Detection of a Superluminous Spiral Galaxy in the Heart of a Massive Galaxy Cluster

In Figure 1, we present the 0.7–2 keV XMM-Newton images of the superluminous disk galaxies and their large-scale environments. The X-ray images are background-subtracted and exposure-corrected, and include data from both the EPIC-PN and EPIC-MOS detectors. Five X-ray images reveal large-scale extended X-ray emission, which originates from massive galaxy clusters. The X-ray images associated with SDSS J113800.88+521303.9 and SDSS J093540.34+565323.8 do not show extended emission within the studied regions. This implies that these galaxies do not reside in the vicinities of galaxy clusters. While the ROSAT X-ray images showed excess emission associated with these galaxies, this emission was likely associated with bright point sources (e.g., AGNs) that were resolved and removed from the XMM-Newton images. Because the detection significance in the ROSAT images was rather low, it is also possible that the excess X-ray emission in the ROSAT data is due to an upward fluctuation. The properties of the detected galaxy clusters are discussed in Section 4.5.

Zoom In Zoom Out Reset image size

To determine the centroid of the galaxy clusters, we utilize the background-subtracted X-ray images (Figure 1) and apply Gaussian smoothing with a kernel size of 6''. The peak of the X-ray emission on the smoothed image defines the cluster center. We then measure the offset between the position of the superluminous disk galaxy and the cluster center. This offset is the projected distance between the superluminous disk galaxy and the center of the cluster's gravitational potential.

For one galaxy, J16273, the projected distance from the cluster center is only 24, which corresponds to 9.6 kpc at the redshift of the galaxy (z = 0.2599). This offset is comparable to the typical offset between the X-ray peak of galaxy clusters and the position of BCGs (Zhang et al. 2011). Additionally, this small offset may also be caused by projection effects. The Legacy Survey image of the galaxy is presented in Figure 2. Visual inspection of this image indicates that all galaxies in the proximity of J16273 are less luminous. To qualitatively study the neighboring galaxies around J16273, we performed an environment search using the NASA/IPAC Extragalactic Database (NED). We conservatively searched for galaxies within the redshift range of z = 0.23–0.29, considering the uncertainty of the photometric redshift of the neighboring galaxies. We identified six galaxies in this redshift range within a projected distance of 06 (145 kpc). By deriving the r-band luminosity of the galaxies, we found that the neighboring galaxies are factor of 4–630 less luminous than J16273. This implies that J16273 is the BCG of this galaxy cluster.

Zoom In Zoom Out Reset image size

The offsets between the superluminous disk galaxies and the centroids of the galaxy clusters are much larger for the remaining four systems. Specifically, for three clusters, the offset ranges between 1072–1621, which corresponds to a projected distance of 323–612 kpc. These large offsets exclude the possibility that the superluminous disk galaxies are the BCGs of the galaxy clusters. However, given that the redshifts of the disk galaxies and the galaxy clusters are similar, and the projected distance is smaller than the R₅₀₀ radius of the galaxy clusters, it is likely that these superluminous disk galaxies are physically associated with the galaxy clusters. We note that 2MASX J09572689+4918571 is associated with a merging galaxy cluster. Therefore, it cannot be excluded that this superluminous disk galaxy was the central galaxy of the northeastern subcluster, whose core is located at a projected distance of 60''(≈223 kpc) from the superluminous disk galaxy. For the remaining one system, 2MASX J14072225+1352512, the offset is 5646, which represents a projected distance of 2459 kpc. This offset is ≈3.3 times larger than the R₅₀₀ of the galaxy cluster, therefore it is unlikely that this galaxy belongs to the cluster. Instead, this galaxy is likely projected onto the same region of the sky.

To explore the physical properties of J16273, we reanalyzed its SDSS optical and WISE infrared data. The global stellar mass and star formation rate are estimated from the WISE mid-infrared photometry, where the short-wavelength bands of WISE are sensitive to the host stellar component and the longer wavelengths the star formation activity (Jarrett et al. 2013). Based on the calibration of Cluver et al. (2014), the W1–W2 color may be used to estimate the stellar mass-to-light ratio, and the light derives from the W1 in-band luminosity. The stellar mass therefore follows, with uncertainties (10%–20%) propagating from the calibration and WISE photometry. For the star formation rate, we use the calibration from Cluver et al. (2017), where the mid-infrared luminosity (the W3 and W4 bands) is tightly correlated to the total-infrared luminosity. The typical uncertainties, from the calibration and photometric measurements, are 20%–40%. Using the color–color relations in Jarrett et al. (2019), we note that the WISE colors W1–W2 = 0.01 ± 0.05 and W2–W3 = 1.42 ± 0.28 [Vega mag], essentially relating the stellar population to the star formation activity, and have values that indicate a transition between spheroid-dominated (e.g., triaxial early types) and intermediate disk galaxies (disk + bulge). The colors are consistent with the relatively intermediate-to-large bulge-to-total ratio—B/T = 0.49—that is inferred from the W1 radial profile using the decomposition detailed in Jarrett et al. (2019).

We also inspected the optical spectrum of the galaxy to search for potential AGN signatures. To this end, we utilized the SDSS DR12 optical spectrum of the galaxy and computed the standard optical diagnostic line ratios, [O iii]/Hβ, [N ii]/Hα, and [S ii]/Hα, which allow us to place the galaxy on the [N ii]–BPT and [S ii]–BPT diagrams (Baldwin et al. 1981). We obtained $\mathrm{log}([{\rm{O}}\,{\rm\small{III}}]/{\rm{H}}\beta )=-0.83$, $\mathrm{log}[{\rm{N}}\,{\rm\small{II}}/{\rm{H}}\alpha ]=-0.38$, and $\mathrm{log}[{\rm{S}}\,{\rm\small{II}}/{\rm{H}}\alpha ]=-0.21$. These values place J16273 below the star-forming locus, with no indication of an AGN.

In Section 4.2, we demonstrated that there is only one superluminous spiral galaxy in our sample that resides in the center of a galaxy cluster. Because our X-ray analysis identified each center of each galaxy cluster, we can search for the true central galaxies in the four other clusters. To identify these galaxies, we investigated SDSS observations around the cores of the clusters and searched for the optically brightest galaxies. We identified massive galaxies in the center of each cluster, which are shown in Figure 3.

Zoom In Zoom Out Reset image size

To probe the characteristics of the true BCGs, we first measured their u − r color indices, which hint at whether the galaxies are quiescent of star-forming. For the BCG galaxies, we found u − r = 3.18–4.01, suggesting that these systems are red and quiescent (Strateva et al. 2001). Their morphologies also indicate that these are massive elliptical galaxies. Thus, the other galaxy clusters detected in the XMM-Newton images host common massive ellipticals at their core.

Interestingly, the brightness of the central elliptical galaxies is comparable to or even lower than the luminosity of the superluminous disk galaxies. The superluminous disk galaxy 2MASX J14475296+1447030 is slightly fainter in the r band (r_disk − r_elliptical = 0.37 mag), but is brighter in the u band (u_disk − u_elliptical = −0.50 mag) than the elliptical galaxy (SDSS J144756.93+144532.6) at the cluster core. Surprisingly, 2MASX J09254889+0745051 and 2MASX J09572689+4918571 are both brighter in the five SDSS bands than the elliptical galaxies at the cluster cores, with r_disk − r_elliptical =−0.54 mag and r_disk − r_elliptical = −1.07 mag, respectively. Therefore, while the superluminous disk galaxies do not reside in the central regions of the galaxies, two of them (2MASX J09254889+0745051 and 2MASX J09572689+4918571) should still be considered as the brightest cluster galaxy of their clusters.

Using the XMM-Newton observations, we determined the properties of the galaxy clusters that were detected in the vicinities of the superluminous disk galaxies. To determine the best-fit temperature, X-ray luminosity, and total mass of the clusters within the R₅₀₀ region, we used an iterative process. Specifically, we computed the temperature in the region that had the highest signal-to-noise ratio (S/N). We then obtained the first estimate of the R₅₀₀ through the kT − M₅₀₀ relation established in Lovisari et al. (2020). Then we extracted a second spectrum, using the obtained R₅₀₀, to obtain a new cluster temperature. We continued this procedure until the R₅₀₀ converged, i.e., until the new R₅₀₀ did not differ by more than $0.1^{\prime}$ from the previous one. This region was used to measure the final best-fit temperature and luminosity. To determine the total mass of the cluster, we relied on the kT − M₅₀₀ relation (Lovisari et al. 2020).

To fit the spectra throughout this procedure, we used an APEC model, assumed Galactic column density (Willingale et al. 2013), and used the metallicity tables of Asplund et al. (2009). During the fitting process, the temperature, metallicity, and normalization were free parameters. The redshift of the cluster was fixed at the redshift of the superluminous disk galaxy. The best-fit ICM temperatures are in the range of kT = 2.5–4.7 keV. The R₅₀₀ radii of the clusters vary in the range of R₅₀₀ = 659–953 kpc. We also derived the X-ray luminosity of the clusters within this radius and obtained L_{0.1–2.4 keV} = (0.40–3.14) × 10⁴⁴ erg s⁻¹. The best-fit temperatures, luminosities, and inferred R₅₀₀ values are listed in Table 1.

We did not detect extended X-ray emission around two superluminous disk galaxies. While this nondetection suggests that these galaxies do not reside in luminous galaxy clusters, it cannot be excluded that they are members of galaxy groups. To explore this possibility, we derive 3σ upper limits on the X-ray emission around these two galaxies. We compute the upper limits within a 500 kpc region, by assuming an optically thin thermal plasma model (the APEC model in XSpec) with kT = 1.2 keV and Z = 0.3 Z_⊙ abundance. The obtained X-ray upper limits are <2.9 × 10⁴³ erg s⁻¹ and <1.0 × 10⁴³ erg s⁻¹ for SDSS J113800.88+521303.9 and SDSS J093540.34+565323.8, respectively. These values are comparable to those obtained for galaxy groups (Lovisari et al. 2015), hence it cannot be excluded that these superluminous disk galaxies are members or possibly the brightest galaxies of groups.

We also probed the morphologies of the five detected galaxy clusters to determine whether they are relaxed or disturbed systems. Lovisari et al. (2017) demonstrated that concentration and centroid shift are the two most robust parameters for probing the morphology of clusters. Therefore, we follow their method to measure these parameters, and list the obtained values in Table 3. We find that the galaxy cluster hosting J16273 is the most relaxed, based on the concentration parameter, and it is also in the range of relaxed clusters based on the centroid-shift parameter. The galaxy cluster hosting 2MASX J14475296+1447030 is the most relaxed cluster based on its centroid shift. The other end of the range is represented by the clusters in the vicinities of 2MASX J09572689+4918571 and 2MASX J14072225+1352512, which exhibit disturbed morphologies.

While we studied the basic properties of all detected clusters in the previous section, we now carry out a more in-depth analysis of the galaxy cluster that hosts a super spiral galaxy in its center. Specifically, we probe its morphology, derive surface brightness and temperature profiles, and further explore whether it is a cool-core or non-cool-core cluster.

In Section 4.5, we demonstrated that this galaxy cluster exhibits relaxed morphology. To further explore the morphology of the cluster, we present the 0.7–2 keV band X-ray image in Figure 4. To extract its surface brightness profile, we define the regions that have constant S/Ns, which result in wider bins on the outskirts of the cluster. The left panel of Figure 5 presents the 0.7–2 keV band azimuthally averaged surface brightness profile of the cluster. The profile does not reveal sharp surface brightness edges, which often signify past interactions or mergers. This evidence suggests the relaxed nature of the cluster.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Based on the surface brightness profile of the galaxy cluster, we also derived its density profile. To this end, we fit the surface brightness profile with a double β-model, where the surface brightness of the cluster is described following Lovisari et al. (2015):

$$\begin{eqnarray*}&&{S}_{{\rm{X}}}={S}_{1}{\left[1+{\left(\displaystyle \frac{r}{{r}_{{\rm{c}},1}}\right)}^{2}\right]}^{-3{\beta }_{1}+0.5}+{S}_{2}{\left[1+{\left(\displaystyle \frac{r}{{r}_{{\rm{c}},2}}\right)}^{2}\right]}^{-3{\beta }_{2}+0.5},\end{eqnarray*}$$

where r_c,1 and r_c,2 are the core radii and S₁ and S₂ are the amplitudes. By fitting this model and assuming spherical symmetry, we model the gas density profile as:

$$\begin{eqnarray*}&&{n}_{{\rm{e}}}(r)={\left\{{n}_{{\rm{e}},1}^{2}{\left[1+{\left(\displaystyle \frac{r}{{r}_{{\rm{c}},1}}\right)}^{2}\right]}^{-3{\beta }_{1}}+{n}_{{\rm{e}},2}^{2}{\left[1+{\left(\displaystyle \frac{r}{{r}_{{\rm{c}},2}}\right)}^{2}\right]}^{-3{\beta }_{2}}\right\}}^{\tfrac{1}{2}}.\end{eqnarray*}$$

The best-fit parameters of the density profile are provided in Table 4, which allows the derivation of the gas density at any radius.

In the right panel of Figure 5, we present the temperature distribution of the cluster. To define the spectral extraction regions, we required an S/N of >30. In the innermost region, we also required a minimum extraction region of 30'' to reduce the impact of the point-spread function of XMM-Newton. To fit the spectrum, we used an optically thin thermal plasma emission model (apec in XSpec), where the temperature, metallicity, and normalization were allowed to vary, but the line-of-sight column density was fixed. The profile reveals that the temperature is somewhat cooler in the core, suggesting that it is a cool-core cluster.

Based on the iterative method introduced in Section 4.5, we found that the R₅₀₀ radius of the cluster is 797 kpc. Within this region, the best-fit temperature is ${kT}={3.55}_{-0.20}^{+0.18}$ keV and the total ICM luminosity is L₅₀₀ = (1.17 ± 0.12) × 10⁴⁴ erg s⁻¹. Based on the kT − M₅₀₀ scaling relation, we infer a total mass of M₅₀₀ = (2.39 ± 0.19) × 10¹⁴ M_⊙. Thus, the superluminous disk galaxy J16273 is the central BCG of a massive cluster.

The galaxy J16273 is the second superluminous disk galaxy that resides in the center of a galaxy cluster. However, this galaxy exhibits a disk morphology, while 2MASX J10405643-0103584, presented in Bogdán et al. (2018a), is a lenticular galaxy. Given that we followed up 12 candidate superluminous disk galaxy BCGs with XMM-Newton, we can now estimate the frequency of galaxy clusters hosting such curious galaxies at their centers.

To identify superluminous disk galaxies, Ogle et al. (2019) searched all the galaxies in NED that had available r-band photometry and a spectroscopic redshift of z < 0.3. Of these galaxies, 1525 galaxies are high-luminosity galaxies with L_r > 8L^*. While 1400 of these objects are elliptical galaxies, 84 and 16 galaxies exhibit spiral and lenticular morphology, respectively. The remaining 25 galaxies are either peculiar or nonspiral (e.g., quasar host) galaxies. As discussed in Section 2 and in Ogle et al. (2019), most of the superluminous disk galaxies reside in low-density environments, and only a fraction of them are cluster members. Based on Bogdán et al. (2018a) and the results of the present work, we conclude that of the 100 superluminous disk galaxies, only two systems, i.e., 2%, are the central BCGs of galaxy clusters, and that most cluster-member superluminous galaxies reside at the outskirts of clusters. Considering all the massive galaxies with L_r > 8L^* within z < 0.3, we conclude that only ≈0.13% are superluminous disk galaxies and central BCGs.

The existence of massive disk galaxies is surprising because massive galaxies grow through a series of minor and major mergers, but major mergers are known to destroy galaxy disks. To explain the existence of massive disk galaxies in the present universe, two formation channels are considered. First, it is possible that a fraction of superluminous disk galaxies have experienced a very quiet merger history. Ogle et al. (2019) analyzed the environments of the superluminous disk galaxies and concluded that most of them reside in relatively low-density environments. In agreement with this, our study also pointed out that only ∼2% of superluminous disks reside in the centers of rich clusters. Therefore, it is feasible that a substantial fraction of superluminous disk galaxies did not undergo major mergers and gradually increased their mass by a series of minor mergers. Additionally, due to their large mass, these galaxies are protected from major mergers since most galaxies in their (low-density) vicinities will have significantly lower masses, resulting in minor mergers.

In the second formation scenario, the original galaxy disks are destroyed, but they are reformed at a later epoch. Because of the high galaxy density in the central regions of clusters, the occurrence rate of major mergers is significantly higher, which in turn is more likely to result in the formation of a spheroid. Therefore, in these environments, it is more likely that the galaxy disks were accreted at a later time (Jackson et al. 2020). In this picture, the presence of an extended disk can be explained by the merger between a spheroid and a gas-rich satellite, where the latter galaxy spins up the spheroid. As a result of this merger, a rotationally supported disk can form. According to the simulations presented in Jackson et al. (2020), about 70% of massive disk galaxies originate from this scenario, and only 30% of them can be attributed to the quiet merger history.

Given the rich environments of cluster cores, it is possible that J16273 also has a reformed disk. This scenario is also consistent with the relatively massive bulge component of the galaxy, which bulge would be sufficiently massive as a traditional spheroid even without its large disk. In Bogdán et al. (2018a), we also speculated that the massive superluminous lenticular galaxy with B/T = 0.66, 2MASX J10405643-0103584, was an elliptical galaxy whose disk was reformed by a gas-rich merger. Because both superluminous disk galaxies in the cluster cores have relatively large B/T ratios, we hypothesize that reforming the disk is the scenario that is more likely to result in the formation of these galaxies. However, in isolated environments or on cluster outskirts, massive disk galaxies could have formed via a quiet merger history.

In the present study, we have probed whether superluminous disk galaxies are BCGs of clusters. However, our analysis could not be extended to galaxy groups. Indeed, our initial selection criteria utilized short ROSAT observations to detect the ICM emission from clusters. Although these data are suitable for hinting at the presence of luminous galaxy clusters at z ∼ 0.2–0.3, fainter galaxy groups remain hidden at these redshifts. Therefore, it is feasible that some of the superluminous disk galaxies are located in galaxy groups or in low-mass clusters.

To probe the superluminous disk galaxies residing in such systems, the eROSITA All-Sky Survey will be ideal. Thanks to its large collective area, after the 4 yr survey it will detect galaxy groups with a few times 10⁴² erg s⁻¹ luminosity within z < 0.3. Although its point-spread function is significantly broader than that of XMM-Newton, these data could still be used to identify the centers of groups. Thus, these data will allow us to probe whether superluminous disks are members of galaxy groups or low-mass clusters, and will reveal whether they reside in the core or on the outskirts of these systems.