts.

Complex mm-wave structures appear within a radius of ∼3''. The central emission is slightly elongated to the north and is surrounded by several blob-like components. Combes et al. (2014) presented a submm-wave ALMA Band 7 image with a beam size of 05, comparable to our mm-wave data of 04. The authors argued that nonnuclear submm-wave emission exists and is from dust considering that the submm-wave emission follows a two-arm spiral structure detected in the CO(J = 3–2) line. Also, they estimated that the dust emission is likely to be stronger than synchrotron emission in the submm-wave band. However, our Band 6 continuum distribution appears to be different from the Band 7 one, suggesting that the Band 6 emission may have a different origin. Also, we mention the paper by Mezcua et al. (2015), which presented ≈01 Hubble Space Telescope (HST) optical and VLT near-infrared (NIR) data including K-band emission and H₂ line emission. Apparently, none of the distributions seems to be spatially similar to the Band 6 structures. Eventually, nothing is found probably relevant to the extended and patchy components.

This is an ultraluminous IR galaxy and is considered to be in the final stage of a galaxy merger (e.g., Ricci et al. 2017c, 2021). Around the central peak emission, diffuse emission within a radius of ∼01 is revealed at a resolution of 004. The encompassing area contains emission of ∼4 mJy, of which ≈1.3 mJy beam⁻¹ is ascribed to the peak emission. Thus, the contribution of the diffuse component is strong. Although there were few images taken at comparable resolutions, to our knowledge, NIR (2.2 and 3.8 μm) images at resolutions of 01–02 were obtained by Imanishi & Saito (2014), who utilized a camera of Subaru/IRCS in adaptive-optics mode. The obtained NIR nuclear images seem to be overwhelmed by only a point source. Using shorter-wavelength (≈1–2 μm) HST/NICMOS data at similar resolutions of 01–02, Scoville et al. (2000) also reported an unresolved component, but additionally mentioned the possible presence of low-level extended emission. To discuss more about whether the mm-wave diffuse emission has an NIR counterpart, higher-resolution images with better sensitivities are desired.

A core-like component appears to be surrounded by a halo, whose major axis is oriented at a PA of ∼140°. In the adjacent cm-wave band, a jet in a north–south direction was clearly revealed up to 4'' on each side (e.g., Ulvestad & Wilson 1983). The difference in angle between the radio jet and the mm-wave halo would suggest that the mm-wave emission would not be dominated by the jet. Instead, ionization cones traced by optical ionization lines (e.g., [O iii] and [S ii]+Hα) have similar angles to that of the halo (e.g., Mulchaey et al. 1994). Spatially resolved X-ray iron emission that could be due to X-ray irradiation of gas by the AGN also has a similar angle (Kawamuro et al. 2020, 2021). Thus, the mm-wave halo could be more relevant to the AGN photoionization process than the cm-wave jet.

In addition to the central nuclear emission with ∼0.7 mJy beam⁻¹ (a), our image at a resolution of ∼05 unveils a comparably bright component (b) with ∼0.5 mJy beam⁻¹ ∼2'' away from the nucleus in the south. An 8.4 GHz image was obtained at a similar beam size of ∼04 by Mundell et al. (2009) using the VLA, and shows radio emission (∼1 mJy) spatially consistent with a. On the other hand, there is no significant radio emission around the south mm-wave component (b). The upper limit of the radio emission, corresponding to 3 × rms, is estimated to be ∼0.05 mJy beam⁻¹, leading to an upper limit of <−0.7 for the spectral index. This is smaller than that for the central emission (∼0.7). As an insight from different wavelength data, Ma et al. (2021) spatially revealed ionized gas at resolutions below 01. According to optical emission line diagnostics with the [N ii]/Hα and [O iii]/Hβ, AGN-like or LINER-like optical emission line ratios were found around b. Thus, the mm-wave emission in b may be due to the AGN.

Millimeter-wave continuum emission in the central 1'' × 1'' region of the object was discussed in detail by Alonso-Herrero et al. (2019), who used the same ALMA data as we used. Here, we give a brief summary of their discussion, and those who are interested in more details should check Alonso-Herrero et al. (2019) and references therein.

The authors suggested that there exist two mm-wave components with different PAs of ∼35° and ∼−20°, indicated in Figure 7. The centrally concentrated component with a PA ∼ 35° may be relevant to an ionization cone and a radio jet. Using HST observations of [N ii] and Hα, an ionization cone was mapped at a similar PA. The presence of a jet in the same direction, or synchrotron emission, was inferred from a spectral-index map that Alonso-Herrero et al. (2019) made by combining mm-wave data with additional submm-wave ALMA data.

The other component at PA ∼ −20° (a) is located 05 away from the nuclear peak, and the angle almost aligns with the axis of a nuclear disk. mm-wave to submm-wave spectral indices therein were found to be consistent with thermal processes and free–free emission. While considering this fact and other multiwavelength data (e.g., Brγ), Alonso-Herrero et al. (2019) discussed that SF may contribute to the mm-wave emission. In the direction of the mm-wave component, cm-wave radio emission was also revealed by Mundell et al. (1995) and could originate from supernova remnants in the above context.

A northwest–southeast linear structure, parallel to a PA of ∼100°–130°, can be seen. In contrast, Schmitt et al. (2001) reported only unresolved radio (8.46 GHz) emission with a size upper limit of <60 pc based on a spatial resolution of ∼02. This is better than that achieved by the ALMA observation (05). As a spatially extended signature of AGN activity, a narrow-line region was mapped by using optical [O iii] emission in Schmitt et al. (2003), and this extends toward the northeast. Eventually, from the radio and optical observations, we cannot find mm-wave counterparts. Instead, it seems that the major axis of the galaxy disk has a similar angle of ∼130° (Schmitt et al. 2003). Thus, the linear structure could be associated with gas, or SF processes, distributed in the galaxy disk.

This object is known to have three radio spots in the northwest, nuclear, and southwest areas (e.g., Cooke et al. 2000; Koss et al. 2015; Maksym et al. 2017; Finlez et al. 2018), and our ALMA image at a resolution of 03 shows two significant components: a in the nucleus and b in the southwest direction. Thus, mm-wave emission spatially corresponding to the northeast radio spot is undetected. At a similar resolution of ∼02, Schmitt et al. (2001) estimated the 8.5 GHz flux densities for the southwest, nuclear, and northeast components to be ∼1 mJy, ∼1 mJy, and ∼8 mJy, respectively. On the other hand, their mm-wave flux densities per beam are ≲0.1 mJy beam⁻¹ (5σ_mm upper limit), ∼0.2 mJy beam⁻¹, and ∼0.4 mJy beam⁻¹, respectively. If the southwest and northeast radio components have similar spectral indices of 0.7–0.9, as inferred from the study of Morganti et al. (1999), the nondetection of the northeast component in the ALMA image is reasonable. In contrast, the nuclear component looks relatively bright compared with the southwest one (i.e., a higher ratio of mm-wave and radio flux densities). Indeed, the nuclear emission may have a spectral index of ∼0.5 according to the 8.5 GHz and 229 GHz flux densities. This may suggest different physical origins for the central and other components.

According to Maksym et al. (2019), a radio jet drives a shock that produces the X-ray emission and optical [O iii] emission. Moreover, the authors suggested that the shocked gas has enough energy to evacuate gas from the galaxy. Thus, the mm-wave emission (b) may also be tracing such a phenomenon, suggesting that mm-wave observations can be a tool for unveiling such shocked regions.

We note that Fabbiano et al. (2011) suggested the presence of another AGN in the southeast region. No significant emission is, however, found around that position in our ALMA image.

In addition to the central core emission, a north–south component appears to extend up to 003 on each side, and also there exist blob-like components. Data publicly presented to date seem to have achieved resolutions of ∼02 at best (e.g., Thean et al. 2000; Schmitt et al. 2001; Fischer et al. 2021). Thus, we cannot discuss the origin with the help of the other data.

Bright mm-wave emission is seen across the galaxy disk in the northeast–southwest direction. The flux density for the disk that we define as a 3'' × 14'' rectangular region with a PA of 43° is ∼600 mJy, and its intraband spectral index is constrained to be −2.1 ± 0.2. These are largely different from those at the central peak (a) of ${S}_{\nu ,\mathrm{mm}}^{\mathrm{peak}}\sim 40$ mJy beam⁻¹ and α_mm = −0.4 ± 0.2, indicating that the significant component is different between the central peak and the disk region. An SED in the 5–350 GHz band was analyzed by Bendo et al. (2016) with a focus on the central disk region within 12'' × 30'' (see also Henkel et al. 2018), and the 230 GHz continuum flux of 1.3 Jy was basically reproduced by combining thermal dust emission and free–free emission, resulting in an index of ∼−3. This index may be expected in our data as well, but is different from our constraint of −2.1 ± 0.2. This could be because thermal dust grains are more broadly distributed than free–free-emitting photoionized gas, as mentioned by Bendo et al. (2016), and the mm-wave slope is flattened in our high-resolution data where dust emission with a steep slope (α_mm = −4 to −3) is relatively missed.

At a PA of ≈140°–150°, a blob-like component (a) is revealed at a resolution of ∼01. Radio observations at 8 GHz and 1.5 GHz were performed (Nagar et al. 1999), but the resolutions were much coarser (a few arcseconds), and detected radio components were not spatially resolved. HST images were reported in some works (Ferrarese & Merritt 2000; Schmitt et al. 2003), and show that the host galaxy and also a narrow-line region traced by the optical [O iii] emission have similar PAs of −65°, slightly different from the PA of the mm-wave component. Thus, the origin could be due neither to phenomena in the main body of the host galaxy nor to the AGN, and currently the origin is unclear.

In our image at a resolution of 01, nuclear emission is prominent, and patchy components are seen. Radio (8.4 GHz) emission was imaged with a beam of 03 × 07 by Thean et al. (2000) and significant radio emission with 4.9 mJy beam⁻¹ exists around the nuclear mm-wave emission. However, for further discussion, as the radio image is very noisy, better radio images with higher sensitivities and also higher spatial resolutions are desired to be compared with the ALMA one.

Around the bright nuclear core (a), extended emission in the east–west direction (b) exists on a scale of a few parsecs. In addition, there appear to be two components: one located to the northwest (c) and the symmetrically distributed structure from north to south (d). In an area within a radius of 035 which includes all the above structures, the flux density is ∼30 mJy. Since the peak flux density per beam is ∼15 mJy beam⁻¹, an important fraction of the mm-wave emission within the area is resolved at our resolution of ∼0.5 pc.

Submm-wave ALMA Band 7 continuum emission was imaged at a resolution of ∼03 by Izumi et al. (2018). The resolution is coarser than achieved with our data (i.e., 002), thus hampering a visual comparison with our image. However, the authors reported the presence of an elongated structure on a scale of 005 (∼1 pc) at a PA of ∼300° (blue dotted arrow) based on visibility modeling. They suggested that this would trace the polar dust structure found by MIR observations (Tristram et al. 2014). Component c appears to coincide with that, thus perhaps indicating dust emission as the origin of c. The more prominent east–west emission of b could have a different origin, however, given its direction, and currently, it is unclear what its origin is. Finally, the origin of component d is also unclear due to the lack of meaningful data available. Each of the north and south components spreads on a scale of 1'' and has a flux density of ∼0.9 mJy. To determine the physical mechanism for each component, it is important to determine the index, but our ALMA data are not sufficient to constrain it well. Thus, additional data at adjacent different wavelengths or those obtained at higher sensitivities are desired.

At a ≈02 resolution, in addition to the bright central emission, diffuse emission is seen, particularly in the east direction, while there may be a west component as well. Exceptionally, dotted contours at 3σ_mm, which are formally insignificant but could be real, are presented. Orienti & Prieto (2010) presented three maps at 4.8 GHz, 8.4 GHz, and 14.9 GHz by analyzing VLA data, whose resolutions are ∼04, 02, and 06, respectively. Interestingly, while diffuse halo structures were revealed at the two lowest frequencies within a radius of ∼2'', the 14.9 GHz map shows that a component, extending to the east and west directions by ∼2'' on each side, is prominent. Its physical origin was not discussed in the paper. The 14.9 GHz structure looks similar to the mm-wave one at least in the east, although it is slightly tilted counterclockwise by ∼20°–40°. This suggests their association. Furthermore, given that the PAs of the 14.9 GHz and mm-wave components are similar to that of the galaxy disk, they may be related to gas in the galaxy disk. Probably, a higher inclination angle (i.e., edge-on view) to NGC 5506, where line-of-sight column densities of gas can be higher, would make it easier to identify emission relevant to the gas in the galaxy disk.

This is a member of a merging system with IC 4518B. The prominent core emission, which appears to have a tail-like structure with a PA of 200°–210°, is seen in our data at ≈02 resolution. To our knowledge, no high-resolution radio images have been published, while Condon et al. (1996) showed a 1.4 GHz image at a resolution ∼10''. As data taken at comparable resolutions, Asmus (2019) showed 12 μm and 18 μm images at resolutions of ∼04–05. The author revealed an elongated MIR component with a PA of 17° (blue dotted arrow), which is almost parallel to the mm-wave tail. Given the argument by Asmus (2019) that such an MIR component is due to dusty wind by an AGN, the mm-wave tail could be also driven by the AGN. As a different supportive result for the AGN-driven emission, Díaz-Santos et al. (2010) found extended [S iv] 10.51 μm line emission out to ∼08 to the south direction, and interpreted this as a sign of the presence of a narrow-line region, consistent with the AGN origin.

In addition to the central core-like emission, our high-resolution image created by a beam of 026 × 024 revealed a horn-like structure at a PA of ∼40°. However, as no meaningful data taken at comparable resolutions are available, discussion on its origin is currently difficult.

This is a well-studied merging system, and two active nuclei have been identified at various wavelengths (e.g., Komossa et al. 2003; Gallimore & Beswick 2004; Gallimore et al. 2004; Puccetti et al. 2016). According to the recent study by Kollatschny et al. (2020), there may be another inactive nucleus to the south of the active nucleus in the south. Our image at 003 resolution shows two mm-wave compact components: one at the center (a) and the other in the north (b), probably associated with the first and second active nuclei (e.g., Gallimore et al. 2004). On the other hand, insignificant emission is found around the third nucleus (c), as already confirmed by Treister et al. (2020), who analyzed the same ALMA data. Medling et al. (2019) suggested that within a radius of 0077 around each of the two active nuclei, thermal dust emission is dominant at 242 GHz. This argument was based on the result that the extrapolated mm-wave flux densities from 5 GHz ones by adopting a spectral index of ≈1 are lower than observed. In contrast, our higher-resolution data suggest a spectral index of α_mm ∼ 1.0 ± 0.4 for the nucleus (a), inconsistent with the dominance of thermal emission. Regarding the north nucleus (b), the spectral index is not constrained, as its mm-wave emission is not detected in individual spectral windows.

We comment on the relation of the mm-wave and X-ray luminosities for the two active nuclei. In Paper I, we added their luminosities together to be consistent with the X-ray flux derived in Ricci et al. (2017c), but we also confirm that the two nuclei follow a mm-wave-to-X-ray regression line (Table 1 in Paper I) within 2σ, even if we consider them separately. The used X-ray luminosities are taken from Puccetti et al. (2016), who derived intrinsic 2–10 keV luminosity for each using NuSTAR, Chandra, XMM-Newton, and BeppoSAX data.

Our ALMA image at a resolution of 03 shows an isolated signal at a PA of −50 to −40° (a). According to a subarcsecond MIR observation by Asmus et al. (2016), an extended MIR component possibly exists, and its PA was constrained to be 42° (see also Isbell et al. 2021). The PA does not seem to indicate that the mm-wave signal is associated with the MIR component. Instead, the host galaxy has a PA of 120°and thus seems aligned to the signal. To discuss further whether a host-galaxy phenomenon is indeed relevant, detailed analyses of host-galaxy images are necessary.

Three horn-like structures in the north, east, and south directions are seen in our mm-wave image constructed with a beam of 037 × 025. Images at comparable resolutions are not found. Additional observations are desired for discussing the nonnuclear components.

Our ALMA image produced at ≈04 resolution shows three compact components: one at the central AGN position (a) and the other two to the southeast and northwest (b and c). They are almost linearly aligned at an angle of ∼115°, which is similar to that found for linear structures at frequencies of ∼10–20 GHz. The mm-wave components were already discussed by Morganti et al. (2015) with a different ALMA data set at ∼230 GHz. Using 17 GHz radio data from Morganti et al. (2007), the authors constrained the spectral indices for the southeast, northwest, and central components to be 0.87 ± 0.02, 0.93 ± 0.01, and 0.33 ± 0.01, respectively. The resolutions of the used 17 and 230 GHz data were ∼04–05, comparable to ours. The indices favor nonthermal processes as the origin of the mm-wave components, and moreover the flatter slope of the central component could suggest that its origin is different from those of the southeast and northwest ones.

We moreover note that IC 5063 exhibits extended X-ray emission along the mm-wave and radio structures (Gómez-Guijarro et al. 2017). Recently, Travascio et al. (2021) performed a spatially resolved X-ray spectral analysis using Chandra and suggested that extended X-ray emission in the soft X-ray band (<3 keV) can be reproduced by either two photoionized gas models or a sum of photoionized and collisionally ionized gas models. Furthermore, hard-X-ray Fe xxv emission was detected at the northwest radio structure. On the basis of these facts, the authors suggested that the X-ray emission may be caused by interaction of a radio jet with the interstellar medium. Thus, the mm-wave emission seems to also trace the interaction regions.

Our mm-wave image at a resolution of ≈03 reveals two bright nuclear peaks (a and b). In addition, to the north, a compact component (c) is seen. Zhao et al. (2016) studied this object using not only ALMA Band 9 data at a similar resolution of ≈02, who observed CO(J = 6–5), but also other supplementary data at different wavelengths (e.g., optical and radio bands). They found a radio core near the southeast nucleus (a) using VLA 8.4 GHz continuum data provided in Thean et al. (2000), and also a higher 8.4 GHz-to-CO (J = 6–5) luminosity ratio than expected for SF regions therein. These support that an AGN is located at the southeast position of a. We note that a is consistent with a peak of optical emission therein (González Delgado et al. 1998). Following the analysis and suggestion of Zhao et al. (2016), we defined the southeast nucleus (a) as an AGN position. On the other hand, Zhao et al. (2016) did not suggest the presence of an AGN in the northwest region (b), due to the absence of compact radio emission and a SF-region-like radio-to-CO (J = 6–5) ratio in the area. We note that we created a high-resolution Chandra image (≈01) in the 3–7 keV band, probably tracing AGN emission (Levenson et al. 2005), and we did not find clear double nuclei.

Lastly, we mention that the Band 9 emission around the north mm-wave component (c) was attributed to dust emission by Zhao et al. (2016). The peak emission is ≈10–20 mJy beam⁻¹, and if a thermal component with α_mm = −3.5 is dominant, the predicted flux at 262 GHz (our mm-wave frequency) is ∼0.5 mJy beam⁻¹. This is consistent with our observed value of 0.5 mJy beam⁻¹, suggesting that the Band 6 emission would also be dust emission.

Our mm-wave image at a high resolution of 006, corresponding to ∼30 pc for the object, shows that overall the morphology looks diffuse. This can be noticed more clearly by considering fainter emission at 3σ_mm (dotted lines) additionally. A 5 GHz image was presented at a much coarser resolution of ∼14 by Lonsdale et al. (1992), and the radio emission has a core-like morphology. This radio emission seems to coincide with the peak of the mm-wave image. However, to confirm whether most of the radio emission is indeed associated with the mm-wave peak or appears like the diffuse mm-wave emission, it is important to obtain a higher-resolution radio image.

In addition to the central core emission, diffuse emission in the east–west direction appears at 3σ_mm, although this is formally insignificant. The morphology appears to trace the galaxy disk (e.g., Malkan et al. 1998). Therefore the diffuse emission would trace gas in the disk, or SF regions. As in NGC 5506 (Appendix C.17), the edge-on view to NGC 7172 would enable us to identify that emission. We note that a possible radio counterpart for the core was found with a 04 beam in Thean et al. (2000).

Around the strongest central component (a), there are three blobs (b, c, and d). These are likely to be encompassed by an annular starburst ring with radii of ∼15–25 found at different wavelengths (e.g., Soifer et al. 2003; Díaz-Santos et al. 2007; Orienti & Prieto 2010), and are spatially coincident with significant 349.7 GHz components presented at ≈050 × 040 resolution by Izumi et al. (2015). At that 350 GHz band, the three components have flux densities per beam of ≈1.0–1.7 mJy beam⁻¹ (Izumi et al. 2015). With Band 6 ALMA data, different from ours, Imanishi et al. (2016) also measured their flux densities at ∼260–266 GHz with a similar beam size of 057 × 052. By correcting for the slight difference in beam size, the spectral slopes between the two bands are estimated to be ≈−2.6. This suggests that thermal emission with α_mm ∼ −3.5 may not be the only component. We note that, in our mm-wave data, the three components were not detected in a sufficient number of windows to derive the spectral index.

Our mm-wave image reveals nuclear emission (a) and also five surrounding components labeled b, c, d, e, and f . Hints on the origins can be obtained by comparing the ALMA image with the MIR [Ne ii] emission line image of Wold & Galliano (2006). The MIR image revealed six areas with bright MIR emission, labeled M1, M2, M3, M4, M5, and M6. The mm-wave components of b, c, d, e, and f appear to coincide with M1, M2, M3+M4, M5, and M6, respectively. The authors described that the six MIR structures may be due to young star clusters embedded in dust and, otherwise, to clusters that are older and obscured by a screen of dust. However, the recent study by Ricci et al. (2018) suggested that AGN-driven jets/winds could be related to the MIR structures, based on their associations with NIR emission lines (i.e., [Fe ii] and Brγ), while the contributions of young stellar clusters were not completely excluded. Thus, the mm-wave components may originate from AGN jets/winds.